摘要
景观指数法作为景观格局分析的重要研究方法和度量手段,也是景观生态学研究的重要组成部分。由于景观指数在空间格局表达时信息高度浓缩,其在应用中普遍地存在"一值多形"问题。文中根据景观水平上聚集度指数和多样性指数对景观空间整体复杂性和异质性的描述,采用二维空间地理现象的四叉树编码表达对四个不同模拟景观格局进行空间划分,形成多层次的四叉树结构,并通过Fragstats软件对不同子区域景观指数计算,以及采用数值矩阵和对应的3D模型来表达不同景观空间格局。该方法明确展现了不同景观格局相对应的空间位置在景观聚集性特征和多样性特征上的不同,并且用同一景观格局中各子区域的景观指数描述自身子区域特征,也从景观空间局部表现了景观空间格局的异质性。研究过程证明了基于四叉树的景观指数分析方法可以区分景观格局在指数计算中的"一值多形"。同时文章展望了该方法对于不同景观空间格局进行聚类划分的应用。
As a measurement for landscape pattern analysis, the method of landscape metrics plays an important role in the study of landscape ecology and it is the simplest method in landscape spatial analysis. The problem, ＂a high value shape＂, which means landscape metrics have same values in different landscape spatial patterns or different shapes have the same fractal dimension, is caused by the highly concentrated information and the thoughtless of the space relation.＂A high value shape＂ is common and it affects the cognition of landscape spatial pattern and its heterogeneity. In this paper, according to the description of complexity and heterogeneity by aggregation metrics and diversity metrics in landscape levels, the quad-tree coding for two-dimensional geographical phenomenon was taken to divide the landscape space into smaller ones, the values of the metrics on each smaller space were calculated and analyzed. The difference of values in these numerical matrixes not only showed the global spatial pattern, but also revealed the local spatial pattern in subdomains. At last, the 3D visualization based on the numerical matrixes gave a good view of the landscape spatial pattern. The study found that the method of quadtree coding could distinguish the spatial pattern well though there was＂a high value shape＂problem in the four landscape spatial patterns on DIVISON, SHDI, and SHEI metrics. The method described the different quantity in the spatial pattern, because the corresponding sub-domains had different values in four landscape spatial patterns. Besides, the 3D visualization of the numerical matrixes helped to understand the difference in these overall landscapes and the spatial heterogeneity in each landscape. At the end of the paper, it has the outlook for the application of the method in the clustering analysis of different landscape spatial patterns.
出处
《地理信息世界》
2016年第2期21-26,共6页
Geomatics World
基金
国家自然科学基金(41222009)
国家自然科学基金项目(41271405)资助
关键词
景观指数
四叉树编码
景观空间格局
空间异质性
landscape metrics
quad-tree coding
landscape spatial pattern
spatial heterogeneity
