Fig. 3: Comparison of the temporal coevolution of malaria cases, regional temperatures and rainfall, and remote climate drivers, for different components (low-frequency and seasonal).

A Low-frequency changes in P_f cases, P_v cases, and T_min (the composite time series from 11 stations, DZ_reg), show concomitant variation and the existence of a slowdown in both malaria and temperatures. B The respective seasonal components of rainfall (R) and minimum temperature (T_min DZ_reg) covary. Note the dip in both variables around the turn of the century and the concordance of the multi-annual envelope. C The respective seasonal components of P_v and P_f cases also covary, including the reduction in amplitude for both variables around the turn of the century. Note the long-term envelope of the seasonal cycles and its concordance with the trend of T_min in A. The time period indicated by the black box corresponds to the years at the turn of the century, the focus of our investigation. All components shown were significant at p < 0.005 against both white noise and a red-noise null model (see “Methods” for technical details).