【作者】 俞晓莹；

【导师】 丁晓利； 李志伟；

【作者基本信息】 中南大学 ， 测绘科学与技术， 2012， 博士

【摘要】 合成孔径雷达干涉测量技术(D-InSAR)可应用于大范围、全天候、实时的快速地表缓慢形变监测,它已日益成为人们监测地表形变、地裂缝、地震等地质灾害的一项重要技术手段,同时它也大大弥补了传统的监测手段如GPS、水准测量等的不足。近年来,在D-InSAR技术的基础上发展起来的PS-InSAR技术、短基线集的D-InSAR技术(SBAS)、点差分干涉测量技术(IPTA)等已经成功运用于地壳形变、城市地表沉降监测、地震监测、滑坡监测以及冰川运动等领域之中。然而,随着应用领域的进一步拓宽,这些技术的研究还处于起步阶段,里面还涉及到的许多算法还有待于进一步改善。在研究D-InSAR关键技术的基础上,本研究重点就PS-InSAR技术、SBAS技术及IPTA技术中的关键问题进行分析研究。PS-InSAR技术可以实现高分辨率的点目标监测,但对于长时间序列的监测存在着无法正确解缠的问题,且受大气影响较大；运用SBAS技术可以减少数据成本,提高数据的应用效率,但由于SBAS技术采用影像多视处理技术,导致空间分辨率降低,降低了PS点的监测精度；IPTA技术运用了矢量数据的高效率的数据结构。该数据结构大大节省了数据的存储空间和提高了运行效率,但同样存在PS-InSAR中解缠困难和受大气影响较大的问题。对以上多种方法进行深入分析和研究之后,本论文发展了一套既能检测线性形变,又能检测周期性和非线性形变的技术流程和方法。基于改进的SBAS算法不是简单的将两种算法的叠加,而是对算法中的仍然存在的问题进行了改进,如利用二元树的复数小波滤波方法对干涉图中的噪声加以去除；针对大气延迟问题提出了利用MERIS数据进行ASAR大气延迟相位的改正；针对低分辨率的传统的SBAS技术利用PS点提高监测精度。在理论研究的基础上,选择河北沧州市区作为实验场,采用了ENVISAT卫星的ASAR数据进行实验分析,得到相应的监测结果。为了验证结果的可靠性,本文利用研究区域监测井的历年水位数据与实验的形变监测结果进行趋势对比分析,并结合形变和地下水位的关系,分析出研究区域地下水的地球物理特性。论文的主要研究工作可分为以下四个方面。1)提出了一种基于二元树复小波的差分干涉图滤波方法。针对干涉相位噪声严重影响InSAR生成数字高程模型和地表形变监测精度的问题,在对比研究多种滤波方法的优缺点的基础上,利用本文提出的二元树小波滤波方法分别对模拟SAR数据和真实的SAR数据进行滤波,并将其滤波结果与其它常用的SAR数据的滤波方法的滤波结果进行精度比较,得出基于二元树复小波的滤波方法能大大提高滤波的精度。该方法有效地去除了差分干涉图中由于多种失相干问题引起的噪声,为后续点差分干涉图的处理提供了更有效的数据源。2)提出了一种基于MERIS水汽数据的ONN模型与Von_karman kriging模型融合的空间插值方法。应用Von karman kriging方法对局部水汽进行空间插值可以有效地提高传统的OK方法的插值精度；同时它避免了由于样本较少时引起的插值结果不连续的现象,提高了插值精度。通过实验表明：在季节性变化很明显的区域,由于水汽引发的相位延迟是不容忽视的,它将严重影响形变监测的精度,甚至得到错误的监测结果,因此利用MERIS数据进行大气改正是非常必要的。本文利用沧州市的差分干涉对和同时间的MERIS数据进行了实证分析,分析结果表明：沧州地区受大气影响的较为明显,必须去除大气效应。3)提出了一种改进的SBAS方法。该方法有效地解决了系列点差分干涉相位方程中的秩亏问题,并有效地分离出周期性形变、非线性形变,为更真实的表达地表形变的变化提供了更精确的监测结果。改进的SBAS方法融合了SBAS的高时间分辨率特点、IPTA的高效处理模式、高空间分辨率的特点以及高效的解缠方法。文中提出相应的算法和数据处理流程。在此基础上,利用ASAR数据对沧州市的地表形变进行了提取实验,同时对大气相位和噪声也采取了时空滤波方法,并且有效地分离了线性与非线性形变,得到了相应的监测结果。4)将地面形变结果与监测井的地下水位进行地球物理特征分析。改进的SBAS技术提取得到不同时期的沧州市的沉降结果。结合地下水地层岩性的特点,利用研究区域内不同监测井得到的地下水位变化的数据,估计出该区域的含水层的弹性释水系数和非弹性释水系数。实验结果表明该区域已经发生较为严重的塑性形变,需要进行人为有效地控制地下水的开采量；最后就地下水位沉降与地面沉降的关系作了分析,得到该实验区域沉降的原因。更多还原

