从规则格网DEM中提取沟谷网络的方法研究

贺晓晖，陈楠

(福州大学 福建省空间信息工程研究中心，福州 350002)

摘要：对TauDEM软件提取沟谷网络的方法进行了研究。以某地区1∶1万地形图为基础数据生成适宜分辨率的规则格网DEM数据；运用TauDEM 5.0软件对DEM数据进行洼地填充、水流方向计算、汇流累积量计算；通过沟谷高差统计分析确定汇流累积量的临界值，提取出该地区最佳沟谷网络，并与ArcGIS水文分析模块提取的沟谷网络进行了比较。研究表明，该方法提取的沟谷网络比较精细，伪沟谷较少，提取效率也明显较高，同时减少了反复实验获取合理阈值受主观因素的影响。

关键词：DEM；TauDEM软件；汇流累积量阈值；沟谷提取

doi:10.3969/j.issn.1000-3177.2015.01.023

中图分类号：P208文献标识码：A

Stream Network Extraction Based on Digital Elevation

Model Data of Regular Grids

HE Xiao-hui，CHEN Nan

(SpatialInformationResearchCenterofFujian，FuzhouUniversity，Fuzhou350002)

Abstract：This paper studied the method of extracting stream networks using TauDEM software.Firstly，regular grid DEM data was generated with appropriate resolution taking the source data from topographic maps with the scale 1：10 thousand.Secondly，depressions in DEM were filled，flow directions and flow accumulation were calculated using the software TauDEM5.0.Finally，the thresholds of the flow accumulation were determined by stream drop analysis and statistics，and the best valley network of this region was extracted and compared with the extraction results by the hydrological analysis module of ArcGIS.This research shows that the extracted stream networks using this method is fine，the pseudo valleys are less，the efficiency of extraction is significantly higher，and at the same time，the influence of subjective factors in the repeated experiments for obtaining reasonable thresholds are reduced.

Key words：DEM；TauDEM software；threshold for flow accumulation；valley extraction

1引言

数字高程模型(简称DEM)通过有限的高程数据实现对地形表面的数字化模拟。DEM包含丰富的地形信息，利用数字地形分析技术可提取其隐含的地形特征信息，众多学者对如何提取地形特征信息进行了深入细致的研究。沟谷在地形上具有明显的物理特征，基于DEM提取沟谷网络的算法可分为3种：基于图像处理的算法、基于地表几何形态分析的算法、基于地表流水物理模拟分析的算法[1-11]。1984年O’Callaghan和Mark提出了基于地表流水物理模拟分析算法[12]，该算法的关键是无洼地高程矩阵、水流方向矩阵和汇流累积量矩阵的生成。由于该算法提取沟谷的效率较高，在正确处理3个矩阵的基础上，精度能够得到保证，因此众多学者致力于对该算法的改进研究[13-18]。……