基于TM影像的景观空间自相关分析——以北京昌平区为例

格局与尺度之间的关系是景观生态学的核心研究内容。景观格局发生在不同的尺度 ,而尺度又影响格局的研究 ,因而 ,在景观生态学研究中应用多种量化研究方法于一系列尺度来确定和特征化空间格局研究 ,并探求空间格局与生态学功能和生态学过程之间的关系是非常必要的。以北京昌平区为例 ,从 TM影像中选取 5个具有突出自然和社会经济背景差异的景观 ,即林地景观、农田景观、都市边缘景观、卫星城景观和灌丛景观为研究对象 ,基于归一化植被指数 (N DVI) ,采用常用空间自相关指数 ,即 Moran的 I系数和 Geary的 c系数进行一系列的空间自相关分析 ,旨在阐明 :变化的空间粒度如何影响空间分析 ?以及空间分析如何响应划区效应 ?此外 ,基于 N DVI和数字高程模型 (DEM)也探讨了对于不同的数据类型 ,格局的尺度依赖性如何变化。研究结果表明 :空间粒度的变化对于景观分析有着显著的影响 ,随着空间粒度的增加 ,空间自相关均呈下降趋势 ;不同景观类型对于空间粒度的变化有着不同的响应 ,人为干扰较多的景观具有较低的空间自相关 ,但对空间粒度的变化表现出较强的敏感性 ;对于不同的数据类型 ,格局分析对空间粒度变化的响应是不同的

Landscape spatial autocorrelation analysis of TM remote sensing data:A case study of Changping District, Beijing, China

A strong motivation for developing landscape ecology is to deal with the relationship between spatial pattern and scales. The patterns of landscape development in time and space result from complex interactions of physical biological and social forces. Numerous studies have shown that the spatial pattern of landscape may have significant influences on ecological processes, such as population dynamics, biogeochemical cycling, and biodiversity. Thus, identifying and characterizing spatial pattern across a range of scales using various quantitative methods in order to appropriately understand the interaction of spatial pattern and ecological process are often necessary in landscape ecological studies. This study, therefore, conduct a series of spatial autocorrelation analyses mainly based on NDVI (Normalized Difference Vegetation Index) for five landscapes with contrasting natural and socioeconomic settings in Changping District, Beijing, to demonstrate: how does changing grain size affect the results of spatial analysis? How do the results of spatial analysis differ in changing zoning alternatives? In addition, we also investigate how do such scale-dependent changes vary with different types of landscape data, based on NDVI and DEM (Digital Elevation Model). Results show that changing grain size have significant effects on the values of landscape analysis,and spatial autocorrelation decline with increasing grain. Different landscapes have different sensitivity response to grain size. Landscapes with stronger human disturbance have lower spatial autocorrelation, and more sensitivity to changing grain size. Landscapes with more disturbances by human, almost have no zoning effect. The effect of changing scale varies in their magnitude and rate of change when different types of landscape data are used.