Potential predictability of the sea‐surface temperature forced equatorial East African short rains interannual variability in the 20th century

• Published in: Quarterly journal of the Royal Meteorological Society 2015-01, Vol.141 (686), p.16-26
• Main Authors: Bahaga, T. K., Mengistu Tsidu, G., Kucharski, F., Diro, G. T.
• Format: Article
• Language: English
• Subjects: climate variability; East Africa; IOD; mechanism; predictability; short rains
• ISSN: 0035-9009; 1477-870X
• DOI: 10.1002/qj.2338

In this article, the predictability of the 20th century sea‐surface temperature (SST) forced East African short rains variability is analyzed using observational data and ensembles of long atmospheric general circulation model (AGCM) simulations. To our knowledge, such an analysis for the whole 20th century using a series of AGCM ensemble simulations is carried out here for the first time. The physical mechanisms that govern the influence of SST on East African short rains in the model are also investigated. It is found that there is substantial skill in reproducing the East African short rains variability, given that the SSTs are known. Consistent with previous recent studies, it is found that the Indian Ocean and in particular the western pole of the Indian Ocean dipole (IOD) play a dominant role for the prediction skill, whereas SSTs outside the Indian Ocean play a minor role. The physical mechanism for the influence of the western Indian Ocean on East African rainfall in the model is consistent with previous findings and consists of a gill‐type response to a warm (cold) anomaly that induces a westerly (easterly) low‐level flow anomaly over equatorial Africa and leads to moisture flux convergence (divergence) over East Africa. On the other hand, a positive El Niño–Southern Oscillation (ENSO) anomaly leads to a spatially non‐coherent reducing effect over parts of East Africa, but the relationship is not strong enough to provide any predictive skill in our model. The East African short rains prediction skill is also analyzed within a model‐derived potential predictability framework and it is shown that the actual prediction skill is broadly consistent with the model potential prediction skill. Low‐frequency variations of the prediction skill are mostly related to SSTs outside the Indian Ocean region and are likely due to an increased interference of ENSO with the Indian Ocean influence on East African short rains after the mid‐1970s climate shift.