融合分布式目标的矿区长时序InSAR地表形变监测.pdf
硕士学位论文 融合分布式目标的矿区长时序 InSAR 地表 形变监测 Long Time-series InSAR Surface Deation monitoring with Distributed Scatterers in Mining area 作 者李 毅 导 师闫世勇 副教授 中国矿业大学 二零一九年五月 万方数据 学位论文使用授权声明学位论文使用授权声明 本人完全了解中国矿业大学有关保留、使用学位论文的规定,同意本人所撰 写的学位论文的使用授权按照学校的管理规定处理 作为申请学位的条件之一, 学位论文著作权拥有者须授权所在学校拥有学位 论文的部分使用权, 即①学校档案馆和图书馆有权保留学位论文的纸质版和电 子版,可以使用影印、缩印或扫描等复制手段保存和汇编学位论文;②为教学和 科研目的,学校档案馆和图书馆可以将公开的学位论文作为资料在档案馆、图书 馆等场所或在校园网上供校内师生阅读、浏览。另外,根据有关法规,同意中国 国家图书馆保存研究生学位论文。 (保密的学位论文在解密后适用本授权书) 。 作者签名 导师签名 年 月 日 年 月 日 万方数据 中图分类号 P23 学校代码 10290 UDC 528 密 级 公开 中国矿业大学 硕士学位论文 融合分布式目标的矿区长时序 InSAR 地表 形变监测 Long Time-series InSAR Surface Deation monitoring with Distributed Scatterers in Mining area 作 者 李毅 导 师 闫世勇 副教授 申请学位 工学硕士 培养单位 环境与测绘学院 学科专业 摄影测量与遥感 研究方向 对地观测与定量遥感 答辩委员会主席 秦凯 评 阅 人 二○一九年五月 万方数据 致谢致谢 时光荏苒,岁月如梭,不知不觉硕士三年的求学生涯就要结束了,仿佛初次 来到矿大还在昨天,在此度过的美好时光依然历历在目。回想这三年里,我在矿 大有幸认识了许多好老师,结识了一些新朋友。借此在论文完成之际,我想向所 有关心和帮助我的老师、亲友们致以最诚挚的谢意 首先要由衷地感谢我的导师闫世勇副教授,作为 InSAR 研究领域的启蒙老 师,感谢您把我领入了一个如此新兴的研究方向,并为我搭建理想的科研环境和 平台;感谢您在论文选题、开展研究,到论文撰写、修改直至定稿过程中给予的 悉心指导及亲切关怀, 使我在硕士期间的学习和研究得以顺利完成。在硕士研究 生的三年里,闫老师包容谦虚、平易近人的为人处事,严谨求实、认真负责的工 作态度, 给我留下了极为深刻的印象, 这都是我在今后工作中值得去学习的地方。 师从三载,闫老师不仅仅传授了我知识,更重要的是教会了我从事科学研究的方 法, 使我受益终身, 在此向您表以我最真挚的感谢, 希望您身体健康, 工作顺利 感谢中国矿业大学环境与测绘学院的谭琨教授、 赵银娣和孙久运副教授在论 文选题过程中提供的宝贵意见, 以及刘涛健、蒋冬梅老师在科研工作中给予过的 极大帮助,感谢你们在我硕士研究生求学期间给予的指导和关怀。感谢实验室已 经毕业的刘茜茜师姐, 吴岳、 周洪月师兄, 以及郑一桐、 刘竞龙、 蒋金雄、 史珂、 原刚、彭锴、冯瀚和张敏等师弟师妹,是你们分担了我的部分工作,使我有更为 充足的时间作研究, 与你们一起学习和生活的时间是我人生中最难以忘怀的一段 美好时光,祝你们工作顺利、学业有成、身体健康 感谢同年级地信遥感班的刘润、 侯增福、 王一博、 武复宇、 张玉沙、 王惠敏、 戴玉玲、樊文智和邹家恒等同学在学习、工作和生活中给予的帮助和鼓励,与你 们在一起的时光充满着欢笑和快乐,同样祝你们工作顺利、学业有成。感谢中国 地质大学(北京)的郑岳泽硕士、中国科学院遥感与数字地球研究所的赵常军博 士在数据处理等方面的交流和给予的帮助。最后,由衷地感谢资源与地球科学学 院杨宝婷硕士对我六年来的支持和陪伴,感谢你给我带来的快乐与鼓励,使我的 研究生生活更加的丰富多彩。希望以后继续有你的支持与陪伴 这里我还要深深地感谢含辛茹苦把我养大, 并且始终大力支持和关心的父母, 感谢你们一直以来的对我的鼓励、 理解与信任,你们永远是我道路前进上最坚实 的后盾。愿你们身体健康,万事如意 最后,再次向所有给予我帮助的师长、亲人和朋友们致以最真诚的感谢 李毅 2019 年 04 月 于徐州 万方数据 I 摘摘 要要 随着在轨 SAR 卫星增多、SAR 数据的海量积累,以及地表形变监测需求的 不断提高,侧重于分析高相干点目标的传统时序 InSAR 分析方法在矿区地表形 变监测中,往往存在测量点数量少与分布密度低等不足,在实际应用中难以直观 反映出矿区地表沉降影响范围。为了充分挖掘多时相 SAR 数据在地表形变监测 中的潜力,克服矿区自然地表对当前时序 InSAR 分析技术的局限性,根据地面 目标散射特性的不同,利用分布式目标开展矿区地表形变监测成为可能。本文以 常规 PS-InSAR 技术为基础, 开展融合分布式目标的时序 InSAR 技术监测矿区地 表形变的方法研究, 并将其应用到沛北矿区地表形变监测中,更好地反映出该区 域地表沉降的时空演变特征。由此,本文主要研究工作和取得研究成果如下 (1) 对 InSAR 及 D-InSAR 技术原理与数据处理进行简要说明, 指出 InSAR 技术的局限性;并在此基础上引出和详细介绍当前 3 种时序 InSAR 分析技术体 系---以单一主影像为体系的 PS 技术、多主影像为体系的 SBAS 技术和混合时序 InSAR 分析技术; (2)在简要说明分布式目标的统计分布特性的基础上,详细介绍用以描述 分布式目标统计特性的二阶统计量---样本协方差/相干矩阵;对当前常用于同质 像元识别的多种非参数假设检验方法与参数假设检验方法进行系统性地阐述与 对比分析; (3)以干涉相位一致性作为基本约束条件提出了分布式目标的相位优化函 数模型;随后,从分布式目标的统计分布特性和多元散射机制考虑,基于相位优 化函数模型详细介绍利用最大似然估计MLE方法、样本协方差矩阵和样本相干 矩阵特征值分解C/T-EVD方法进行分布式目标相位优化的方法研究,并得到统 一形式的相位优化函数。从计算效率上讲,基于样本协方差/相干矩阵特征值分 解的相位优化方法无需迭代处理,存在计算效率高的优势。 (4) 基于常规 PS-InSAR 技术和融合分布式目标的 DS-InSAR 方法, 利用不 同时期的 SAR 数据对沛北矿区进行地表形变监测, 得到沛北矿区采时 (2007.02- 2011.03)与现阶段(2017.01-2018.12)长时序地表沉降监测结果。首先,结合矿 区地物覆盖类型探讨了融合分布式目标进行地表沉降监测的优势所在, 并分析了 重点沉降区域的沉降情况;最后,同常规 PS-InSAR 监测结果采取交叉验证的方 式说明了分布式目标监测结果的可靠性。研究表明①融合分布式目标的 DS- InSAR 监测结果在空间分布趋势与常规 PS-InSAR 监测结果基本一致;而且对分 布式目标采取不同相位优化方法得到的 DS-InSAR 沉降监测结果之间一致性较 万方数据 II 好,沉降结果差异分布较小。②融合分布式目标的 DS-InSAR 方法通过提高测量 点MPs数量与空间分布密度,可以有效改善地表形变信息的连续性与完整性, 这样有助于定位出地表沉降发生位置, 确定出由煤矿开采引起的地表沉降影响范 围;也有助于增加对矿区地表沉降区域的探测几率。③融合分布式目标的 DS- InSAR 方法在提高测量点MPs数量与空间分布密度的基础上,一定程度上可以 提高相位解缠的精度,获取可靠的地表沉降结果;而且还可以提高地表形变相位 在空间分布上的连贯性,有利于较大量级地表形变信息的提取。 该论文有图 42 幅,表 11 个,参考文献 112 篇。 