基于DEM的平缓地区水系提取和流域分割的流向算法分析

本文以地势较为平坦的秦淮河流域为实验样区,以高精度DEM(5m分辨率)和地形图水系为基础数据,对比分析了单流向算法(D8算法)、多流向算法(Dinf法)以及添加数字化河道信息后的单流向算法(AgreeD8算法)3种算法下水系提取和流域分割的结果。实验结果表明,提取得到的水系更逼近于实际河网,该算法既有效提高了水系提取和流域分割的精度,又保留了D8算法模型简单、稳定性强、运行效率高等优点。

DEM-based flow direction algorithms study of stream extraction and watershed delineation in the low relief areas

The paper explored performances of several different flow direction algorithms,namely D8 algorithm,Dinf algorithm and Agree & D8 algorithm.Among these algorithms,D8 algorithm is a kind of single flow direction algorithms,Dinf algorithm is one of multiple flow direction algorithms,while Agree & D8 algorithm is a method which introduces a real river channel vector map to DEM firstly,and then calculates the flow direction using D8 algorithm based on DEM.Quanhuai Basin,located at Nanjing of Jiangsu province,a large,flat catchment with an area of over 2600 km2was selected to perform the comparisons.And the DEM with 5 meter resolution and the real river channel vector map with the same scale were used.The experimental results showed that Agree & D8 algorithm offered the best stream extraction and watershed automatic delineation results,which approached to the real channel network and real sub-watershed most.Agree & D8 algorithm,which not only took the advantage of D8 algorithm,such as simple,steady and high efficiency,but also effectively improved the accuracy of river channel extraction and watershed automatic delineation,was a proper method to extract catchment properties in large and low relief area.