【作者】 向荣；

【导师】 丁晓利； 李志伟；

【作者基本信息】 中南大学 ， 地图学与地理信息系统， 2007， 硕士

【摘要】 差分合成孔径雷达干涉测量技术(Differential Interferometry SyntheticAperture Radar,D-InSAR)在最近10来年有了很大的发展。但是,由于时间和空间的失相干和大气效应的影响,传统的差分干涉测量的应用受到很大的限制。永久反射器技术(permanent scatterer,PS)虽然能克服这些问题,但也有缺点,它需要有一定密度的稳定的反射点,对影像数目的要求也很高(大于30张)。为了克服以上问题,基于角反射器的合成孔径雷达干涉测量(Corner Reflector InSAR,CRInSAR)技术被提出来了。它在低相干地区的地表形变监测中具有很大应用潜力。本文首先介绍了InSAR和D-InSAR的工作原理、数据处理流程和数据处理中的若干关键技术,介绍了目前国际上已有的CRInSAR算法理论,并特别介绍了由德国GFZ的夏耶博士提出的CRInSAR算法理论和处理流程。然后,本文分析了CRInSAR中的大气影响。对大气的结构和成分进行了分析,根据InSAR中大气的延迟情况,将大气影响分为对流层延迟和电离层延迟。由于电离层对CRInSAR的影响很小,大气延迟的主要部分在对流层,因此本文系统分析了对流层对CRInSAR的影响,对比了现在主要的大气校正模型,最后给出了用Continuous GPS消除大气影响的算法和模型。接着本文详细针对CRInSAR算法中相位解缠问题,提出了利用GPS整周模糊度解算中的LAMBDA方法进行相位整周模糊度搜索,并通过试验证明通过增加虚拟观测值,考虑研究区域的变形先验信息,LAMBDA方法能够成功的应用到CRInSAR技术的中来,并且结果比较符合实际的观测结果。然后在夏耶博士的CRInSAR算法的基础上改进提出了CRInSAR变形解算模型,在假设形变包括线性形变和周期性形变的基础上进行变形模型的解算。最后,本文详细介绍了在香港地区的试验验证CRInSAR变形算法和大气改正的过程和结果。先用香港地区12个CGPS站的数据对8幅SAR图像做了大气改正,在此基础上分别用传统DInSAR和使用LAMBDA方法的CRInSAR计算了试验区域的变形状况。在两个试验区,DInSAR和CRInSAR得出了非常相似的结果,证明了CRInSAR变形监测算法的可靠性。更多还原

【Abstract】 As well known,temporal and geometric decorrelation represents the most important limiting factors in synthetic aperture radar interferometry(InSAR) Due to this fact,analyses of low coherence areas(e.g.,forested and vegetated areas)with conventional InSAR methodologies cannot be performed successfully.Even the application of advanced InSAR algorithms,such as the permanent scatterer(PS)technique,may prove to be ineffective wherever the density of stable radar targets(i.e.,the PS)is extremely low.In order to overcome this limitation,the usage of artificial reflectors can be very promising.In this paper,at first we give the basic theory and algorithm-processing flow chart of InSAR and DInSAR.Then,the deformation monitoring theories of CRInSAR were introduced,especially the algorithm of CRInSAR presented by Xia Y.of Germany GFZ.Second,we analyzed structure and composition of the atmosphere.Based on the differences of effects,the layered structure of the atmosphere is redefined as tropospheric and ionospheric effects.Because the ionospheric effects are very small in CRInSAR and the majority of atmosphere effects are from tropospheric,we analyzed the tropospheric effects to CRInSAR and compare atmospheric correction models.At last,the method and process of using CGPS data to correct atmospheric effects are analyzed.In the third part,we applied the Least Squares Ambiguity Decorrelation Adjustment(LAMBDA)method which has been applied in GPS ambiguities resolution to solve the problem of phase unwrapping,in order to solve the phase unwrapping problem of CR-InSAR.Real data is used to test the unwrapping method and reliable results are achieved compared with other method.Then the new CRInSAR algorithm was proposed.Based on the assumption that the deformation includes the linear and the periodic parts,we modeled the deformation and solved it.At last,we investigate and compare the process and results achieved in three test fields in Hong Kong in order to demonstrate some algorisms in CRInSAR and to estimate the reliability and precision of the method in slop movement and ground deformation monitoring.We used the data from CGPS observations to correct the interferometric phases for the temporal atmospheric inhomogeneity first.Then we used the standard DInSAR and the Corner Reflectors technique(CR).The results of deformation of the CR point extracted from the DInSAR are similar with solutions from CRInSAR technique.The comparison of the results demonstrates that the Algorithms of CRInSAR are reliable and precise.更多还原