地理科学 ›› 2010, Vol. 30 ›› Issue (1): 22-29.doi: 10.13249/j.cnki.sgs.2010.01.22

• 论文 • 上一篇    下一篇

基于ESDA的城市社会空间研究——以上海市中心城区为例

宣国富1, 徐建刚2, 赵静3   

  1. 1. 东南大学旅游学系, 江苏 南京 210096;
    2. 南京大学城市与区域规划系, 江苏 南京 210093;
    3. 南京晓庄学院地理系, 江苏 南京 211171
  • 收稿日期:2009-07-26 修回日期:2009-12-10 出版日期:2010-01-20 发布日期:2010-01-20
  • 作者简介:宣国富(1977- ),男,安徽铜陵人,博士,讲师。主要从事城市社会地理、旅游地理研究。E-mail:xgfu99@sina.com.
  • 基金资助:
    国家自然科学基金(40901078、40371038)和江苏省教育厅高校哲学社会科学研究指导项目(08SJD7900048)、 江苏省哲学社会科学基金(08EYD022) 资助。

An Analysis of Urban Social Space Based on ESDA ——A Case Study of the Central Urban District in Shanghai

XUAN Guo-fu1, XU Jian-gang2, ZHAO Jing3   

  1. 1. Department of Tourism, Southeast University, Nanjing, Jiangsu 210096;
    2. Department of Urban and Regional Planning, Nanjing University, Nanjing, Jiangsu 210093;
    3. Geography School of Nanjing Xiaozhuang College, Nanjing, Jiangsu 211171
  • Received:2009-07-26 Revised:2009-12-10 Online:2010-01-20 Published:2010-01-20

摘要: 以上海市中心城区为实证,在因子分析的基础上将ESDA方法应用于城市社会空间研究。运用全局Moran’s I指数、Moran散点图、LISA等指标和方法,从全局和局部两个层面研究了城市社会空间主因子的空间关联特征。结果表明,各主因子都存在显著的空间正相关,呈现趋同集聚,其中社会经济地位因子和居住条件因子的相关性明显强于其他因子,相近社会经济地位和居住条件的社会群体在空间上的集聚对形成城市社会空间的作用更为显著;各主因子都存在不同于全局的局部空间关联模式,存在显著的"热点"和"冷点"地区,其中社会经济地位因子和居住条件因子呈现出更为明显的"热点"和"冷点",具有显著的"同质集聚、异质隔离"特征。相关研究结论对于理解城市社会空间的形成具有重要意义,也可为进一步深入研究的典型案例选择及相关政策制定提供参考。

Abstract: Spatial association is the essential characteristics of spatial related things and phenomena. Exploratory Spatial Data Analysis (ESDA) provides an effective method to reveal the spatial association. The formation of urban social space and its characteristics make the phenomenon of urban social spatial pattern also show significant spatial association. Based on factor analysis, the ESDA methods were applied to urban social space research with the Central Urban District in Shanghai as a case study. From the global and local dimensions, spatial association characteristic of the main factors of urban social space were revealed, by using the indicators and methods of Global Moran’s I index, Moran scatter plot and LISA(Local Indicators of Spatial Association). Global spatial autocorrelation analysis showed that the main factors of urban social space were all with significant spatial association, but there were differences in the degree of spatial association. The factors of socio-economic status and living conditions had stronger spatial association than other factors. Spatial agglomerations of similar socio-economic status and living conditions groups had more prominent contributions to the formation of urban social space. Local spatial autocorrelation analysis demonstrated that the main factors had different local spatial association from the overall pattern, there are obvious "hot spots" and "cold spots", and also some spatial "outliers". The socio-economic status and the residential condition factors show more obvious "hot spots" and "cold spots ", which manifested the characteristics of significant "homogeneous agglomeration, heterogeneous segregation ".

中图分类号: 

  • K928.5