大尺度景观结构指数的因子分析

基于遥感和GIS技术,尝试对大尺度下陆地景观结构指数进行因子分析。根据生态区划计算描述中国陆地景观结构的13种景观结构指数,包括斑块密度、最大斑块指数、边界密度、面积加权斑块形状指数、面积加权斑块分维度、平均最近斑块距离、最近斑块距离的标准差、最近斑块距离的相异系数、蔓延度、分布邻接指数、结合度、香农多样性指数、辛普森多样性指数。在相关分析的基础上,对其中12种指数进行因子分析,并采用前3个公因子分析中国陆地景观结构,3个公因子的方差累积贡献率达88.8%,分别定义为斑块多样性因子、斑块形状复杂性因子和斑块分布均匀性因子。计算表明,在人类活动影响强烈的生态区,这3个公因子取值都比较小,反映斑块类型多样性减少、形状复杂性降低、分布趋于均匀。

A Factor Analysis on Landscape Structure Metrics with Large Scale

While landscape ecological metrics is widely used to describe micro-or meso-scale regions now,there is little literature on using them to macro-scale regions.This paper illustrates an example of using thirteen common landscape structure metrics on terrestrial China.Firstly,thirteen metrics were calculated according to ecological regions of China by means of remote sensing and GIS.The metrics include patch density,largest patch index,edge density,area-weighted shape index distribution,area-weighted fractal index distribution,mean Euclidean nearest neighbor distance distribution,standard deviation of Euclidean nearest neighbor distance distribution,coefficients of variation of Euclidean nearest neighbor distance distribution,contagion,interspersion and juxtaposition index,cohesion,Shannon's diversity index and Simpson's diversity index.Secondly,methods of correlation analysis and factor analysis were adopted to analyze these metrics.Related to patch diversity,complexity of patch shapes,evenness of distribution respectively,the first three factors from the factor analysis were adopted to describe the landscape structure of terrestrial China.At last it comes to the conclusion that the ecological regions more affected by human beings show less value for the three factors,meaning less patch types,simpler patch shapes and patches being more evenly distributed.