Contribution of built environment factors and their interactions with subway station ridership

Exploring the built environment's impact on subway station ridership can aid in developing effective built environment update strategies. Previous studies lack an analysis of the impact of the interaction effects of built environment explanatory variables on subway station ridership. Beijing is divided into three zones with different buffer scales, and 18 built environment explanatory variables are selected as independent variables based on the ‘7D’ dimension of the built environment, and these are calculated based on a hypothetical circular scale range centered on a subway station. The inbound ridership of subway stations during morning peak hours, outbound ridership of subway stations during morning peak hours, inbound ridership of subway stations during evening peak hours, and outbound ridership of subway stations during evening peak hours are taken as dependent variables. For different dependent variables, the optimal parameters-based geographic detector (OPGD) model determines the recommended scale combination of the built environment around subway stations for three zones. Moreover, the impact of single explanatory variables and the interaction of explanatory variables on subway ridership were explored at the recommended scale combination based on the OPGD model. The results show that: (1) the recommended circular buffer radius combinations for the inbound ridership of subway stations during morning peak hours, outbound ridership of subway stations during morning peak hours, inbound ridership of subway stations during evening peak hours, and outbound ridership of subway stations during evening peak hours are 800–800–2000 m, 800–1000–2000 m, 800–1000–2000 m, and 800–800–2000 m, respectively. (2) The density of apartment facilities and the density of office facilities are the variables that contribute substantially to the ridership of Beijing subway stations. (3) The interaction between multiple explanatory variables has a much stronger contribution to ridership than single factors. In particular, the contribution of the explanatory variables to inbound ridership during the morning peak and outbound ridership during the evening peak increased significantly. Based on the analysis results, targeted built environment updating strategies are provided from the perspective of supply–demand balance, which can provide an important decision-making basis for updating the built environment around subway stations. In addition, it can also provide a theoretical basis for delineating the scope of transit-oriented development (TOD) in Beijing.