Unraveling the Multi-Scale Spatial Relationship between Built Environment and Walk Access to Metro Stations: A Case Study in Nanjing

In order to better understand how the built environment affects the ratio of walk access to metro stations, a multiscale geographically weighted regression (MGWR) model was used to analyze the spatial relationship. This paper selected 72 metro stations in Nanjing city as the research objects, and calculated the “5D” elements of built environment as independent variables by using open-source data. A sampling survey of access mode was taken to calculate the ratio of walk access at each station during morning peak as the dependent variable. Model results show that compared with multiple linear regression model (MLR) and geographically weighted regression (GWR) model, the proposed method has better explanatory power. At the same time, the built environment factors including entropy, intersection density, and the distance to city center, have significantly different influence patterns between the traditional downtown area and the new urban area.