基于ICESat-2数据的山区细小河流断面提取  

Extraction of small river cross-section in mountainous based on ICESat-2 data

作　　者：李丹[1,2] 覃超[2,3,4] 薛源 徐梦珍[2,3,4] 刘英杰 LI Dan;QIN Chao;XUE Yuan;XU Mengzhen;LIU Yingjie(Emergency Science Research Academy,China Coal Research Institute,Beijing 100013,China;State Key Laboratory of Hydroscience and Engineering,Tsinghua University,Beijing 100084,China;Key Laboratory of Hydrosphere Science of the Ministry of Water Resources,Tsinghua University,Beijing 100084,China;Department of Hydraulic Engineering,Tsinghua University,Beijing 100084,China)

机构地区：[1]煤炭科学研究总院有限公司应急科学研究院,北京100013 [2]清华大学水圈科学与水利工程全国重点实验室,北京100084 [3]清华大学水利部水圈科学重点实验室,北京100084 [4]清华大学水利水电工程系,北京100084

出　　处：《清华大学学报（自然科学版）》2024年第12期2122-2131,共10页Journal of Tsinghua University(Science and Technology)

基　　金：清华大学水圈科学与水利工程全国重点实验室项目(sklhse-2023-B-03,2023-KY-02)。

摘　　要：河流以一定的边界形态承载着水流及物质的输移,受现场量测困难、水文观测站点有限及数字高程模型(digital elevation model,DEM)精度限制,导致分布在江河源区、山区和偏远地区的细小河流断面信息积累极为匮乏,阻碍了河流水文、水动力过程等方面的研究。国际水文协会(IAHS)自本世纪初力倡解决无测站流域水文预测难题,使得高精度提取无测站流域河流水文信息变得越来越重要。ICESat-2 (ice,cloud and land elevation satellite-2)卫星测高任务具有在缺资料地区采集高密度地表点云的能力。该文以黄河中游一级支流皇甫川流域缺资料地区河宽小于10m的细小河流为研究对象,提出了一种利用ICESat-2 ATL03数据提取山区细小河流断面形态的方法。通过光子点云中、高置信度选择、平滑滤波去噪处理和DEM构建,提取了3个不同点位细小河流断面形态并与无人机原位测量结果进行对比分析。结合中、高置信度点云选择和滤波去噪可以有效地去除光子点云噪声,去噪率在63%以上;基于ATL03数据提取的地面点完整度和丰富度优于利用ATL08产品对ATL03分类得到的地面点;基于ATL03数据提取的河流断面形态结果与无人机原位测量结果基本一致(R^(2)>0.96,RMSE=0.69m)。研究结果初步证明了ICESat-2测高数据用于缺资料地区细小河流断面提取的可行性,可为全流域三维河网构建、水文水动力过程模拟等提供技术支持。[Objective] Rivers carry water and material transport within certain boundary forms.Due to the difficulties of in-site measurement,limited hydrological observation stations,and the precision constraints of digital elevation model(DEM),there is a significant scarcity of information on small river cross-sections distributed in river source areas,mountainous regions,and remote areas,which hinders research on river hydrology and hydraulic processes.Since the beginning of this century,the International Association of Hydrological Sciences(IAHS) has been advocating for solutions to the challenges of hydrological prediction in ungauged basins(PUB).Despite large-scale remote sensing technology is increasingly applied to the extraction of river hydraulic parameters,the spatial resolution of satellite altimetry data is too low for small rivers(river width less that 150 m),which account for a high proportion of river networks.The National Aeronautics and Space Administration(NASA) launched the ICESat-2(Ice,Cloud and land Elevation Satellite-2) satellite in 2018.This satellite was equipped with the Advanced Topographic Laser Altimeter System(ATLAS),a photon-counting LiDAR system for the first time.The light spot(footprint) it projects onto the Earth's surface has a diameter of about 17 m,with a center-to-center distance between spots of only 0.7 m.This allows for the acquisition of photon point cloud data with smaller spots and higher density along the track.The high-density photon point cloud provided by ICESat-2 offers the possibility of extracting hydrological parameters of narrow rivers with high precision.This study focuses on the small rivers with a river width of less than 10 m in the Huangfu River basin,a first-order tributary of the middle reaches of the Yellow River,which is a data-scarce region.[Methods] A method for extracting the cross-sectional morphology of small rivers using ICESat-2 ATL03 data has been proposed.First,photons with medium to high confidence levels are selected to eliminate most of the noise.Th

关 键 词：ICESat-2 数字高程模型 河流断面形态 平滑滤波 无人机 

分 类 号：TV211.1[水利工程—水文学及水资源]