| 引用本文: | 郑 倩,史海滨,李仙岳,等.基于AGREE算法与BURN-IN算法的平原灌区DEM河网提取问题剖析[J].灌溉排水学报,2019,38(5):122-128. |
| ZHENG Qian,SHI Haibin,LI Xianyue,et al.基于AGREE算法与BURN-IN算法的平原灌区DEM河网提取问题剖析[J].灌溉排水学报,2019,38(5):122-128. |
|
| 摘要: |
| 基于数字高程模型（DEM）的区域水系提取应用较为广泛，当提取对象为平原灌区时存在突出问题，如河道断裂、平行河道的出现以及末端支流的提取未果等造成的子流域划分结果与实际轨迹不符等现象，数字化河网的精确程度直接影响后续的区域水文分析，故该问题值得关注。【目的】优化DEM河网提取结果。【方法】基于Agree算法和Burn-in算法下的提取河网，并从浅表、DEM因子与河网提取的相关关系等方面进行原因分析，进一步提出解决方法。【结果】2种算法下平原灌区河网提取问题的主要原因是DEM离散误差、栅格像元值的真实程度、栅格像元的精度以及地表的无渗漏假设；提出的解决方法为栅格像元水平分辨率的提高与主河道断裂部分的栅格重计算。对于坡度范围为0°~10°的平原灌区，坡度对地表水流方向的形成影响较小，不具备影响河网提取问题的能力；栅格像元水平分辨率提高10 m×10 m时，平原灌区2种算法下数字河道提取首要问题成功解决；主河道断裂部分的栅格重计算虽成功提取原断裂部分河道，但有新的断裂部分生成，未从根本解决。【结论】Agree算法和Burn-in算法下的基于DEM的平原灌区河网提取问题，经一系列原因剖析和解决方法的提出与应用，提取结果较为满意，提倡学者能够结合2种方法思路作为解决类似平原灌区的提取数字河网出现的问题。 |
| 关键词: Agree算法； Burn-in算法； DEM； 灌区； 河网 |
| DOI：10.13522/j.cnki.ggps.20180143 |
| 分类号: |
| 基金项目: |
|
| Identifying the Sources of Errors in Using DEM to Extract Irrigation-drainage Network in Plain Areas Using AGREE and BURN-IN Algorithm |
|
ZHENG Qian, SHI Haibin, LI Xianyue, LI He, BU Huailiang, LI Zhengzhong
|
|
1.College of Water Conservation and Civil Engineering, Inner Mongolia Agricultural University, Huhhot 010018, China; 2.Shahao Canal Experimental Station, Hangjinhouqi, Bayannaoer City, Inner Mongolia, Bayannur 015000, China
|
| Abstract: |
| 【Objective】 The digital elevation model (DEM) has been widely used to map irrigation network but its application in irrigation district in plain region is problematic, characterized by network discontinuity, parallel channel and failure and missing end branches. This could have a consequence for the hydrological analysis. This paper analyzes the source of the errors.【Method】 The network was extracted by the Agree and the Burn-in algorithm, we analyzed the relationship between the characteristics with ground surface elevation and DEM factors. 【Result】The sources of the errors in extracting channel network in plain regions are the discrete error of the DEM, inaccuracy of grid cell values, resolution of grid images, and the assumption of no surface water percolation. 【Conclusion】 For channel network in plain with slope angle of 0°~10°, there was limited surface runoff and it thus cannot used to extract the channel network. The solution is to improve the resolution of the grid cells and recalculate the channel fracture part of the grid recalculation. Among them, the method improving grid pixel precision successfully solved the problem of digital river channel fracture under the Agree algorithm and Burn-in algorithm in the plain irrigation area. Although the grid recalculation of the main channel fracture part has successfully extracted some of the original fault channels, partially generated, failed to solve fundamentally. Therefore, this paper advocates that scholars can combine the two methods and ideas as a solution to the problem of extraction of digital river network in similar plain irrigation areas. |
| Key words: Algorithm of Agree; Algorithm of Burn-in; DEM; irrigation area |
