区域产业布局模式识别:指标体系与实证检验

关于区域产业布局模式的已有研究多是针对特定地区或特定产业的经验归纳,缺乏连续性和可比性。在综合性、典型性、独立性、可比性和可操作性原则的指导下,从中心性、分布集中性、空间集聚性和耦合关联性4个维度构建了表征产业布局模式的综合指标体系。以南充市2位数产业的模式分类为例,综合使用相关分析、聚类分析、因子分析、图谱分析等多元统计方法,结合定性分析,从4个维度分别归纳了其产业布局模式,进一步总结了单中心极度集中型、“中心-外围”均衡簇群型、点状散布型和人口依赖的遍在均衡型等4种产业布局的综合模式。指标体系的合理性和灵活性得到了实证检验。

Industrial Layout Pattern Recognition:Index System and Empirical Test

Regional industrial layout patterns have drawn the attention of many geographers. Focusing on specific industry and (or)specific area, however, most of previous researches are lacking of continuity and comparability. A rational, comparable, and feasible index system is essential for describing and evaluating layout patterns of various industries in different areas. This paper tries to establish such an index system and estimate its rationality and feasibility through the followed case study. The index system is constructed to represent the industrial layout patterns from perspectives of centricity, inequality, spatial agglomeration, and correlation between observed industry and other industries and economic factors. Firstly, the proportion of employment in the center to that in the region and to that in the sub-center, are recommended to represent the centricity of the industrial layout. Secondly, the authors set up indices of ubiquity, measured by the proportion of subregions with one or more enterprises of the observed industry, and concentration ratio to represent the inequality of industrial distribution among subregions. The concentration ratio can also be estimated excluding the center. Thirdly, spatial agglomeration of an industry is measured by distances between the center and sub-centers and other industrial agglomeration areas, and distances among industrial agglomeration areas without regard to the center. Finally, it is also important for recognition of industrial layout patterns to investigate their relationship with the spatial distribution of other industries, population, land use and other socio-economic factors, which can be measured by Pearson correlation coefficient, similarity coefficient or grey relational degree. Taking Nanchong as a case, we classify 2-digital industries from the aforesaid perspectives respectively with quantitative analysis methods. At last, four typical comprehensive patterns of industrial layout are concluded. The case study provides evidence for rationality and flexibility of the index system, which can also be used in industrial layout pattern recognition and classification in all kinds of areas, at different times and on multi-scales.