地理科学进展 ›› 2010, Vol. 29 ›› Issue (6): 757-768.doi: 10.11820/dlkxjz.2010.06.016

• 方法模型与应用 • 上一篇    

空间统计学进展及其在经济地理研究中的应用

杨振山,蔡建明   

  1. 中国科学院地理科学与资源研究所,北京100101
  • 收稿日期:2009-12-01 修回日期:2010-02-01 出版日期:2010-06-25 发布日期:2010-06-25
  • 通讯作者: 蔡建明(1961-),男,博士,研究员,博士生导师,主要研究领域城市可持续发展、城市发展战略、世界城市和都市农业 等.E-mail: caijm@igsnrr.ac.cn
  • 作者简介:杨振山(1979-),男,博士,助理研究员,主要研究方向为城市规划,GIS 和RS 空间信息技术在城市规划研究中的应 用,空间经济,产业集群和都市农业.E-mail: yangzspat@gmail.com
  • 基金资助:

    中国科学院第三期创新项目(KZCX2-YW-321).

Progress of Spatial Statistics and Its Application in Economic Geography

YANG Zhenshan,CAI Jianming   

  1. Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China
  • Received:2009-12-01 Revised:2010-02-01 Online:2010-06-25 Published:2010-06-25

摘要:

空间统计是20 世纪90 年代以后在经济地理,尤其是城市和区域研究领域中发展起来的重要研究方法。这 一方法考虑到事物发展的空间依赖性,大大革新了原有经典统计,并借助于地理信息技术增强了可视化效果,丰富 了在城市和区域研究中对空间的认识和预期。本文简明地综述了空间统计主要领域和内容,指出探索式空间分析, 局部空间统计和空间回归模型是空间统计与经济地理研究主要的结合点,代表了未来发展趋势;并就空间尺度、空 间权重矩阵、边缘效应和解释谬误等方面着重阐述了应用时应注意的问题。在此基础上,对近年来空间统计在社会 经济要素集聚、土地利用和城市空间结构、交通和房方产研究中的应用进行了回顾,总结其主要应用方面和价值。 空间统计将会大大提高对城市微观尺度的研究,为城市研究的基本理论假设和社会经济发展机理提供重要研究手 段,但对数据库建设提出了更高的要求。

关键词: 经济地理, 空间数据, 空间统计, 空间相关

Abstract:

Since the 1990s, spatial statistics has been emerged as an innovative and important method in human geography, especially in urban and regional studies. This method is valued at understanding socioeconomic elements based on their intrinsic spatial dependency. By deciphering the spatial dependency, it modifies conventional statistics, with geographical information systems on data storing, management, visualization and analysis, to improve knowledge of understanding urban and regional space. Allowing for the requirement of urban and regional studies, the paper introduces the main concepts, principles, techniques and problems when spatial statistics is applied. The paper is particularly interested in three main application trends: exploratory spatial data analysis, local statistics, and spatial statistic modeling. With the purpose of properly applying the method of spatial statistics, it is argued that applicants should pay attention to spatial scale, spatial weights, edge effects and ecological fallacy. The paper further reviews the application and progress in the themes of economic agglomeration, land use and urban structure, transportation and real estate by addressing associated application values. It is pointed out that the application of spatial statistics will strengthen the analysis at micro-levels but will need high resolution data set.

Key words: economic geography, spatial data, spatial dependency, spatial statistics