节点文献

基于景观指数的南雄市景观结构特征及其聚类分析

Characterising landscape Pattern and Cluster Zoing in Nanxiong City Based on Landscape Index

【作者】 赵宇

【导师】 钱乐祥;

【作者基本信息】 广州大学 , 土地资源管理, 2010, 硕士

【摘要】 景观生态分区是景观生态学的重要组成部分。通过分区系统的建立,不仅可以全面反映一定区域景观的空间分异状况和组织关系,还能够揭示其空间结构与生态功能特征,并以此作为景观生态评价和规划管理的基础,为城市景观生态规划奠定基础。传统的分区方法,通常采用尺度分析、生态功能指数分析、遥感影像解译等方法,所采用的方法有限,受人为主观影响较大。景观指数聚类分区是一个尚未成熟分区方法。本文在景观格局分析的基础上,通过对研究区景观生态子区基于景观指数的聚类分区,探索并完善这一新方法,同时,挖掘景观指数的信息含量。论文以2009年广东省韶关市南雄市IKONOS遥感影像为数据源进行实证研究,综合运用遥感、GIS、景观生态学、聚类分析方法,研究内容主要包括:(1)计算并分析南雄市景观在类型水平上的空间结构特征及形成机制,南雄市及各景观生态子区景观水平上的空间结构特征及形成机制;(2)以景观指数为变量因子,运用因子分析与聚类分析方法,对聚类因子进行数据压缩和聚类,并对聚类分区结果进行定量分析与探讨;(3)归纳并分析各分区的分布规律和景观生态结构特征,并提出对策建议。研究结果包括两方面:(1)在类型水平上,呈现出耕地景观、林地景观凝聚度高,连通性好,破碎化小的特征,而城乡建设景观、疏林地景观、水体景观、未利用地景观则相反;在景观水平上,南雄市景观结构特征呈现出了显著的阶梯性差异;(2)应用各生态子区景观指数作为变量因子进行聚类分析,把南雄市18个景观生态子区分为了四个大区,分别为:人类干扰主导区、山区盆地过渡区、山区自然景观主导区、低山丘陵自然景观主导区。聚类结果较好的反映出了南雄市景观生态结构的显著特征。景观指数聚类分区的以其突破景观生态分区方法的局限性和人为主观干扰性为特点,初步显示了在景观生态分区中的应用前景。

【Abstract】 Landscape ecological zone classification is an important part of landscape ecology. Through establishment of classification systems, not only the spatial variation conditions and organizational relationships of the regional landscape could be reflected, but also reveal the feature of spatial structure and ecological function. Finally, use the result landscape ecological zone classification as a basic of landscape ecological assessment and management for city landscape ecology planning. Traditional classification such as scale analysis, ecological function analysis, remote sensing image interpretation are often used, it’s limited by method and influenced by human subjectivity.The Cluster Analysis which based on landscape index is an un-mature method. In this paper, this method is exploring and improving through the units cluster analysis witch based on landscape pattern analysis, mining the unknown information of landscape index at the same time.IKONOS images of Nanxiong city, Guangdong province, are used in this paper. Integration remote sensing analysis, GIS analysis, landscape ecology analysis and cluster analysis, it is included that: firstly, calculated and analyzed the spatial structure feature and formation mechanism at class level and landscape level of Nanxiong, landscape ecological units as well; secondly, Factor Analysis and Cluster Analysis are applied for data compression and cluster witch used landscape index as the variable factors, finally, quantitative analysis and discussion the results; finally, summarizes and analyzes the distribution rule of each region and the structure of landscape ecology pattern, then give some suggestions for the city planning of landscape ecology .The research results showed that: firstly, at class level, cultivated landscape and forestland landscape with high cohesion, well connectivity, low fragmentation, Urban and Rural Construction landscape, water landscape, unused land on the contrary. The structure has shown a significant difference at the landscape level. Secondly, 18 landscape ecological units have been clustered to four regions by cluster analysis witch based on application of the landscape index as variable factor, there are Human disturbance dominated areas, Mountain-basin transition areas, Mountainous landscape dominated areas and Mountain landscape dominated areas. Cluster results reveal the landscape pattern feature well.Cluster analysis by use landscape index witch broken the limitations of landscape ecology method and human interference, has shown applications and perspective in landscape ecology classification.

  • 【网络出版投稿人】 广州大学
  • 【网络出版年期】2011年 05期
节点文献中: 

本文链接的文献网络图示:

本文的引文网络