【作者】 陈强；

【导师】 李永树； 刘国祥；

【作者基本信息】 西南交通大学 ， 大地测量学与测量工程， 2006， 博士

【摘要】 星载合成孔径雷达干涉是正在发展中的极具潜力的遥感新技术，其差分干涉模式可应用于大范围区域地表形变探测，具有高精度和高空间分辨率等突出优势，可与具有高时间分辨率基于点特征的观测技术(如GPS、精密水准)形成优势互补，为地球物理研究和形变灾害监测提供一种经济的空间对地观测新途径。然而，目前它的应用却受到两大因素的严重制约，即时间、空间失相关引起的低相位信噪比和大气相位延迟影响。 近年发展起来的基于永久散射体(PS，具有稳定散射特性的自然或人工地物)的时序差分雷达干涉技术为间接克服相位失相关和大气延迟提供了可能。本文以永久散射体雷达差分干涉技术为研究对象，探索克服差分干涉应用中失相关和大气影响的关键理论和方法，旨在改善雷达干涉技术在区域形变测量中的精度和可靠性，为形变灾害监测提供经济而有效的遥感新途径。具体来说，本论文完成了如下研究内容并得到了相关结论。 第一，论文从理论和实际两方面深入分析了粗／精轨道状态矢量的不确定性对参考相位、形变提取及DEM重建的精度影响。结果表明DEOS精轨数据能够满足高精度干涉测量，而ESA粗轨数据精度约为1m左右，对形变探测和DEM重建会产生明显的线形系统偏差。 第二，基于时间基线、有效空间基线和Doppler质心频率差的最佳组合原则，本论文提出时序差分干涉公共主影像的优化选取算法，并给出具体的综合相关(JC)函数模型。实验采用不同的指数因子计算综合相关系数，结果表明指数因子的取值差异对综合相关系数不会产生显著摆动，JC模型及其求解方法用于PS-DInSAR公共主影像的优化选取是稳定和可靠的。 第三，为改进PS识别的准确与合理性，本论文提出并实现了双重阈值PS自动探测算法，即时序相关系数和振幅离差指数两道阈值串行结合。通过对上海更多还原

【Abstract】 As a promising remote sensing technique, satellite synthetic aperture radar interferometry (InSAR) has been evolving remarkably in recent years. Its extended mode, i.e., differential InSAR (DInSAR), can be applied to detect large-area ground deformations with some prominent advantages like high accuracy and fine spatial resolution, which greatly complements some conventional ground- and point-based geodetic techniques such as GPS and leveling. This provides a viable and valuable space-geodetic approach for ground deformation detection and geophysical studies. However, the full operational capability of DInSAR has not been achieved until now due to two major negative factors. First, the spatio-temporal decorrelation may result in unacceptable noise in phase observations, particularly over slowly-deformed areas with dense vegetation coverage. Second, the atmospheric artifacts make deformation measurements contaminated.It is possible to mitigate the aforementioned negative impacts and to study the temporal behavior of deformations by tracking some natural and man-made objects, i.e., permanent scatterers (PS), with temporally coherent radar reflectivity, in the frame of time series of SAR images covering the same area (the technique is often termed PS DInSAR). Therefore, this thesis selects PS-based DInSAR as a research topic, and explores its key theoretical and practical issues to overcome decorrelation and atmospheric effects. The motivation of this research is to improve both accuracy and reliability in PS-based DInSAR and to offer a new approach for monitoring deformation-related hazards. The investigations performed in this thesis and relevant conclusions are outlined as follows.First, from both theoretical and practical viewpoint the thesis analyzes the influence of uncertainty in orbital data on accuracy of reference-surface phase, deformation measurement and reconstruction of digital elevation model (DEM). The testing results indicate the precise orbit state vectors provided by DEOS can meet the accuracy requirement of radar interferometry, while the coarse orbit state vectors更多还原