基于夜光遥感影像与百度POI数据的中国城市群空间范围识别方法

提出基于夜间灯光遥感影像、电子地图兴趣点和社会经济统计数据等,以经济地理学中的“点-轴”理论为基础,以“点-轴集聚区”的识别为核心,利用迭代自组织聚类、聚合分析、指标阈值筛选等方法,识别中国城市群及其空间范围的技术方法。通过该方法识别出中国14个城市群,其中8个城市群的空间范围与规划范围接近;与规划范围不一致的则表现为三种情况,分别揭示出规划中需要考虑的不同问题。研究结果表明,本文提出的方法能突破行政边界限制,科学反映城市群辐射范围,客观反映城市之间的社会经济联系强度,并基于“现状-动态”视角有利于深入发掘潜在的城市群对象。研究结果可以为城市群规划和管理提供参考。

Identifying the spatial range of urban agglomerations in China based on night light remote sensing and POI data

Urban agglomeration, as an emerging phenomenon in many urbanized areas worldwide, is considered as a highly developed spatial form of integrated cities. Cities are highly linked within an urban agglomeration, which renders the agglomeration one of the most important carriers for global economic development. In recent years, the study of urban agglomeration has become an important agenda both for urban planning and urban sustainable development. However, in the research community, there is still a lack of a consensus with regard to how to delineate the urban agglomerations in geographic space. Particularly, in many urban planning cases, functional links among cities are often neglected, resulting in overestimated spatial ranges of the planned urban agglomerations. The aim of this paper is to develop a method for the identification of the spatial range of urban agglomerations by using night-light remote sensing data, digitally mapped points of interest (POI) and the "point-axis" theory in economic geography. Firstly, based on a review of the "point-axis" theory in economic geography, we developed a concept of "developing axis" with two basic characteristics and used the concept to describe the four development stages of urban agglomerations. Then, we calculated two indexes to quantify the intensity and its changes in socio-economic activities by combining nighttime light remote sensing images and POI data. After that, we conducted a clustering analysis on the two indexes to identify and extract the "point-axis cluster", and overlaid it with the administrative boundaries to obtain a set of candidate urban agglomerations. Finally, we used socio-economic statistic data and formulated criteria based on previous studies to select urban agglomerations. Using this method, a total of 14 urban agglomerations in China are identified. Among which, eight have spatial ranges match their planning documents. As for the mismatching urban agglomerations, three different types of mismatch are distinguished, which indicate that different types of problems need to be considered in the planning. The results show that the proposed method can overcome the restriction of administrative boundaries in the identification of the spatial range of urban agglomerations, objectively reflect the strength of social and economic links among cities, and help to identify potential urban agglomerations with a dynamic perspective. This research can provide useful implications and suggestions for urban agglomeration planning and management.