A hierarchical network approach for modeling Rift Valley fever epidemics

Rift Valley fever is a vector-borne disease which causes high morbidity and mortality in livestock and also can infect humans. In the event Rift Valley fever virus is introduced to the United States, mitigation strategies need to be rapidly set in place to prevent the disease from becoming endemic. Presented here is a stochastic network-based mathematical model of Rift Valley fever. Given a lack of empirical data for this region, this di erence equation model uses stochastic parameters following several Pert distributions. A large directed, asymmetric network was established with 3621 nodes in the Uvalde and Panhandle areas of Texas, U.S.A. noting that the climate in Texas can be similar to that of the Rift Valley in Eastern Africa. This model predicts how Rift Valley fever can spread spatially based on biotic and antibiotic factors and identifies the relevant parameters in RVFV propagation. Humans, cattle, Aedes mosquitoes, Culex mosquitoes are independently treated with di erent contact networks to assess virus propagation and spread under di erent scenarios. The importance of the factors a ecting Rift Valley fever virus spread was assessed under di erent scenarios. Spatial e ects and climate factors are also addressed in the model. It has been shown that cattle movement, mosquito population and favorable climate for mosquito development contribute to the spatial spread of Rift Valley fever virus.