关键词关键词地表形变监测;时序 InSAR 分析技术;分布式目标;沛北矿区;煤炭开 采 万方数据 III Abstract With the increase of orbital SAR satellites, the accumulation of SAR datasets, and the increasing requirements for surface deation monitoring, the conventional time- series InSAR analysis techniques mainly focused on the analysis of high coherence point targets and has the problem of insufficient measurement points and low distribution density in mining area, causing the difficult to visually reflect the influence range of surface deation of the mining area in practical application. In order to fully exploit the potential of multi-temporal SAR data in surface deation monitoring, to overcome the limitation of the natural environment of the mining area on the current time-series InSAR analysis technology, it is possible to explore the distributed targets for ground deation monitoring according to the different scattering characteristics of the ground targets. Based on the conventional PS-InSAR technology, this paper researches on of monitoring the deation by using the time-series InSAR technology with distributed targets, which had then been applied to monitor and analyzes the surface deation of Peibei mining area, making its temporal and spatial evolution characteristics more clearer. Therefore, the main research work and concusions of the study are as follows 1 The principle and data processing of InSAR and D-InSAR were briefly explained, and point out the limitations of InSAR technology; three kinds of Time - series InSAR analysis technology system -- PS-InSAR technology with Single master image, SBAS technology with Multi master image and hybrid Time-series InSAR analysis technology were then presented; 2 The statistical properties of distributed targets are briefly described, and the second statistical moments of statistical characteristics of distributed targets -- Sample Covariance/Coherence matrix are described in detail; Then, a variety of Non - Parametric hypothesis test and Parametric hypothesis test s used for Statistically Homogeneous PixelsSHPs selection have been explained and summarized, systematically; 3 The basic concept of interferometric phase consistency is elaborated, and as a basic constraint condition, the phase optimization function model of Distributed target is proposed. Then, from the statistical properties and the multi-scattering mechanism of distributed target, the Maximum Likelihood EstimationMLE , the sample 万方数据 IV Covariance matrix and Coherent matrix eigenvalue decomposition C/T-EVD are used to implement the distributed target based on the phase optimization function model, and a unified of phase optimization function is introduced. In terms of optimize efficiency, the sample Covariance/Coherence matrix eigenvalue decomposition does not need iterative analysis, and has the advantage of high computational efficiency. 4 The conventional PS-InSAR technology and the DS-InSAR by integrating distributed targets had been utilized to monitor the surface deation of the Peibei mining area, the mining 2007.02-2011.03 and at the present stage 2017.01- 2018.12 of long-term surface deation of the Peibei mining area was obtained by using SAR data from different periods. First, combined with the type of land cover in mining area, we demonstrate the advantages of integrating distributed targets for surface deation monitoring, and analyze the deation of major subsidence areas. Compared with conventional PS-InSAR result, the reliability of distributed target monitoring results was illustrated in a cross-validated way. The research shows that① the DS-InSAR deation results are basically consistent with PS-InSAR results, in spatial distribution trend; the DS-InSAR deation results obtained by different phase optimization s for distributed targets keep well, and the deation rate difference is small. ②DS-InSAR combinded with Distributed targets can improve the continuity and integrity of deation ination by increasing the number of measurement points MPs and spatial distribution density, which help to locate the location of land subsidence and delineate the influence extent of surface subsidence caused by coal mining, and also helps to increase the probability of detecting surface deation areas in mining areas.