Publication by: Gitau et al.

Publication Date: 2014
Abstract: Despite earlier studies over various parts of the world including equatorial Eastern Africa (EEA) showing that intraseasonal statistics of wet and dry spells have spatially coherent signals and thus greater predictability potential, no attempts have been made to identify the predictors for these intraseasonal statistics. This study therefore attempts to identify the predictors (with a 1-month lead time) for some of the subregional intraseasonal statistics of wet and dry spells (SRISS) which showed the greatest predictability potential during the short rainfall season over EEA. Correlation analysis between the SRISS and seasonal rainfall totals on one hand and the predefined predictors on the other hand were initially computed and those that were significant at 95% confidence levels retained. To identify additional potential predictors, partial correlation analyses were undertaken between SRISS and large-scale oceanic and atmospheric fields while controlling the effects of the predefined predictors retained earlier. Cross-validated multivariate linear regression (MLR) models were finally developed and their residuals assessed for independence and for normal distribution. Four large-scale oceanic and atmospheric predictors with robust physical/dynamical linkages with SRISS were identified for the first time. The cross-validated MLR models for the SRISS of wet spells and seasonal rainfall totals mainly picked two of these predictors around the Bay of Bengal. The two predictors combined accounted for 39.5% of the magnitude of the SST changes between the July–August and October–November–December periods over the Western Pole of the Indian Ocean Dipole, subsequently impacting EEA rainfall. MLR models were defined yielding cross-validated correlations between observed and predicted values of seasonal totals and number of wet days ranging from 0.60 to 0.75, depending on the subregion. MLR models could not be developed over a few of the subregions suggesting that the local factors could have masked the global and regional signals encompassed in the additional potential predictors.

Download the full report

Publication by: Camberlin et al.

Publication Date: 2009
Abstract: The inter-annual and spatial variability of different rainfall variables is analysed over Equatorial East Africa (Kenya and northeastern Tanzania). At the station level, three variables are considered: the total precipitation amount (P), the number of rain days (NRD) and the daily rainfall intensity (INT). Using a network of 34 stations, inter-station correlations (1958–1987) are computed for each of these variables. The spatial coherence of monthly or seasonal P and NRD is always much higher than that of rainfall intensity. However, large variations in spatial coherence are found in the course of the seasonal cycle. Coherence is highest at the peak of the short rains (October–December) and low during the long rains (March–May), except at its beginning. The inter-annual variability of the onset and cessation of the rains is next considered, at the regional scale, and the study extended to 2001. In the long rains, the onset and cessation dates are independent of NRD and INT during the rainy season. Hence, the long rains seasonal rainfall total depends on a combination of virtually unrelated factors, which may account for the difficulty in its prediction. However, the onset, which exhibits a large inter-annual variability and a strong spatial coherence, has a prime role. Conversely, in the short rains, though the onset is again more decisive than the cessation, the different intra-seasonal descriptors of the rains are more strongly inter-related.

Download the full report

Publication Date: 2014
Abstract: This study evaluated the skill of forecasting seasonal rainfall over the Greater Horn of Africa (GHA) using Ensemble Model Technique from a cluster of four General Circulation Climate Models (GCMs) from Global Producing Centres (GPCs). The spatial distribution of rainfall anomalies of the observed models output during extreme events showed that the ensemble model was able to simulate El-Niño (1997) and La-Niña (2000) years. The ensemble models did not show good skill in capturing the magnitude of the extreme events. The skill of the ensemble model was higher than that for the member models in terms of its ability to capture the rainfall peaks during the El-Niño Southern Oscillation (ENSO) phenomena. The analysis for the correlation coefficients showed higher values for the ensemble model output than for the individual models over the Equatorial region with the stations in the northern and southern sectors of the GHA comparatively giving low skill. The ensemble modeling technique significantly improved the skill of forecasting, including the sectors where individual models had low skill. In general, the skill of the models was relatively higher during the onset of the ENSO event and became low towards the decaying phase of the ENSO period. Generally, the study has shown that the ensemble seasonal forecasting significantly adds skill to the forecasts especially for October-December (OND) rainy seasons. From the study, ensemble seasonal forecasting significantly adds skill to the forecasts over the region. Blending dynamical ensemble forecasts with statistical forecast currently being produced during Regional Climate Outlook Forums (RCOFs) would add value to seasonal forecasts. This significantly reduces the impacts and damages associated with climate extremes over the region.