一种融合Priority-Flood算法与D8算法特点的河网提取方法

发布时间：2018-04-10 02:23

本文选题：河网提取　切入点：Priority-Flood算法　出处：《地理与地理信息科学》2017年04期

【摘要】：基于数字高程模型(Digital Elevation Models)进行河网提取是数字地形分析的重要研究内容和水文分析的关键步骤,而流向算法是其中最重要的部分。然而,传统的流向算法在平坦地形区域容易形成平行河网。为解决平行河网问题,该文提出了一种结合Priority-Flood算法漫水思想和改进D8算法的流向算法。借助Priority-Flood算法的漫水思想,可以保证流向符合宏观地形特征;同时将地形划分为上坡、平地和下坡3种类型,辅以PriorityFlood算法可以实现在不填洼的基础上计算流向,而且针对不同地形,采用相应的改进D8流向算法,从而保证流向符合微观地形特征。通过这两种算法的结合,可以直接使用原始DEM提取河网,消除了填洼过程,提高了流向计算效率,同时消除了填洼造成的大量平地,有效地解决了平地区域平行河网问题。最后,选取全球不同区域SRTM数据对JD算法与新算法进行对比实验,结果表明,新算法能够有效地改善平行河网问题。
[Abstract]:The extraction of river network based on digital elevation model (Elevation models) is an important research content of digital terrain analysis and a key step of hydrological analysis, and the algorithm of flow direction is the most important part.However, the traditional flow direction algorithm is easy to form parallel river network in flat terrain area.In order to solve the problem of parallel river network, this paper proposes a new flow direction algorithm which combines the idea of overflowing with Priority-Flood algorithm and the improved D8 algorithm.With the help of the idea of overflowing of Priority-Flood algorithm, the flow direction can be guaranteed to conform to the macroscopic topographic characteristics, and the terrain can be divided into three types: uphill, flat and downhill, and the PriorityFlood algorithm can be used to calculate the flow direction on the basis of no filling.Moreover, the improved D8 flow direction algorithm is adopted for different terrain, so as to ensure that the flow direction accords with the microscopic terrain characteristics.Through the combination of these two algorithms, the original DEM can be directly used to extract the river network, eliminating the process of filling, improving the efficiency of calculation of flow direction, eliminating a large number of flat land caused by the filling, and effectively solving the problem of parallel river network in the flat land area.Finally, the SRTM data from different regions of the world are selected to compare the JD algorithm with the new algorithm. The results show that the new algorithm can effectively improve the parallel river network problem.
【作者单位】：南京师范大学虚拟地理环境教育部重点实验室;江苏省地理信息资源开发与利用协同创新中心;
【基金】：国家自然科学基金项目(41601411;41671389;41401440)
【分类号】：P208;P333

【参考文献】