基于最佳地形因子组合的地貌形态类型划分研究
发布时间：2019-02-24 11:27
【摘要】：地貌是地球表层空间长期演化的成果,具有时间上的长期连续性和空间上的广阔分异性,传统地貌数据的采集基本采用野外实地人工调绘,工作量大、效率低下,难以获取某一地貌的完整数据。随着计算科学的迅速发展,地貌的定量化分析已成为当前地貌学研究的主要方式。本文以研究区30m分辨率的DEM数字高程数据为基础,提取多种地形因子信息,选取地形因子最佳分类组合,借鉴遥感影像分类方法,将选取的地形因子作为单波段合成一幅RGB多波段影像,以监督和非监督分类相结合的方法,对研究区的地貌形态类型进行划分。论文的主要内容和结论如下:(1)以研究区30m分辨率的DEM数据为基础,提取高程、坡度、坡度变率、坡向、坡向变率、地形起伏度、平面曲率、剖面曲率、地表粗糙度、地表切割深度、高程变异系数、等高线密度等12种地形因子,以地形起伏度为例,采用ArcGIS邻域分析法分别提取了窗口大小从3像元×3像元到35像元×35像元的地形起伏度数据,并利用均点变值法对得到的数据进行分析处理,得到了宏观地形因子的最佳分析窗口大小为14像元×14像元,最佳统计面积为0.1764km2。(2)从定义和地学意义方面对各因子做了定性的分析,以相关系数法和雪氏熵值法对这些地形因子做了定量的分析,确定了适合研究区地貌形态类型划分的最佳地形因子组合为高程、全累计曲率、坡度变率、DEM晕渲图、高程变异系数、坡度和地形起伏度等7个指标。(3)以《四川省地貌区划》为依据结合研究区实际地形情况提出了一种适合研究区的地貌形态类型分类体系。将研究区地貌形态类型分为河谷地貌、丘陵、山地三大类,河谷地貌、缓丘平坝、浅丘、深丘、低山、低中山6个小类。(4)提出了非监督分类和监督分类相结合的地貌形态类型划分方法,对研究区进行了地貌形态类型划分工作,得到了总体分类精度为74.15%,与地表实际参考数据吻合度为68.37%的分类结果,统计了研究区地貌形态类型分布情况数据,制作了研究区地貌形态类型专题图。
[Abstract]:Geomorphology is the result of long-term evolution of the surface space of the earth. It has long-term continuity in time and wide differentiation in space. The traditional geomorphological data collection basically adopts field manual adjustment and mapping, which has the advantages of heavy workload and low efficiency. It is difficult to obtain complete data on a particular landscape. With the rapid development of computer science, the quantitative analysis of geomorphology has become the main way of geomorphology research. Based on the DEM digital elevation data with 30m resolution in the study area, this paper extracts the information of various terrain factors, selects the best classification combination of terrain factors, and draws lessons from the classification method of remote sensing images. The selected terrain factors are used as a single band RGB multi-band image, and the geomorphological types of the study area are divided by combining the supervised and unsupervised classification methods. The main contents and conclusions are as follows: (1) based on the DEM data of 30m resolution in the study area, elevation, slope variability, slope direction variability, topographic fluctuation, plane curvature, profile curvature, surface roughness are extracted. There are 12 topographic factors, such as depth of surface cutting, coefficient of variation of elevation, density of contour line, etc. Taking terrain fluctuation as an example, the topographic relief data of window size from 3 pixels 脳 3 pixels to 35 pixels 脳 35 pixels are extracted by ArcGIS neighborhood analysis method. The data obtained are analyzed and processed by the mean point variation method. The optimum size of the analysis window for the macro terrain factors is 14 pixels 脳 14 pixels. The optimum statistical area is 0.1764km2. (2) qualitative analysis of each factor is made in terms of definition and geoscientific meaning, and quantitative analysis of these topographic factors is made by correlation coefficient method and snow entropy method. The optimum combination of topographic factors suitable for the classification of geomorphologic types in the study area is determined as elevation, total cumulative curvature, slope variation rate, DEM shading map, elevation variation coefficient. (3) based on the geomorphological regionalization of Sichuan Province, a classification system of geomorphologic types suitable for the study area is put forward according to the actual topographic conditions of the study area. The geomorphologic types of the study area are divided into three types: valley geomorphology, hilly area, mountain area, valley geomorphology, gentle hill level dam, shallow hill, deep hill, low mountain, and so on. (4) the method of combining unsupervised classification with supervised classification is put forward, and the geomorphological classification of the study area is carried out, and the overall classification accuracy is 74.15. According to the classification results of 68.37% of the actual surface reference data, the distribution data of geomorphologic types in the study area are statistically analyzed, and the topographic map of geomorphologic types in the study area is made.