【Abstract】 Differential Synthetic Aperture Radar Interferometry(D-InSAR) had been applied to survey large range, all-weather, real-time surface slow deformation information. It has increasingly become an important technical means in monitoring ground deformation, ground fissure, seismic geological disaster. What’s more, it also greatly makes up for the traditional monitoring means such as GPS, leveling and other deficiencies. In recent years, Based on D-InSAR technology, the key techniques of differential synthetic aperture radar interferometry advanced techniques were deeply analyzed and studied in the dissertation such as PS-InSAR technology, small baseline subsets of D-InSAR Technology (SBAS), point differential interferometry (IPTA), which had been successfully applied in variation field,for instance, monitoring surface subsidence of city, crustal and earthquake deformation monitoring, landslide monitoring as well as the glacier variation field. However, with the application of these techniques to further expand, their arithmetic needs to be perfected in high-precision surveying.Based on key technology of D-InSAR, SBAS technology and IPTA technology were deeply analyzed and studied in the dissertation. In order to reduce the cost of data and improve deformation monitoring accuracy and efficiency, this paper presents the application of SBAS technology; In order to reduce the loss of coherence effect, PS-InSAR technology was proposed to achieve the PS point on the deformation monitoring; In order to improve the image processing efficiency, the vector data was used as an efficient data structures, which is namely IPTA data processing model; Based on the above, an integral terrain displacement detection method is proposed which not only can detect linear deformation, but also can detect periodicity and nonlinear deformation. In the study, Improved SBAS is not the two algorithms superimposed as SBAS and IPTA, but the algorithm improved. Interferometric phase noise reduction based on Dual Tree Complex Wavelet Transform is proposed; ASAR atmospheric delay phase correction is solved by using MERIS data. On the basis of theoretical research, the choice of Hebei District of Cangzhou was selected as the studied and experimental area. ENVISAT ASAR data were employed in the data processing. In order to verify the reliability of results, underground water level data in the monitoring wells and deformation monitoring result are used to compare. Deformation data and underground water level data are used to analysis of the regional groundwater geophysical characteristics. The specific research work and contributions are listed from four aspects as follows:1) Because interferometric phase noise seriously affects the InSAR monitoring accuracy, Compare much variety filtering methods, two dual-tree complex wavelet filtering method was proposed, which is conducted in using SAR simulation data and real interferogram data. Comparing with the precision with other methods, two dual-tree complex wavelet filtering method can greatly improve the filtering precision.2) Because of atmospheric delay in InSAR interferogram, InSAR altimetry and deformation precision are affected. On this basis, combined with the InSAR image on the atmospheric delay characteristics, existing methods of atmospheric interpolation model were studied. Von_Kriging spatial interpolation method Based on the terrain model is put forward in MERIS data. After search for the optimal parameters, ordinary Kriging method and Von_Kriging method interpolation are compared at accuracy. Interferogram with the same time of MERIS data is used to study. The result shows Cangzhou region affected by atmospheric effects obviously, which must be removed for precise survey.3) Based on the SBAS technology and IPTA technology characteristics, a new algorithm is put forward. Cangzhou city surface deformation was extracted by the algorithm, which include of the effective separation of linear and nonlinear deformation. Atmospheric delay and noise are obtained by spatio-temporal filtering method. Finally, the deformation results is contrasted the groundwater with abundant water period and low water period.4) Subsidence result is extracted by improved SBAS technology in Cangzhou city with different periods. The change of underground water level data combined with subsidence results was used to estimate the regional aquifer elastic storativity and non-elastic storativity in different monitoring wells in the study area. The experimental results revealed that the area in some places has already occurred more severe plastic deformation, which need to effectively control the exploitation quantity of groundwater. Finally the underground water level subsidence and subsidence were analyzed to find stratum subsidence mechanism.更多还原