关键词: 空间句法, 街道网络结构, 中心度指数, 土地利用, 空间尺度

Abstract:  Previous studies have evidenced that the street network structure is closely related to land use distribution. However, most previous studies only focus on the influences of the global and local centrality indexes on land use distribution. They lack a comprehensive consideration of indices at different scales. In addition, previous studies mainly analyze the relationship between the centrality index and the distribution of commercial functions without conducting a comprehensive investigation on land use. In this study, street network characteristics were expressed by the street centrality index at different scales, namely, proximity and betweenness. Based on previous research conclusions, the relations of street network characteristics with land use intensity and the distribution of functional mix degree were analyzed through correlation and regression analyses under the guidance of the space syntax theory. Moreover, influences of street network morphology at different scales on the distribution of functions and social economic activities were discussed, proposing collaborative distribution strategies. Results can provide references to improve space utilization efficiency. 
　　A case study based on Binjiang Avenue with a grid road network and the Huayuan District with a tree-like road network in Tianjin was carried out. Findings revealed that the street centrality was related to the land use distribution, but the correlation degree influencing the direction and dominant indices varied with street structural characteristics. At Binjiang Avenue, the area with high construction intensity was positively related to betweenness in the mesoscale. Therefore, the land use distribution in this area was “extraversion” to some extent relative to the geographic center away from the region. In other words, it is mainly oriented to primary and secondary urban arteries that connect with other regions and serve the city or for zoning purposes. The mixing degree was positively related to proximity in the mesoscale and microscale, showing “introversion” characteristics. This implied that it is mainly oriented to local access roads, serving residents. The high construction intensity and high mixing degree in the Huayuan District presented relatively consistent distributions, mainly distributing at two sides of local accessible streets. These roads are mainly urban roads that connect residential groups in the region or roads in residential communities. Such distribution characteristics provide convenient and accessible traffic conditions and diversified services for the short-distance travels of residents, conducive to increasing land utilization efficiency. 
　　The above conclusions interpreted to some extent that regions have different centrality at different scales and present different location characteristics by connecting streets in different ways. Different street structures correspond to different land use distribution patterns. Further, the land use distribution is determined by the space structure to use groups with different location advantages and services. Hence, further research should explore the location of the region in the urban system, including the distribution patterns of accessibility and trafficability through street centrality at different scales. The research should judge its services and attraction to pedestrian and vehicle flow rates in different ranges when determining land use distribution patterns. This should be done to increase the coupling between land use layouts and space locations at different scales, thereby increasing land use efficiency and regional vitality. In future studies, the land use types can be further divided, as can the influences of centrality at different scales on various land use types like commercial land, residential land, and public service land. Then, the location characteristics formed by the street network structures and dominant layout functions could be used fully to increase space utilization efficiency. On the other hand, pedestrians on streets can be enlisted into the uniform analysis framework. How they choose and use spaces according to street network structures and land use distribution characteristics can be investigated via a multi-subject model and other methods to provide a more human-oriented perspective to improve space utilization efficiency and vitality. 

Key words: space syntax, street network structure, centrality, land use, spatial scale