③Compared with the conventional PS-InSAR results, the DS-InSAR based on the distributed target increase the number of Measurement Points MPs and spatial distribution density, which can make phase unwrapping more accurately to a certain degree, leading to improve the reliability of deation estimation. Moreover, its can improve the consistency of deation phase in spatial domain, which is beneficial to the extraction of significant surface deation ination. There are 42 figures, 11 tables and 112 references. Keywords ground deation monitoring; Time-series InSAR analysis; Distributed target; Peibei mining area; underground coal-mining 万方数据 V 目目 录录 摘摘 要要 ........................................................................................................................... I 目目 录录 .......................................................................................................................... V 图清单图清单 ........................................................................................................................ IX 表清单表清单 ..................................................................................................................... XIII 变量注释表变量注释表 ............................................................................................................. XIV 1 绪论绪论 ........................................................................................................................... 1 1.1 选题背景与研究意义............................................................................................. 1 1.2 国内外研究现状..................................................................................................... 3 1.3 主要研究内容......................................................................................................... 7 1.4 论文结构与章节安排............................................................................................. 8 2 InSAR 技术基本原理技术基本原理 .............................................................................................. 10 2.1 InSAR 技术 ........................................................................................................... 10 2.2 InSAR 技术局限性 ............................................................................................... 15 2.3 时序 InSAR 技术 ................................................................................................. 18 2.4 章节小结............................................................................................................... 25 3 分布式目标选取及其相位优化分布式目标选取及其相位优化 .............................................................................. 26 3.1 分布式目标的统计特征....................................................................................... 26 3.2 同质像元识别....................................................................................................... 29 3.3 分布式目标相位优化........................................................................................... 34 3.4 章节小结............................................................................................................... 51 4 基于基于 DS-InSAR 的沛北矿区长时序地表沉降监测的沛北矿区长时序地表沉降监测 .............................................. 53 4.1 研究区概况与数据源........................................................................................... 53 4.2 矿区采时长时序地表沉降监测与分析............................................................... 