Abstract：Exploring the spatial layout of service industries in new urban districts is significant for guiding the planning of these districts and optimizing the spatial layout of service industries. This study examined the overall and sub-industry spatial distribution of service industries in Zhengdong New District, Zhengzhou, Henan province, China. An analysis was carried out using points of interest (POI) data (Big Data closely related to human-economic geography) and the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm in MATLAB and Arc GIS. The main findings from this research were: (1) In terms of the overall spatial layout, the clusters showed an obvious "414" spatial hierarchy and the regions with these clusters had better accessibility. Most clusters were located in only one functional area, but some others were located in two or more functional areas. The distribution of noise points was relatively scattered and isolated from the potential regions of service industry agglomeration in future. The effect of spatial proximity, administrative drivers and market orientation led to the apparent spatial polarization of service industries, which were centered in the lower left corner of a diagonal from the northwest to the southeast of Zhengdong New District. Future urban planning should target the construction of 'multiclustering centers' to avoid excessive clustering of service industries (especially low-level service industries) in the CBD and residential/commercial logistics areas, and to avoid increasing the pressure on population, transportation, resources and the environment in these two functional areas. (2) The analysis by sub-industry showed that some industries had differences in the distribution of spatial nodes, although the CBD and residential/commercial logistics areas were the gathering points of spatial nodes for all industries. Some industries showed features of coincidence and dislocation between the spatial nodes and the location of functional areas. From the planning point of view, the positioning of each functional area should be clearly differentiated, the overlap of service industries should be avoided, and the location of each sub-industry should be considered carefully. In the end, the paper proposed the optimization of the industrial structure of functional areas, which provides the basis for improving the spatial structure of service industries in Zhengdong New District.
Keyword：POI data; DBSCAN algorithm; service industry; Zhengdong New District; functional area;
【source】Geographical Research 2018年01期
