关键词: 空间交互网络, 空气传播疾病, 个体接触, 疫情防控, 社团结构, 公共交通

Abstract:

Transportation is a necessary way to realize human mobility. Spaces in transportations, such as buses and subways, are confined and crowded, which could induce a large volume of physical contact between passengers and thus promotes disease spread between areas with no human connection. Mobility and physical contact are widely regarded as two critical components in disease spread over space. However, current studies only considered a single aspect, human mobility or physical contact, in identifying high-risk areas of airborne disease. Without comprehensive consideration of human mobility and physical contact, health planners may misidentify areas with a high risk of airborne disease and design less-effective interventions. To fill the gaps, this study first proposed a theoretical framework that comprehensively considers human movement and physical contact during movement to construct a "movement-contact". Using public transit system in Beijing, we built "movement-contact" network with the Thiessen polygon centered by bus station as the node and the number of individuals moved or contacted as weight. Then, the weighted degree centrality of each node was calculated. Further, the K-shell decomposition algorithm was used to divide the importance degree of each node, thereby extracting the nodes (areas) with a high risk of disease transmission. Accordingly, we offer some suggestions to control the spread of the disease. Communities, despite being the basic unit of human activity, are rarely used to evaluate the effectiveness of disease control measures. To fill the gap, the effectiveness of control measures was evaluated by comparing the modularity of community structure before and after taking control measures. The results showed that there were three types of high-risk areas. The first type is the intercity transportation hubs, such as Beijing Capital International Airport, Beijing South Railway Station, and Beijing West Railway Station. The second type is some residential areas, e.g. Qinghe, Shahe, and Anhuiqiao. The third type is the business centers represented by Wangjing, Sanlitun, and Guomao. Based on the findings, this study suggests stopping bus services in high-risk areas to impede the spread of disease. After taking control measures, the modularity of the community structure reduced significantly. This study is of great significance for identifying high-risk areas of airborne disease, developing prevention measures for disease transmission, and evaluating the effectiveness of control measures. It provides some suggestions to prevent the spread of disease in a timely manner under the promise of guaranteeing the normal life of residents.

Key words: spatial interaction network, airborne disease, physical contact, epidemic prevention and control, community structure, smart card data