RS和GIS支持下的山地景观粒度效应研究

发布时间：2018-05-04 11:27

本文选题：山区 + 土地利用　；参考：《昆明理工大学》2013年硕士论文

【摘要】：随着当今社会经济的快速发展,人口的日趋增多,人地矛盾关系变得越来越紧张,人类的活动使土地利用状况和地表景观时时刻刻发生着重大的变化。目前,景观格局变化研究成为景观生态学研究的一个热点方向,应用景观指数来定量分析景观格局成为一种有效的研究手段。同时,对景观格局进行研究时要受到尺度效应的影响,准确掌握尺度效应带来的规律有利于正确认识景观格局的变化。本文选取云南省具有山地代表性的寻甸县为研究区,在遥感软件ERDAS与地理信息系统软件ArcGIS以及景观分析软件Fragstats支持下,运用空间分析技术和景观格局分析方法,对滇中山区土地利用动态变化及景观格局进行定量的研究与深入地分析,选取了18种景观指数,从景观水平出发,探讨了景观指数对粒度变化的响应规律。文章前期运用遥感技术对获取的三期TM影像进行了解译,得到了土地利用现状数据。研究发现,寻甸县三个时期土地利用以林地和耕地为主,两种土地类型面积之和占到了全县总面积的90%以上,水域面积和其他用地面积为最少。在两个时间段内,通过土地转移矩阵可以看出各种土地利用类型基本都有转入和转出。1992-2001年期间,面积变化较大的为耕地和林地,耕地减少了315.78km2,林地增加245.76km2,2001-2006年期间,各类土地利用变化相对上个时期变化较小。将景观格局图与提取的地形信息相叠加分析。在高程范围内,耕地和林地占据各高程面积的80%以上,坡度分布上,各类景观类型主要分布在坡度为0-25°范围内,均占到各自总面积的80%以上,相对来说景观在坡向上分布比较平均。地形分布指数显示,建设用地和水域对高程地形具有较强的选择性,水域、耕地和建设用地趋向于分布在较为平缓的地带,林地在坡度较大的地方显示出较强的分布优势；建设用地和耕地对坡向选择性较强,而林地和水域对坡向选择则不是很强。采用景观指数法,从斑块类型水平和景观水平对研究区景观格局进行了分析,定量地说明了景观格局的变化情况。最后,探讨了空间粒度变化对景观指数产生的效应,研究中选取的18种景观指数的粒度效应分为四类：斑块数、斑块密度、边界总长、边界密度、景观形状指数、聚集度和蔓延度随着粒度的增大,指数呈现出有规律的减小的趋势,具有明确的尺度效应：平均形状指数、平均分维数指数和最大斑块指数随着粒度的增大,指数先是呈现有规律减小的趋势,之后出现较为明显的粒度转折点；平均斑块面积、分离度、分散指数和分割度随着粒度的增大,指数呈现增大的趋势,有明显或者没有转折点：总面积、香浓多样性指数、香浓均匀度指数和丰富度随着粒度的增大,指数呈变化趋势不明显,变化规律呈现出不可预测性。
[Abstract]:With the rapid development of social economy and the increasing of population, the contradiction between man and land becomes more and more tense. At present, the research of landscape pattern change has become a hot research direction of landscape ecology, and quantitative analysis of landscape pattern by landscape index has become an effective research method. At the same time, the study of landscape pattern should be affected by the scale effect, and it is beneficial to understand the change of landscape pattern by accurately mastering the law brought by scale effect. This paper selects Xundian County, which is the representative mountain area of Yunnan Province, as the research area, with the support of remote sensing software ERDAS, GIS software ArcGIS and landscape analysis software Fragstats, using spatial analysis technology and landscape pattern analysis method. The dynamic changes of land use and landscape pattern in mountain areas of central Yunnan were studied and analyzed quantitatively. Eighteen landscape indices were selected, and the response of landscape indices to granularity change was discussed from landscape level. In the early stage of this paper, the remote sensing technology was used to interpret the three periods TM images, and the present land use data were obtained. It is found that the land use in Xundian County is dominated by woodland and cultivated land. The total area of the two types of land accounts for more than 90% of the total area of the county, and the area of water area and other land use is the least. In the two time periods, it can be seen from the land transfer matrix that all kinds of land use types were basically transferred and transferred from 1992 to 2001. During the period of 1992-2001, the area of cultivated land and forest land decreased by 315.78 km2, and the forest land increased by 245.76 km2 during 2001-2006, the area of cultivated land decreased by 315.78 km2, and the forest land increased by 245.76 km2 during the period of 2001-2006. All kinds of land use changes are relatively small in the last period. The landscape pattern map is superposed with the extracted topographic information. In the range of elevation, cultivated land and forest land occupy more than 80% of each elevation area. In the slope distribution, all kinds of landscape types are mainly distributed in the range of 0-25 掳, accounting for more than 80% of their total area. The landscape is relatively evenly distributed up the slope. Topographic distribution index shows that construction land and water area have strong selectivity to elevation terrain. Water area, cultivated land and construction land tend to be distributed in a relatively gentle area, and woodland shows a strong distribution advantage in places with higher slope. Construction land and cultivated land are more selective to slope than woodland and water to slope. Based on the landscape index method, the landscape pattern in the study area was analyzed from patch type level and landscape level, and the change of landscape pattern was explained quantitatively. Finally, the effects of spatial granularity change on landscape index are discussed. The granularity effects of 18 landscape indices selected in the study are divided into four categories: patch number, patch density, total boundary length, boundary density, landscape shape index. With the increase of particle size, the index of aggregation and spread showed a regular decreasing trend, and had a definite scale effect: average shape index, mean fractal dimension index and maximum patch index increased with the increase of particle size. The average patch area, separation degree, dispersion index and partition degree increased with the increase of particle size. There were obvious or no turning points: total area, aroma diversity index, aroma evenness index and richness with the increase of particle size, the change trend of the index was not obvious, and the law of change showed unpredictability.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P208;P237;P901

【参考文献】