Multimodal Transportation (Mar 2025)
Investigating the non-linear influence of the built environment on passengers’ travel distance within metro and bus networks using smart card data
Abstract
Understanding the travel behavior of transit passengers and its influencing factors is crucial for promoting transit use and alleviating urban traffic congestion. However, limited studies have examined the determinants of spatial expansion in multimodal public transportation and overlooked the nonlinear influence between variables. To address these gaps, this study employs the travel distance indicator to portray the spatial expansion of transit passengers. Using smart card data collected from Beijing, China, we propose a comprehensive trip chain extraction method within the metro and bus network, considering transfer behaviors. From the extracted trip chain data, we calculate travel distances and observe significant variations across different transit networks: an average travel distance of 8.09 km in the bus network, 14.93 km in the metro network, and 23.10 km in the integrated network. Further, we explore the non-linear relationship between transit travel distance and the built environment by employing a Gradient Boosting Regression Tree (GBRT) model. The finding reveals that the built environment exerts the most significant influence on travel distance (46.80 %), particularly regarding the distance to the nearest metro station and the central business district (CBD). Additionally, all variables exhibit non-linear effects on travel distance, with many exhibiting relevance only within specific ranges. For instance, there is a noticeable decline in travel distance when the bus stop density falls within the range of 15 units/km² and the bus coverage rate within a range of 0.8. Beyond these threshold values, the decline in travel distance becomes gradual. These findings emphasize the significance of considering non-linear relationships and threshold effects in transit and urban planning. Finally, this study provides practicable recommendations regarding non-linearities for the government that could be beneficial in promoting transit usage.
