PROGRESS IN GEOGRAPHY ›› 2021, Vol. 40 ›› Issue (12): 2048-2060.doi: 10.18306/dlkxjz.2021.12.006

• Articles • Previous Articles     Next Articles

Impact of the built environment on residents’ commuting mode choices: A case study of urban village in Wuhan City

TONG Zhaomin1(), AN Rui1, LIU Yaolin1,2,3,*()   

  1. 1. School of Resource and Environmental Science, Wuhan University, Wuhan 430079, China
    2. Key Laboratory of Geographic Information System, Ministry of Education, Wuhan University, Wuhan 430079, China
    3. Collaborative Innovation Center for Geospatial Information Science, Wuhan University, Wuhan 430079, China
  • Received:2021-01-18 Revised:2021-05-25 Online:2021-12-28 Published:2021-12-24
  • Contact: LIU Yaolin;
  • Supported by:
    National Natural Science Foundation of China(41771432)


As a space with farmers' characteristics and reshaped by mixed classes, urban villages provide a guarantee for the vulnerable groups to integrate into the urbanization system and realize the anti-capital but not anti-market urbanization process. Exploring the impact mechanism of the built environment on residents' commuting mode choices and carrying out inclusive renovation of the built environment of urban villages will be of great significance for addressing the commuting and employment difficulties of vulnerable groups. Based on the location based services (LBS) data of Wuhan City in 2018 and taking the choice of commuting mode as binary dependent variable, this study used the gradient boosting decision tree to explore the nonlinear relationship between the built environment and commuting mode choices of urban village residents and threshold effect. The results show that the residents of urban villages who commute for a long distance generally depend on the public transport system, and the nonlinear influence and threshold effect of the built environment are obvious, which make up for the insufficient interpretability of the traditional linear model. The contribution of bus station accessibility is the largest (19.89%) in prediction, and the thresholds are about 300 m and 1000 m respectively. In urban village reconstruction, bus station within a Euclidean distance of 300 m should be satisfied as far as possible, there must be bus lines within 1000 m, and other variables have similar interpretability. The impact priority level and the most effective spatial attribute range of the built environment obtained from this study will provide a reference for improving the employment and commuting equity of urban village residents and for realizing regional inclusive and coordinated urbanization.

Key words: urban village, commuting mode choices, gradient boosting decision tree, nonlinear relationship, threshold effect