【学位授予单位】：重庆交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P931;P208
本文编号：2429507
[Abstract]:Geomorphology is the result of long-term evolution of the surface space of the earth. It has long-term continuity in time and wide differentiation in space. The traditional geomorphological data collection basically adopts field manual adjustment and mapping, which has the advantages of heavy workload and low efficiency. It is difficult to obtain complete data on a particular landscape. With the rapid development of computer science, the quantitative analysis of geomorphology has become the main way of geomorphology research. Based on the DEM digital elevation data with 30m resolution in the study area, this paper extracts the information of various terrain factors, selects the best classification combination of terrain factors, and draws lessons from the classification method of remote sensing images. The selected terrain factors are used as a single band RGB multi-band image, and the geomorphological types of the study area are divided by combining the supervised and unsupervised classification methods. The main contents and conclusions are as follows: (1) based on the DEM data of 30m resolution in the study area, elevation, slope variability, slope direction variability, topographic fluctuation, plane curvature, profile curvature, surface roughness are extracted. There are 12 topographic factors, such as depth of surface cutting, coefficient of variation of elevation, density of contour line, etc. Taking terrain fluctuation as an example, the topographic relief data of window size from 3 pixels 脳 3 pixels to 35 pixels 脳 35 pixels are extracted by ArcGIS neighborhood analysis method. The data obtained are analyzed and processed by the mean point variation method. The optimum size of the analysis window for the macro terrain factors is 14 pixels 脳 14 pixels. The optimum statistical area is 0.1764km2. (2) qualitative analysis of each factor is made in terms of definition and geoscientific meaning, and quantitative analysis of these topographic factors is made by correlation coefficient method and snow entropy method. The optimum combination of topographic factors suitable for the classification of geomorphologic types in the study area is determined as elevation, total cumulative curvature, slope variation rate, DEM shading map, elevation variation coefficient. (3) based on the geomorphological regionalization of Sichuan Province, a classification system of geomorphologic types suitable for the study area is put forward according to the actual topographic conditions of the study area. The geomorphologic types of the study area are divided into three types: valley geomorphology, hilly area, mountain area, valley geomorphology, gentle hill level dam, shallow hill, deep hill, low mountain, and so on. (4) the method of combining unsupervised classification with supervised classification is put forward, and the geomorphological classification of the study area is carried out, and the overall classification accuracy is 74.15. According to the classification results of 68.37% of the actual surface reference data, the distribution data of geomorphologic types in the study area are statistically analyzed, and the topographic map of geomorphologic types in the study area is made.
【学位授予单位】：重庆交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P931;P208
【参考文献】
相关期刊论文 前3条
1 龙恩;程维明;周成虎;姚永慧;刘海江;;基于Srtm-DEM与遥感的长白山基本地貌类型提取方法[J];山地学报;2007年05期
2 李本纲,陶澍;用数字高程模型进行地表径流模拟中的几个问题[J];水土保持通报;2000年03期
3 隋刚;郝兵元;彭林;;利用高程标准差表达地形起伏程度的数据分析[J];太原理工大学学报;2010年04期
相关硕士学位论文 前1条
1 刘爱利;基于1：100万DEM的我国地形地貌特征研究[D];西北大学;2004年
，本文编号：2429507
本文链接：https://www.wllwen.com/kejilunwen/dizhicehuilunwen/2429507.html
