Predictive modelling and forecasting of the mosquito abundance and risk of West Nile virus
The transmission and spread of mosquito-borne disease such as West Nile virus (WNV) in north America are weather-sensitive. Weather conditions, such as daily average temperature and precipitation, not only affect the abundance but also the biting behavior of the vector mosquitoes, thus determining the outbreak and spread of WNV. In this talk, I will present a hybrid model for weekly forecasting of the abundance and risk of infection of WNV, integrating a weather-driven statistical model and a compartmental model mimicking the transmission of WNV among mosquitoes, birds, and humans. The predictive models are evaluated using the data of human infection, mosquito surveillance and viral test in Peel Region, Ontario, Canada. With the weekly mosquito surveillance data and the weather forecast, our real time predictive modelling allows us to forecast the weekly vector abundance and risk of WNV infection which has important public health implications. I will conclude with a summary of preliminary dynamics of such hybrid models with weather conditions. This is a joint work with Huaiping Zhu and Nick Ogden in collaboration with Peel public Health.