基于DEM的三川河流域特征提取研究

发布时间：2018-10-08 19:59

【摘要】：数字高程模型(DEM)作为重要的空间数据,已经被广泛应用于诸多领域。在流域数字化的背景下,利用DEM自动提取流域特征已成为水文学研究的重点。流域特征是水文模拟及流域管理等研究的基础性数据,能否准确快速的获去,直接关系到这些研究的精度和效率。近年来随着3S技术的不断发展及DEM精度和获取速度的不断提高,流域特征提取研究得到进一步发展,涌现出了许多新理论和新方法。本文系统概括的介绍了数字高程模型的相关知识,回顾了基于DEM提取流域特征的国内外研究进展,介绍了流域特征的基本概念,探讨了基于DEM提取流域特征的算法基础及方法；以三川河流域为例,利用先进的GIS技术,实现了流域特征的自动提取,分析了提取结果的精度；对整体流域进行分割并对子流域进行了分类。研究成果有： (1)提取的地形因子较好的吻合流域实际情况,表明提取算法和应用的工具是成熟有效的。结果表明：流域内地形平均坡度为13.34。,河源地区较河谷地区要大,坡度大致随着地形高程的不断降低而降低；河谷和残源地区坡度变化微小,沟壑区变化较大；坡向变率清楚的表征了流域内山脊线和山谷线的分布情况；坡向在8个方向上的分布较均匀；由曲率特征值可知流域内河谷区地形平坦,非河谷区沟壑纵横凹凸交替。 (2)提取河网的步骤包括：DEM的预处理、格网流向判定、汇流累计量计算、最佳临界阀值确定、栅格河网生成、栅格河网矢量化及河网的修正和平滑。提出通过汇流累计阀值与河网密度相关关系来确定最佳临界阀值的新方法,经反复试验表明阀值为6500时,提取的河网最符合实际情况,同时说明此方法是可行的。 (3)基于提取的栅格河网及格网流向数据,将整体流域分割成52个子流域,并实现了子流域编码及可视化显示。选取流域面积、河网密度及流域形状系数作为子流域分类因子。采用分层分类法,构建了8类型分类标准体系,完成子流域分类。分类结果表明：子流域以狭长型小流域、狭长型大中流域及卵形低河网密度小流域为主,流域整体河网密度大,易形成陡涨陡落的洪水,另外地形破碎植被稀疏,水土流失严重。
[Abstract]:As an important spatial data, digital elevation model (DEM) has been widely used in many fields. In the background of watershed digitization, automatic extraction of watershed features by using DEM has become the focus of hydrology research. Watershed characteristics are the basic data of hydrological simulation and watershed management, and the accuracy and efficiency of these studies are directly related to the accuracy and efficiency of these studies. In recent years, with the development of 3s technology and the improvement of DEM precision and acquisition speed, the research of watershed feature extraction has been further developed, and many new theories and methods have emerged. This paper systematically introduces the relevant knowledge of digital elevation model, reviews the research progress of extracting watershed features based on DEM at home and abroad, introduces the basic concepts of watershed features, and discusses the algorithm basis and method of extracting watershed features based on DEM. Taking Sanchuan River Basin as an example, the automatic extraction of watershed features is realized by using advanced GIS technology, and the accuracy of the extraction results is analyzed, and the whole watershed is divided and the sub-watershed is classified. The results are as follows: (1) the extracted terrain factors are in good agreement with the actual situation of the watershed, which indicates that the extraction algorithm and the applied tools are mature and effective. The results show that the average slope of the river basin is 13.34. The slope of the river source area is larger than that of the river valley area, and the slope degree decreases with the decreasing of the topographic elevation, the slope of the valley and residual source area changes slightly, and the gully area changes greatly. The distribution of ridge line and valley line in the watershed is clearly represented by the slope direction change rate; the distribution of slope direction is more uniform in eight directions; according to the curvature characteristic value, the terrain of valley area is flat, and the valley in non-valley area alternates with longitudinal and transverse concave and concave convex. (2) the steps of extracting river network include preprocessing of: Dem, determination of grid flow direction, calculation of accumulative amount of confluence, determination of optimal critical threshold, generation of grid network, vectorization of grid river network and modification and smoothing of river network. A new method is proposed to determine the optimal critical threshold value by the correlation between the cumulative threshold value and the density of the river network. The repeated tests show that the extracted river network is the most suitable for the actual situation when the threshold value is 6500, and it is also shown that this method is feasible. (3) based on the grid network flow data, the whole watershed is divided into 52 subbasins, and the subbasin coding and visualization are realized. Watershed area, river network density and watershed shape coefficient are selected as subwatershed classification factors. By using stratified classification method, 8 types classification standard system was constructed to complete subbasin classification. The classification results show that the subbasin is dominated by narrow and narrow small watershed, long and narrow large and medium watershed and small valley with low density of oval river network. The whole river network of the basin is dense and easy to form flood with steep rise and steep fall, in addition, the topographic vegetation is sparse. Soil erosion is serious.
【学位授予单位】：太原理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P208

【参考文献】