Prediction of seasonal rainfall in the sahel region using empirical and dynamical methods

Abstract Long‐range weather forecasting is a notoriously difficult area of environmental science. However, recent improved understanding of atmospheric dynamics and better observations indicate that useful progress, rooted in scientifically sound ideas, may be possible with long‐range forecasting in the tropics. We describe recent research into the mechanisms and prediction of rainfall in the sub‐Sahara during the main summer rainfall season, concentrating on the Sahel region. We use a complex physical model of the atmosphere (a ‘general circulation’ model) and two relatively simple statistical models to show that large‐scale variations in sea surface temperature (SST) can strongly influence seasonal Sahel rainfall. Persistence of patterns of SST anomalies (deviations from long‐term average) is sufficient to allow useful forecasting techniques to be based on fields of SST anomalies observed in the preceding spring. However, persistence of the SST anomalies may not always be sufficient to provide a skilful forecast.