Rainfall variability has become a major agricultural issue in sub-Saharan Africa, especially since crop production is mainly rainfed. Irrigation technologies are expensive and their implementation is slow. Many researchers now believe that some understanding of the causes of rainfall variability would lead to measures that could be used to investigate reduction in total rainfall and/or drought effects.
There is now ample evidence that rainfall in many parts of Africa can be linked to global circulation phenomenon. Ogallo (1988) has shown that rainfall in Kenya is influenced by the Southern Oscillation Index (SOI). West African rainfall especially in Sahel has been known to be linked to sea surface temperature (SST) in the Pacific (Hulme et al. 1992). To be useful for agricultural decision-making, four conditions must be met: (1) sufficient correlations must exist between global circulation phenomena and local rainfall (2) evidence that indeed crop yields differ for the different ENSO phases (3) forecast period must have sufficient lead time before the cropping season commences and (4) ability to translate forecasts into management decisions (e.g. crop choice, planting date, fertilizer application, etc.).
Out of the four issues, it is only in case (1) that there is evidence of research progress in Ghana. Opoku-Ankomah and Cordrey (1994) showed a significant correlation between simultaneous Atlantic SSTs and rainfall in many parts of Ghana. With the view of forecasting seasonal rainfall, Adiku and Stone (1995), in another study, investigated the relationship between the SOI phase established in April and seasonal rainfall (April to July) and obtained a significant correlation for some sites located along the southern coasts of Ghana. However, April-based SOI does not provide sufficient lead time for effective agricultural planning.
Research efforts must now be directed towards establishing relationships between Global Circulation Indices and seasonal rainfall in Ghana having appreciable lead time of at least three months. Some studies by Adiku (2003) seem to suggest that the October-November-December (OND) SSTs in the Niño 3 Pacific region popularly referred to as ENSO may offer a possibility for seasonal rainfall forecasting in southern Ghana with an appreciable lead time.
In this study, we explore further this relationship with a view to identify zones where the forecast skills are significant. We also aim at demonstrating, using coupled climate/crop modeling, that the OND ENSO phase affects the yields of peanut and maize at some localities. Finally, we propose a working scheme to operationalize ENSO for agricultural planning.
Was this article helpful?