Abstract: The traffic accessibility can be observed by studying the spatial connection of traffic network and the degree of connectivity between network nodes. In this paper, the three-dimensional topological network composed of underground rail transit and road traffic is modeled. The accessibility model is established based on railway stations using space syntax morphology analysis variables. Then, this paper analyzes the accessibility level and spatial characteristics of railway stations, and discusses the relationship between the station accessibility and the surrounding land use. The results show that the average accessibility of the railway stations in Shenzhen is 0.088 4, with a low dispersion degree. There have developed four high-value hotspots, namely Futian Central District, North Huaqiang Business District, Dongmen Business District and Houhai District. The business and office land have the highest correlation with the accessibility of the stations, while the industrial land is less affected by the stations. So the railway stations tend to locate at spatial nodes like business districts or office areas where the traffic is more concentrated. The railway stations can basically meet the development requirements of the surrounding land. 27 stations of the "high-medium" type or the "medium-low" type need to improve the functions of the integrated traffic system. This paper evaluates the accessibility of subway stations in Shenzhen and its relationship with the usage of land based on the coupling of the topological network. It proposes a research perspective of transformation from traffic lines to traffic stations, providing a reference for similar studies.

Key words: subway station, accessibility, space syntax, topological network, GIS