基于因子分析的景观指数筛选与分区研究  

Study on Selection of Landscape Indices Based on Factor Analysis and Division of Landscape Areas

作　　者：杨丽[1] 谢高地[2] 甄霖[2] 

机构地区：[1]山东科技大学测绘科学与工程学院,山东青岛266510 [2]中国科学院地理科学与资源研究所,北京100101

出　　处：《山东科技大学学报（自然科学版）》2010年第5期19-23,共5页Journal of Shandong University of Science and Technology(Natural Science)

基　　金：山东科技大学科学研究“春蕾计划”项目(2008AZZ023)

摘　　要：以泾河流域为例,在总结前人研究的基础上,筛选出7个一致性较好的景观指数,应用相关性分析和因子分析对其进行了筛选,并据此对研究区进行了分区。结果表明:面积加权平均分维数与面积变化具有较高的相关性,予以剔除。其余景观指数可以用景观分布与斑块形状因子和景观破碎化因子2个公因子代替,并据此将泾河流域32个县市分成3个类型区,分别处于流域的上游、中下游和下游地区,表现出一定的分布规律。Taking Jinghe watershed as an example,this paper selected 7 landscape indices with good consistencies by using the correlation analysis and factor analysis and also made a division of the study area based on the previous studies. The results showed that AWMPFD had higher correlation with extent change,so it was deleted. The results also showed that landscape indices selected in the paper could be replaced with two common factors,i, e. the factor 1 of landscape distribution and patch shape and the factor 2 of landscape fragmentation,respectively. According to the score of factor 1 and factor 2,32 counties of Jinghe watershed could be divided into 3 types of landscape areas, showing regular patterns along the upstream,midstream and downstream areas of Jinghe watershed.

关 键 词：景观指数 幅度效应 因子分析 聚类分析 泾河流域 

分 类 号：P901[天文地球—自然地理学]