Abstract:

Accurately and effectively estimating the effects of urban pattern on industrial location is always an important issue that has received much attention in urban geography. However, current studies have mainly focused on the influence of a single type of urban functional spaces on industrial location using statistical data based on administrative units. And these studies have not conducted in-depth research into the spatial heterogeneity of influences. Against the existing shortcomings, taking the central Chongqing as an example, this research aimed to investigate how different urban functional spaces influenced the catering industry distribution based on kernal density, spatial autocorrelation and geographic weighted regression from point of interest (POI) data. The research reveals the following points: (1) the spatial distribution of restaurants was characterized by a multi-center spatial structure of "one main, two subs and four third-level centers", which directly reflected the urban pattern. (2) Not only the scale of restaurant agglomeration was closely related to urban expansion sequence, but also the direction of restaurant agglomeration was consistent with urban expansion direction. (3) The urban functional spaces had different or even opposite influences on the restaurant distribution in different city groups. The effect of residential space on the restaurant distribution was all positive, which increased from the central to the peripheral groups. The influence of commercial space on restaurant distribution was weakened from the inside to the outside of the barrier of the mountains, but there might be a phenomenon of commercial dependence in the groups where the location was isolated and the business development was immature. Since the restaurants in the peripheral groups was more dependent on the transport accessibility, the effect of urban traffic space on restaurant distribution in the peripheral groups was greater than that in the central groups, which resulted in a phenomenon of traffic dependence. The urban public space in central groups played a greater role in promoting restaurant assembled than that of the peripheral groups on account of high-quality public service in central groups. The influence of urban leisure space on restaurant distribution was related to the number and popularity of scenic spots. (4) Moreover, different urban functional spaces had different influences on the spatial distribution of restaurants, and urban commercial space had the greatest impact due to its high density of urban construction and population density. This study is especially valuable for understanding the function mechanism of urban pattern on industrial location and providing a scientific basis for making rational urban development plan.

Key words: urban space, mountainous city, catering industry location, spatial heterogeneity, point of interest (POI)