To explore the factors influencing the passenger flow of metro stations and the spatial heterogeneity of these factors,this study focuses on the metro stations in Nanjing.Thiessen polygons are used to delineate the service areas of the stations and identify the influencing factors within these areas,including Points of Interest(POI),road network density,and resident population.Spatial autocorrelation analysis and multicollinearity analysis are conducted on the independent variables.OLS,GWR,and MGWR models are constructed based on multiple data sources,and the results of different models are compared.The findings indicate that the degree of the station and the road network density around the station are local variables,while the resident population and shared bicycle usage around the station are global variables.The MG-WR model outperforms the OLS and GWR models in terms of interpretability and fit.Increasing the road network density in areas with high population density and unsaturated road network density can effectively improve the passenger flow of metro stations in those areas.Adding connecting bus routes in areas with high population density can enhance the conven-ience of interconnection within the public transportation system.It is recommended to increase the number of shared bicy-cles in areas with low passenger flow at metro stations to improve the passenger flow.
mass transitmetro station passenger flowspatial heterogeneitymultiscale geographic weighted regressionthiessen polygon
NETL
NSTL
万方数据
