A spatiotemporal transmission simulator for respiratory infectious diseases and its application to COVID-19

Abstract

The present study introduces a transmission dynamic simulator for respiratory infectious diseases by incorporating human movement data into a spatiotemporal transmission model. The model spatially divides areas into multiple patches according to administrative regions. The transmission of respiratory pathogens within each patch is depicted using an improved Susceptible-Exposed-Infectious-Removed (SEIR) compartmental framework, which incorporates quarantine and isolation measures. The risk of transmission between patches is determined by a gravity-constrained model that considers passenger volume and the spatial distance between patches. We simulate changes in intervention policies and detection methods by adjusting quarantine and detection rates at different stages of the epidemic, thereby capturing spatial variations in pathogen transmission through altering the transmission rate. Ultimately, we apply this simulator to accurately replicate the spatiotemporal dynamics observed during the initial COVID-19 outbreak across all 31 provinces in the mainland of China, successfully capturing the temporal variations in both case numbers and affected provinces. Additionally, it demonstrates a remarkable level of accuracy in predicting the outbreak of epidemic in each province.