摘要
在城市地价分布信息提取中,因地价信息空间位置的相关性和样本数量的有限性,传统统计学的两个基本假设(样本独立与大样本)不易实现。为此,采用探索性空间数据分析(ESDA)技术,对地价数据作空间统计分析。在ESDA分析中,数据结构必须满足正态分布,同时,要对全局趋势分析和方向分析等予以考虑,才能得到误差较小的分析结果。镇江市实例研究表明,该地区地价样点的空间数据结构呈对数分布,存在二阶趋势和北北东方向,合理步长为375 m。
The measurement of GIS Exploratory Spatial Data Analysis in the land price distribution information extraction was study in this paper. As it is very difficult to satisfy the two premises in the traditional statistics for land price samples, other methods should be sought for. The Exploratory Spatial Data Analysis （ESDA） has been set up on the mathematic basis, which has some spatial methods for land price distribution. Besides, a few samples should be needed and the spatial relationship has been considered. It can therefore be used to study the land price distribution. The actuality research of ESDA was expounded in a case of urban commercial land price by GIS Arc / Info 8.0 in this paper. First, its mathematic basis was described. Then the measurements of GIS Exploratory spatial Data Analysis were studied, which included the data structure research, full trend analysis, searching direction and so on. In the end, the GIS Exploratory Spatial Data Analysis measurement was investigated with the Kriging Analysis as an example. The error analysis indicates that, if the frequency distribution of the samples assumes Normal Q - Q Plots and the measurements are used properly, the result is acceptable.
出处
《国土资源遥感》
CSCD
2006年第3期47-50,60,共5页
Remote Sensing for Land & Resources
基金
国家杰出青年科学基金(70425002)
江苏省高校研究生科技创新计划项目"农地非农化空间格局及其形成机制研究"资助
