A climate-based model for malaria prediction in southeastern Africa

Malaria is a major health problem in southeastern Africa. In this study, we explore relationships between malaria and regional climate. Malaria incidence data from the eastern border of South Africa and Swaziland over a 33-year period were de-trended and used to identify epidemics. Composite weather maps were then constructed for seasons with high and low malaria incidence and evaluated. Surface-air temperature rose over the east coast during malaria epidemics and rainfall doubled over a large area including Swaziland, Zimbabwe and southern Mozambique. Remote climatic signals that anticipated malaria epidemics were found in composite analyses, and a statistical model was developed for prediction. Upper-level winds over the Western Pacific were found to predict 57% of malaria variance at a lead time of 6 months.