The impact of COVID-19 on subway passenger flow in Chicago: A study of spatial variation of influencing factors

Abstract

At the beginning of 2020, with the rapid spread of COVID-19 around the world, the passenger flow of subway has suffered from a serious impact. Based on the subway passenger flow data in Chicago, this article analyzes the impact of COVID-19 on rail transit passenger flow. ArcGIS is used to visualize the spatial-temporal distribution of the passenger flow of different stations during different time periods. Based on the fluctuation characteristics of passenger flow before and after the outbreak of COVID-19, one of the deep learning methods, the LSTM (Long-Short Term Memory) neural network model, is constructed to predict the passenger flow of each station in the scenario of no virus. The decline of passenger flow is calculated for each station. Stepwise regression model is constructed to determine factors that explain the decline in passenger flow, and significant factors are obtained: the original passenger flow, number of houses and jobs within 800m buffer zone, number of bus stops within 800m buffer zone, whether the station is a transfer station, distance from the station to the city center, and the number of low-income people. The results of the study show that after the outbreak of COVID-19, the passenger flow of the subway in Chicago experience a “cliff-like” decline in the short term. The passenger flow in most areas dropped by more than 80%, and the passenger flow of some severely impacted stations dropped by more than 90%. Characteristics of the station and built environment factors of different stations influence the decline of passenger flow.