Nonlinear Effects of the Built Environment on Subways at Station Level: Average Travel Distance Changes under the Influence of COVID-19

Abstract

The travel behavior of subway passengers has shown systematic differences with the onset and progression of the COVID-19 pandemic. This study focused on subway stations in Xi’an and systematically analyzed the spatiotemporal features of the average travel distance of passengers (ATDP) during the peri- and post-COVID-19 periods. Using the gradient boost regression tree analysis framework, the study explored the nonlinear relationship between various built environment factors and changes in ATDP during pandemic periods. The results indicated that the distance to the city center, land-use mix, and road network density contributed significantly more than other variables in different periods. Furthermore, the study found that the ATDP in subway stations does not vary significantly based on the structural characteristics of the station network. These findings hold significant value for operational organizers seeking to comprehend the station-level changes in travel distances under the influence of public health crises, offering effective scientific data support for the rational development of operational plans to balance the distribution of operational capacities.