Quantifying the effects of built environment on travel behavior in three Chinese cities during COVID-19

The COVID-19 pandemic has directly impacted human travel behavior. Studies have revealed the relationship between the built environment and infection risk. The impact of the built environment on the individual travel behavior of confirmed COVID-19 cases and the differences in impact across different cities require further investigation. To fill this gap, using the reported individual-level activity data of confirmed cases in three Chinese cities, this study applies the gradient boosting decision tree method to quantify and compare the relative importance and nonlinear effects of built environment characteristics and sociodemographic characteristics on travel frequency and travel purpose. The results show that age is the most important predictor, and it has an inverted U-shaped relationship with travel frequency. However, the impact of age on travel purpose varies depending on the city. Built environment characteristics have a greater impact on travel purpose than on travel frequency with heterogeneities observed across different cities. These findings reveal the heterogeneity in the impacts of the built environment on travel behavior in cities of different sizes and across different regions within cities, and can help city planners and administrators develop flexible and resilient strategies to mitigate the impacts of future sudden infectious disease outbreaks.