高分三号和Sentinel-1/2融合监测锡亚琴冰川厚度变化研究  

Monitoring of Siachen glacier thickness change through integration of Gaofen-3 and Sentinel-1/2

作　　者：张瑞宇 蒋弥 李刚[1,2] 钟敏[1,2] 程晓[1,2] ZHANG RuiYu;JIANG Mi;LI Gang;ZHONG Min;CHENG Xiao(School of Geospatial Engineering and Science,Sun Yat-sen University,and Southern Marine Science and Engineering Guangdong Laboratory(Zhuhai),Zhuhai Guangdong 519082,China;Key Laboratory of Comprehensive Observation of Polar Environment(Sun Yat-sen University),Ministry of Education,Zhuhai Guangdong 519082,China)

机构地区：[1]中山大学测绘科学与技术学院,南方海洋科学与工程广东省实验室(珠海),广东珠海519082 [2]极地环境立体观测与应用教育部重点实验室(中山大学),广东珠海519082

出　　处：《地球物理学报》2025年第3期802-812,共11页Chinese Journal of Geophysics

基　　金：国家自然科学基金项目(42074008)资助。

摘　　要：冰层厚度变化是评估冰川物质平衡的重要环节,利用表面平行流假设并结合升降轨SAR数据集估算冰厚变化的方法近年来被视为研究热点.然而,这类方法往往需要通过正则化手段抑制观测不足引发的解算不确定性,而传统二范数的解算策略也易受到观测粗差引起的误差传播.为解决这一问题,本研究以锡亚琴冰川为例,提出结合Sentinel-1/2和高分三号数据估算偏移量,通过联合解算偏移量和迭代重加权最小二乘得到冰川三维流速和冰层厚度变化.通过多源数据增加观测量的思想取代正则化,消除正则化因子主观性对结果的影响.在表面平行流假设的条件下监测显示2018年7—9月锡亚琴冰川非表面平行流速约为-0.018 m·d^(-1),冰川主要表现为消融.与仅利用Sentinel-1升降轨道数据的解算结果相比,结合Sentinel-2和高分三号数据显著提高了三维流速的估计精度.在东西、北南和垂直方向上,精度分别提升了51%、30%和31%.研究结果表明多影像联合解算有助于提高冰川厚度变化精度,而我国高分三号卫星数据的应用显著提升了三维流速的解算精度.Ice thickness variation plays a critical role in assessing glacier mass balance.In recent years,the method of estimating changes in ice thickness,combining the assumption of surface parallel flow with ascending and descending synthetic aperture radar(SAR)data sets,has gained considerable research attention.However,this method often employs the regularization technique to handle the rank-deficiency of design matrix due to limited observations,while the outliers in SAR offset observations distort the results from ordinary least-squares inversion.To address these issues and further extend previous studies into glacier thickness monitoring more accurately,this paper presents a novel approach that integrates Sentinel-1/2 and Gaofen-3 datasets over the Siachen glacier area.The idea behind the integration is that the increase of multi-source observations from different look angles reduce the subjective nature of regularization parameters and simultaneously improving the estimate accuracy of 3D displacements and therefore the thickness change.Moreover,the introduction of iteratively reweighted least-squares method can mitigate the influence of outliers.Under the assumption of surface parallel flow,the results reveal that during the period of July to September 2018,the Siachen glacier exhibited a non-Surface Parallel Flow of approximately−0.018 m·d^(−1),indicating significant glacial ablation.Compared to the results obtained from Sentinel-1 ascending and descending orbit data,the integration of Sentinel-2 and GF-3 data has significantly improved the accuracy of three-dimensional flow velocity estimation.In the east-west,north-south,and vertical directions,the accuracy has increased by 51%,30%,and 31%,respectively.The research also demonstrates the importance of GaoFen-3 data in exploring 3D glacier flow velocities and thickness change.

关 键 词：冰川厚度变化 高分三号 Sentinel-1/2 偏移量跟踪 抗差估计 

分 类 号：P343[天文地球—水文科学]