Climate and disease data sets used in the study.

All anomaly indices were computed as monthly departures from their respective climatological values (long-term means) defined by the periods shown above. NINO 3.4 SST index was computed by the National Oceanic and Atmospheric Administration Climate Prediction Center (NOAA/CPC) as part of operational ENSO monitoring activities. We computed the WIO index directly from the global SST data based on previous research by Linthicum et al. (1999). SPOT Vegetation data were processed by Vlaamse Instelling voor Technologisch Onderzoek (VITO) in Belgium into 10-day composite data. Monthly composites, long-term means, and anomalies from these data were processed by the NASA/Global Inventory Modeling and Mapping Studies (GIMMS) group.