Abstract:
Process-based models were constructed for computing the risk of malaria epidemic using temperature and rainfall data. The model has a lead-time of two to four months between detection of the epidemic signal and evolution of the epidemic. Malaria data was collected from eight sites in Kenya, Tanzania and Uganda with temperature and rainfall data from meteorological stations closest to the source of the malaria data. The sensitivity, specificity and positive predictive power were used to validate the models. Results validate the additive and multiplicative models, which were shown to be robust and with high climate-based, early epidemic predictive power.