58 4.3 矿区现阶段长时序地表沉降监测与分析........................................................... 70 4.4 本章小结............................................................................................................... 82 5 结论与展望结论与展望 .............................................................................................................. 84 万方数据 VI 5.1 研究成果和结论................................................................................................... 84 5.2 问题与展望........................................................................................................... 85 参考文献参考文献 ..................................................................................................................... 86 作者简介作者简介 ..................................................................................................................... 94 学位论文原创性声明学位论文原创性声明 ................................................................................................. 95 学位论文数据集学位论文数据集 ......................................................................................................... 96 万方数据 VII Contents Abstract ......................................................................................................................... I Contents ....................................................................................................................... V List of Figures ............................................................................................................ IX List of Tables ........................................................................................................... XIII List of Variables...................................................................................................... XIV 1 Introduction .............................................................................................................. 1 1.1 Research Background and Significance ................................................................... 1 1.2 Research Status at home and abroad ........................................................................ 3 1.3 Main Research Contents .......................................................................................... 7 1.4 Route and organizational Structure .......................................................................... 8 2 Basic Theory of InSAR Technology ...................................................................... 10 2.1 Interferometric SAR Technology ........................................................................... 10 2.2 The limitations of InSAR technique ...................................................................... 15 2.3 Time Series Synthetic Aperture Radar Interferometry........................................... 18 2.4 Summary ................................................................................................................ 25 3 Distributed Target Selection and Its Optimal Phase Estimation ........................ 26 3.1 Statistical description of Distributed targets .......................................................... 26 3.2 SHP Identification .................................................................................................. 29 3.3 Optimal phase estimation ....................................................................................... 34 3.4 Summary ................................................................................................................ 51 4 Long Time Series surface deation monitoring based on DS-InSAR in PeiBei Mining Area ......................