结构化地物的异质性分析及其计算方法  

Heterogeneity Analysis and Its Calculation Method for Structured Geographic Objects

作　　者：刘丰[1,2] 郭建文[1] 

机构地区：[1]中国科学院寒区旱区环境与工程研究所,甘肃兰州730000 [2]中国科学院研究生院,北京100049

出　　处：《遥感技术与应用》2014年第1期122-129,共8页Remote Sensing Technology and Application

基　　金：国家973计划项目(2012CB026106)

摘　　要：将具有明显空间结构的非连续性地物数据的结构分析研究,归纳为空间数据的结构化异质性分析问题。在此概念的基础上引用曲线论的方法分析了结构化异质性的变异函数的构成,得出变异距离受样本点间地物弧长及其空间复杂度共同影响的结论,由此推出变异距离函数的解析表达公式。之后给出了结构化异质性的实验变异函数的一种近似计算方法,该方法在样本合理分布的前提下一致收敛于真实变异值。最后介绍了结构化异质性模型在青藏铁路路基沉降数据的结构分析计算中的一个应用实例。The structural analysis problem of the non-continuity geographic objects with significant spatial structures hasn＇t been adequately addressed. Because of the non-linearity characteristic of structured geo- graphic objects,the classic spatial interpolation methods such as Ordinary Kriging, which is based on Eu- clidean distance, can not complete this mission. To solve this issue,a concept of structured heterogeneity of the spatial data was put forward. After introducing the curve theory of differential geometry,the semivario- gram model of the structured heterogeneity was analyzed, which indicates that the lag depends on the arc length between sample points and the complexity of the research region. In the next section, the analytic formulas of the lag was deduced. The above steps were followed by an approximation calculation method for experimental semivariogram of structured heterogeneity, which is able to uniformly converge to true semivariogram in the condition of the reasonable distribution of sample points. Finally,an application using the embankment settlement data of Qinghai-Tibetan Railway was provided.

关 键 词：变异函数 结构化异质性 结构分析 曲率 三次样条插值 

分 类 号：P208[天文地球—地图制图学与地理信息工程]