基于Globcover和DMSP/OLS数据的中国大陆城市潜能分析

城市潜能作为城市体系空间相互作用的代表性指标,反映了城市与其所在的城市体系内所有城市间相互作用的强度,即城市所具备的的集聚能力。合理分析城市空间布局及结构、探寻城市发展的空间差异,以及制定区域发展政策日益迫切,研究城市空间相互作用具有重要意义。本文以Globcover数据集提取中国大陆范围内26 619个城市斑块,获取了城市潜能计算时所需的城市斑块规模总量;同时,以DMSP/OLS夜间灯光影像为数据,提取能综合反映人类社会经济活动的城市斑块夜间灯光总强度,参与城市潜能的计算。利用潜能模型,依据城市斑块分层结果,以城市斑块规模和时间成本为参数,计算得到公里格网的城市潜能值模拟分布图,并分析了中国大陆城市潜能的空间分异特征。

Analysis of Urban Potential in Chinese Mainland Based on Globcover and DMSP/OLS Nighttime Light Data Sets

With the acceleration of urbanization in China, it has become increasingly urgent to analyze city space structure, explore spatial difference of urban development, and make regional development policy. Therefore, the research on urban spatial interaction has great instructive significance. Urban potential is a representative index of urban spatial interaction. It reflects the spatial interactions between one city and the other cities from the same urban system. In fact, urban potential indicates the city’s capability of agglomeration. Previous studies on urban potential mostly focused on using point data to represent the spatial position of a city, and then simulate its urban potential. However, simulations based on this method would deviate from the real situations to some extent. The reason is that, on one hand, the geographic coordinates of the administrative center do not coincide with the urban center; on the other hand, most cities are composed of several separated urban patches, and these patches may also interact with each other. To solve the problems caused by using point data in representing city’s spatial position in the urban potential researches, we took 26,619 urban patches extracted from Globcover data sets as the fundamental research objects. In this way, not only the spatial interactions between different cities, but also the spatial interactions between different urban patches within the same city could be taken into consideration. We adopted DMSP/OLS stable nighttime light data as the data source to attain the size of urban patches, which is taken as one variable in the potential model. The DMSP/OLS stable nighttime light data has advantages including the rapid update frequency and the wide coverage. The total night light index derived from the image reflects the social economic activities of human comprehensively. A raster map, in which its cell value represents urban potential, was produced by using a potential model that is one of the typical spatial interactive models after assigning values to its two variables. Based on the map, we analyzed the spatial differentiation of urban potential. As a conclusion, this paper provided a method for using new data sources to simulate urban potential.