The climate and dengue relationship were examined using different temperature measures (maximum, mean and minimum temperature) in the DLM. The deviance was used to choose the best model. Model including maximum temperature was associated with lower deviance value (Supplementary Table 2). We also compared the deviance for the association between each weather variables and dengue using DLM. The deviance was also lower compared to other models when maximum temperature and relative humidity were included (Supplementary Table 3). The goodness-of-fit was performed to check the model adequacy using auto-correlation functions of residuals and normality of the residuals. Fig. 1 shows that there was no significantauto-correlation between residuals at different lags in the DLM when maximum temperature and humidity were used as predictor variables. The scatter plot shows that the residuals in the model fluctuated randomly around zero with no obvious trend. Thus the goodness-of-fit analyses show that the model fits the data reasonably well. Therefore, we selected the model including maximum temperature and relative humidity as the best model and used it to estimate the effects of weather variation on dengue transmission.