Spatially Explicit Modeling of 2019-nCoV Epidemic Trend based on Mobile Phone Data in Mainland China

Xiaolin Zhu; Aiyin Zhang; Shuai Xu; Pengfei Jia; Xiaoyue Tan; Jiaqi Tian; Tao Wei; Zhenxian Quan; Jiali Yu

Our proposed model stems from the SIR model, a classic approach to simulate epidemiological dynamics. We modified the SIR structure based on the unique characteristics of the outbreak of 2019-nCoV. First, in all cities other than Wuhan, the initial infectious cases are most likely imported from Wuhan 10 . Second, many Wuhan residents moved to other cities due to the Spring Festival and this mobility was closed after the quarantine on January 23. Third, those people from Wuhan have low contacts with local residents because Chinese government required them to implement self-isolation. Last, during the past 40 days, all cities took efforts to control the virus spreading, which slows down the daily increase of new infections ( Figure 1 ). Accordingly, in the modified SIR model, the susceptible variable was divided into two groups: 1 , the number of local susceptible, and 2 , the number of susceptible with Wuhan travel history. These Wuhan-inbound groups ( 2 ), have transmission rate ( 2 ) different from transmission rate ( 1 ) of local residents ( 1 ) as Chinese government took measures to reduce the person-to-person contacts. A decay rate a was introduced to tune the value of β in each day, accounting for the gradual impact of prevention interventions in each city. In our modified SIR model, recovered population R was extended to include those cured, died, and isolated in hospital because they cannot transmit the virus. The differentiate equations of our modified SIR model is as follows: