Nonlinear impact of built environment on people with disabilities’ metro use behavior

Ensuring equal access to public transportation for people with disabilities is crucial for social equity and sustainable urban development. The physical characteristics and travel preferences of people with disabilities may lead to unique spatial and temporal characteristics of metro use behavior and make them more likely to be influenced by built environment characteristics. We tested this hypothesis by comparing metro travel behavior in Wuhan, China, for people with and without disabilities. The results showed that people with disabilities take more and shorter metro trips than people without disabilities. The relative importance of built environment features in influencing metro ridership displays notable distinctions between people with disabilities and those without, as indicated by the outcomes of the gradient boosting decision tree (GBDT) model. For metro ridership among people with disabilities, street density, the number of medical facilities, and the building plot ratio are the top three influencing variables on weekdays, while the number of shopping centers, medical facilities, and recreational facilities are the top three on weekends. Additionally, these variables exhibit clear nonlinear associations on metro ridership for both groups. The findings offer new insights for providing a better travel environment for people with disabilities and thus promoting transportation equity. © 2024 Elsevier Ltd.