基于DEM的江西潦河流域河网信息提取方法

介绍了GIS环境下数字高程模型（DEM）的预处理、水流方向确定、汇流累积量确定、河网生成，以及其他水文信息提取的原理和方法；基于DEM对江西潦河流域进行河网提取和流域边界的划分，并对试验结果进行了比较和分析。结果表明，利用GIS空间分析功能，基于DEM提取的潦河中上游数字河网与实际河网基本吻合；DEM精度和汇流能力阈值对提取的河网精度有着明显的影响，对于中小流域的潦河来说，1：50000精度的DEM和汇流能力为1000的阈值可满足实际应用的要求；基于DEM和ArcGIS提取的数字化河网，其水文要素（如坡度等）可直接作为流域预报模型参数，不仅大大提高了模型的效率，而且还可以方便地利用遥感等数据源，数据获取快速且成本低。从宏观效果上看，利用现有的GIS软件，基于DEM数据提取河网、流域范围等水文信息，从提取效率和结果的精度来说都是切实可行的；从微观上讲，已有的算法还无法确保计算结果与某些特殊地形区域（如平坦区域）的实际河网一致。

Drainage Network Extraction in Liaohe Based on DEM

In this paper, the principles and methods of the digital elevation model （DEM） preprocessing, the flow direction determination and afflux accumulation determination, watershed generation, as well as other watershed information extraction principles and methods under the GIS environment are introduced. The watershed extraction and the basin boundary division based on DEM to the Liaohe drainage area are carried on, and the experiments are comparatively analyzed. The results indicate that using the spatial analysis function of GIS, the digital watershed based on the DEM extraction on the middle and upper reaches of the Liaohe is basically coincident with the actual one. The precision of DEM and the threshold value of affiux ability influence obviously on the precision of extraction watershed. To the small basin, 1：50 000 with DEM and 1 000 with the afflux ability may satisfy the practical application. Based on the digitized watershed extracted by using DEM and the ArcGIS, the hydrological features （such as slope） may be used directly as the basin prediction model parameter. It can not only raise the efficiency greatly, but also use remote sensing data pools conveniently. It can gain the data quickly and cost low. Macroscopically, to extract hydrology information from watershed, basin perimeter based on DEM by using the existing GIS software is practical and feasible. However, the existing algorithm is still unable to guarantee the computed result be consistent to the actual watershed dealing with certain extreme terrain （such as flat site）.