论文部分内容阅读
利用多基线距DInSAR技术反演城市地表缓慢形变已经成为当前DInSAR研究的热点.该技术的基本流程包括干涉图像对组合,差分干涉图生成,高相干点提取,Delaunay三角网建立,网络增量计算和集成,残余相位解缠和定标,大气相位和非线性形变相位估计等.其中网络增量计算是线性形变反演的关键,残余相位解缠和定标是非线性形变反演的关键.针对前人在这两方面研究的不足,提出基于改进的模型相关系数方程的增量计算方法,以及基于三角形“环”处理方法进行离散点相位解缠和定标的算法.在上述算法的基础上,利用24景ERS数据,反演了苏州市在1993-02~2000-12之间的地面沉降.结果表明,在此期间内苏州市大部分城区平均沉降速率在20~30mm/a之间,而且沉降速率呈逐年递减趋势,并在2000年之后趋于稳定.水准数据证明了结果的准确性和算法的可行性.
In this paper, DInSAR technique is used to retrieve the slow deformation of the urban surface, which has become a hot topic in current DInSAR research.The basic flow of this technique includes the combination of interferometric images, the generation of differential interferograms, the extraction of high coherence points, the establishment of Delaunay triangulation, And integration, phase unwrapping and scaling of residual phase, phase estimation of atmospheric phase and nonlinear deformation, etc. Network incremental computation is the key to linear deformation inversion. Residual phase unwrapping and scaling are the keys to nonlinear deformation inversion. Aiming at the deficiencies of predecessors in these two aspects, this paper proposes an incremental calculation method based on the improved model of correlation coefficient equation and an algorithm of phase unwrapping and scaling discrete points based on the triangle loop method. , The land subsidence of Suzhou City was retrieved using the ERS data of 24 scenes from 1993-02 to 2000-12.The results show that the average settlement rate of most urban areas in Suzhou during this period is between 20-30 mm / a , And the sedimentation rate is decreasing year by year and tends to be stable after 2000. The level data prove the accuracy of the result and the feasibility of the algorithm.