Redai dili (Jan 2021)
A Simulation Study on the Influences of Job-Housing Spatial Relationship on Spreads of Infectious Diseases
Abstract
The novel coronavirus pandemic in early 2020 has been profoundly transforming people's lives. In China, countermeasures such as city and community lockdowns were successful in mitigating the epidemic. Such measures predominantly take effect by limiting close contacts between humans through work and school closures, for instance. However, high concentrations of human interactions are a feature of large urban areas, and daily commuting is a major source of the interactions, determined by the job-housing spatial relationship of the city. This study aims to explore the effects of optimizing the urban job-housing spatial relationship on mitigating disease spread. Taking Shanghai as an example, a multi-agent simulation model was established using NetLogo with a spatial granularity of 523×523 m grids to simulate disease spread through the daily basic activities of individual residents, including residence, working and commuting. The data used for the simulation were based on the current job-housing spatial relationship reflected by mobile phone grid data, from which a sample of 20 000 people with home and job locations was extracted to simulate the process of disease spread from 9 designated disease source locations. The selection of the source locations considered different location types of the Shanghai region, covering the central city area, the near suburbs and the outer suburbs. The measure of job-housing spatial relationship optimization was based on excess commuting. Using a method to simulate individual exchanging residences according to the principle of Pareto Optimality, the current excess commuting of Shanghai was found to approximately reach 69%. The job-housing spatial relationship was then measured as the extent of the excess commuting reduction. Analyses on the relationships between the optimization and the disease spread in space and time were conducted based on 20 simulations under each scenario of the 9 sources. The results showed that the proportions of the infected agents develop in S-shape curves with time, and the fastest spreads occur in the commuting periods. Diseases originating from sources that are closer to the city center spread faster than those originate from farther sources. Although 75% of the optimization constitutes the turning point for the job-housing spatial relationship to significantly postpone the spread, before this point, the influence of job-housing optimization on the spread time is limited. Such optimization can steadily reduce the spatial area of the spread, limit the scope of spread, localize the spread, and change its spatial mode from the "width-first mode" to the "depth-first mode." When the disease source is in the outer suburbs, the localization effect is particularly strong. Based on these findings, methods like job-housing balance, decentralizing jobs from central urban areas, and multi-center regional structures that can reduce commuting distances may constitute appropriate long-term strategies for urban epidemic intervention. The major contributions of the study are the exploration on the spatio-temporal regularities of infectious disease spread and epidemic intervention from the perspective of urban spatial structure, and the use of mobile phone data to provide relatively realistic base environment for disease spread simulation.
Keywords
