Ensemble Forecasts of Seasonal Dengue Epidemics

As a common vector-borne disease, dengue fever still remains a lot of challenges to forecast for which the significant distinction of epidemic scale is affected by multiple factors, such as mosquito density, meteorological conditions, geographical environment, travel and so on. To track down the epidemic scale and forecast the remaining time of epidemic season, the population size affected by the epidemic is evaluated before the compartmental model is optimized by assimilation observation with filtering method. In retrospective forecast of dengue pandemic for Guangzhou from 2014-2015 seasons, accurate forecast of dengue cases is generated with an accurate prediction of peak time in all time periods. The real-time forecast system shows a good performance on capturing the trajectory of dengue transmission and scale of epidemic.