Seasonal forecasting of African rainfall: Prediction, responses and household food security

The main approaches, assumptions and methods used in seasonal forecasting are described in this paper. Examples of methods ranging from simple correlation to multi-annual, multivariate models of seasonal rainfall prediction are given for several regions in Africa along with an overview of operation forecasts for the Sahel, East Africa and Southern Africa. Recent developments in climate prediction suggest that seasonal rainfall forecasts for Africa are increasingly reliable and should be of widespread interest to resource managers and consumers. Climate forecasts may indeed revolutionize resource management in Africa. Yet, their utility depends on the linkages between geophysical, economic and social aspects of resource use. Progress in rainfall forecasting is placed in the context of the use of seasonal predictions in Africa, with a particular emphasis on ameliorating vulnerable livelihoods. Targeting users, reaching vulnerable livelihoods, messages that are distorted over space and time, the lag between forecasts and dissemination, maladaptive responses and false alarms are difficulties that can be expected in many developing countries.