GH002-13
Modeling Spatio-temporal Spread of Epidemic based on Its Triggered Earthquake Analogy

Monday, 7 December 2020: 11:06
Virtual
Bei Li, National University of Singapore, Civil and Environmental Engineering, Singapore, Singapore, Yunyue Li, National University of Singapore, Singapore, Singapore and Yanwen Wei, Tsinghua University, Beijing, China
Abstract:
Spatio-temporal modeling for the spread of infectious diseases plays a significant role in predicting the epidemic development and policy making for mitigation (Colizza et al., 2006). Although epidemic modelling has been widely applied to derive time series of the infected population, simulations that incorporate spatio-temporal evolution of the epidemic are rarely reported.

We develop a spatio-temporal model to simulate the dynamic spread of an epidemic based on its analogy to the triggered earthquake process. The infectious potential is modelled by wave solutions from the epicentre, which describes the local spreading process. Additional infections are modelled as triggered earthquake sources, which describes the long-range travel-based spreading process. Under high infectious potential, the infected population is calculated based on the number of contacts per person and the infection rate. Government public health policy changes the number of contacts inversely proportional to its stringency, quantified by Oxford University.

The proposed model is applied to Singapore during the period from January 1st, 2020 to July 15th, 2021. Figure 1(a) to (d) shows the snapshots of the COVID-19 spread at Days 40, 70, 100, and 130, respectively. The warm colors indicate high-risk areas due to the concentration of infectious potential. The numbers on the bottom right corner of each subplot are the expected new infection on that day. Our simulation shows that the expected new infection increases from Days 40 to 70, then decreases from Days 70 to 100, and almost drops to 0 at Days 130. Such results are in general agreement with Singapore’s circuit breaker starting from April 7th (Days 98).

The example suggests that the proposed model is able to simulate the spatio-temporal evolution of the epidemic with valid reflection of the government policy. Therefore, it can be used in predicting epidemic spread while testing tentative policies.

<< Previous Abstract | Next Abstract >>