Periodic synchronisation of dengue epidemics in Thailand over the last 5 decades driven by temperature and immunity
Fig 2
WMFs for (a) temperature, and ln-transformed dengue cases from the (b) passive surveillance data and (c) model output. Cross-WMFs between (d) temperature and ln-transformed dengue cases (data) and (e) temperature and ln-transformed dengue model output. Model output assumes a mean cross-protection of 1 year (see S13 Fig in S1 Appendix for results using other mean cross-protections). The same temperature time series are used for both (d, e). Grey bars above panels indicate the times for which phases (in a–c) and phase difference angles (in d, e) are highly consistent across provinces and statistically significant (see section “Material and methods”). In (a–c), higher values in the mean fields indicate timescales and points in time where the phases are more consistent across provinces, and where the amplitudes of oscillations are more correlated. In (d, e), higher values correspond to timescales and points in time where the agreement between dengue and temperature is itself more consistent across provinces. S14 Fig in S1 Appendix shows which multiannual timescales dominate the (C) WMFs for each panel in this figure. Edge effects in the WTs may influence results before and after the dashed lines (see section “Materials and methods”). The underlying data are in S1 Data at https://github.com/UF-IDD/synchrony_dengue_figures. WMF, wavelet mean field; WT, wavelet transform.