Seasonal rainfall forecasting for Africa

Abstract

An experimental seasonal rainfall forecast for the African Sahel as a whole has been issued by the UK Met. Office for each year since 1986. The forecast is issued in the May or June preceding the wet season (mainly the months July to September). Currently, this forecast is communicated directly to National Met. Agencies in the African Sahel who make various uses of it. For the experimental forecast to become operational, several questions need to be addressed: to what extent can the spatial and temporal resolution of the forecast be refined? can the likelihood of a particularly poor forecast be estimated in advance? how best, and to whom, should the forecast be communicated in order to maximise usefulness? what potential benefits are likely to accrue from operational forecasts? and how can the abuse of forecast information be minimised? This paper defines some preliminary answers to these questions drawing upon a range of disciplinary perspectives. In Part I, the current, and possible future, status of the rainfall forecast is summarised and placed in the context of existing climate information in Africa. In Part II the application of the forecast and its possible impact on five groups of potential users of these rainfall forecasts is discussed. It is suggested that forecasts will have their most immediate use at national and international level, but that improvements in institutional efficiency and interaction will need to be made before the potential benefits of the forecasts can be realised. Rural communities are least likely to obtain direct benefits from the forecast, but potentially there are indirect benefits to such communities if the information is correctly used at other levels.