顾及局部微地形特征的DEM洼地处理算法

A Method of Depression Filling with Consideration of Local Micro-Relief Features

基于数字高程模型(DEM)的洼地处理过程是分布式水文建模的基础步骤,但也是极其耗时的环节。随着DEM数据量的不断提高,洼地处理的效率提升成为解决当前分布式水文分析的重要突破口。该文在对局部微地形特征剖析的基础上,提出了局部微地形漫水算法(MFF算法),该方法分析了洼地处理过程中的8种局部微地形模式,并提出了两种洼地处理过程中的冗余点判别方法。通过对微地形中冗余点、洼地、平地的优化处理,实现了算法效率的有效提升。最后,以70个不同数据量的DEM为实验数据,分析了MFF算法的正确性与计算效率。实验结果表明:MFF算法在保证运算正确性的基础上,相比W&L算法,执行效率平均提高40.13%,最大提高57.21%,可望为DEM高效水文分析提供新的方法。

Depression filling is a basic operation step in the study of distributed hydrological model based on Digital Elevation Models （DEM s） . T he Priority- Flood （ PF） approach is a widely used method for traditional depression filling process due to its relative high efficiency when dealing with a small size of DEM.However,it is still a timeconsuming procedure which should be improved significantly.With the increase of DEM resolution, it is a new breakthrough that the efficiency of depression filling should be further improved. In this paper,we find that significant redundant calculations in local micro- relief areas widely exist in conventional Priority- Flood algorithm. Although these redundant calculations seem acceptable when dealing with a small size of DEM, while it is quite time consuming when dealing with large size of DEM. A Micro- relief Flood Fill （ MFF） algorithm is proposed which takes local micro terrain features into full consideration. In this method, redundant points, depressions and flats will be processed optimally, in order to further promote the efficiency of depression filling. And then, 70 DEM data with differ-ent sizes are used to test the newT algorithm. A comparative analysis is also conducted to investigate the accuracy and efficiency. Experimental results show that the M FF algorithm can not only fill the depression accurately, but also improve the efficiency up to 40 13% than that of W＆ L algorithm on average and 57. 21% at most.This method is expected to provide anew approach to high- efficient hydrological analysis based on DEMs.