The relationship between street network and land use

Since the opening-up policy in 1980s, there has been a fast expansion of China’s transport facilities, and the construction of a transport network covering the whole country. With the high-speed development and huge amount of financial investments, it is important to better understand urban development and make policy based on these objective regularities, so to improve the resources use efficiency and achieve the goal of sustainable development. Based on this background, this research aims to investigate the relationship between land use and street network, so to find whether street properties can capture the land development patterns of different land-use types. This paper investigates this relationship in Ningbo, China. Street property is measured by centrality indicator: closeness, betweeness and straightness on the street network. Six land use types are collected in the study area: commercial land, residential land, industrial land, public services, grass land and forest. Kernel density estimation is used to convert datasets into a basic raster unit. The relationship between each land use and each street network centrality is analyzed by Pearson’s correlation. In addition, geographically weighted regression was used to show how this relationship varied across space. The results indicate that commercial land, public services land and residential land are highly correlated with betweeness and straightness centralities. For the industrial, forest and grass land, there is almost no correlation between three centrality indices and these land uses. Furthermore, the relationships vary not only among different land-use types but also in a single land-use type. Spatial heterogeneity also exists in these relationships. This finding confirms that street centrality can capture the land development of different land uses and plays a important role in shaping urban fabrics.