Diagnostic Study of Seasonal Prediction of Malaria: A Case Study in Senegal, West Africa

Abstract This study aims to improve malaria predictability in Senegal using sea surface temperatures (SSTs) data. It assesses malaria predictability during the September–November period using the canonical correlation analysis (CCA) and two malaria models: the Liverpool Malaria Model (LMM) and the vector-borne disease community trieste (VECTRI). Both models use CPC-gridded global air temperature and African Rainfall Climatology, version 2 (ARC2). SST predictors from the Extended Reconstructed SST, version 5 (ERSSTv5), and the North American Multi-Model Ensemble (NMME) are incorporated. The study compares observed malaria data from the National Malaria Control Program (PNLP) in Senegal with VECTRI model outputs, validating the models against real-world malaria cases. Focusing on the tropical Pacific (TROP_PAC), Gulf of Guinea (GG), and tropical Atlantic (TROP_ATL) ocean basins, which influence the West African monsoon and malaria outbreaks, the research covers 1982–2010. Negative SST anomalies in TROP_PAC correlate with increased malaria transmission due to increased rainfall, while a dipole pattern in TROP_ATL corresponds to decreased malaria transmission in Senegal. The study finds a strong correlation, with a coefficient of 0.7 for TROP_PAC SSTs 5 months in advance and significant correlation scores nearing 0.6 for TROP_ATL with a 5-month lead time. Furthermore, the study identifies significant lagged correlations in parts of the Indian Ocean, suggesting its potential influence on West African rainfall during the July–September (JAS) season. This finding underscores the complex interactions between oceanic basins, indicating that variations in the Indian Ocean can also affect West African rainfall and, indirectly, malaria incidence by interacting with Pacific Ocean influences. These findings provide valuable insights for planning malaria prevention and control initiatives in Senegal, improving public health strategies. The research faced challenges due to complex data and limited existing research.