Study On Time-series InSAR Method For Monitoring Surface Subsidence With Persistent And Distributed Scatterers

Surface subsidence is a common geological disaster,which has a great impact on people’s normal life and the sustainable development of society and economy.Therefore,it is significant to conduct the research on time-series surface subsidence monitoring.Time series InSAR technology has a wide application in subsidence monitoring,but the number of measurement points selected by current methods is limited.Therefore,this thesis focuses on the subject of surface subsidence monitoring combined with PS and DS,researches on the method of homogeneous pixels identification,the selection of PS and DS,and phase optimization.The main research work and results are as follows:(1)Aiming at the problem of the identification accuracy of statistically homogeneous pixel(SHP),on the basis of the hypothesis test of confidence interval(HTCI)algorithm,a SHP recognition algorithm based on the optimal confidence interval estimation(OCIE)is put forward.The proposed algorithm shortens the length of the confidence interval and determines the optimal confidence interval according to the extreme values of the lower and upper confidence limits.The experimental results of simulated and real data show that the OCIE algorithm effectively reduces the identification error of SHP and provides a more meaningful SHP set.(2)Aiming at the problem of the performance of DS phase optimization,an improved EVD-PT algorithm(IEVD-PT)is presented on the basis of the coherencematrix eigenvalue-decomposition PT(EVD-PT)algorithm.The proposed algorithm analyzes multiple components to avoid losing more information,and the goodness-offit is calculated by using those interferometric phases with less noise and clear fringes.The two algorithms before and after improvement are used to optimize the DS phase respectively.The results show that the interferogram quality optimized by IEVD-PT algorithm is improved,and residue point number and sum of phase difference are relatively small,which verifies the effectiveness of IEVD-PT algorithm.(3)In view of the problem that how to select as many DS points as possible from the available data sets without losing the measurement accuracy,a DS encryption method is proposed based on hierarchical analysis.According to the goodness-of-fit value,DS candidate pixels are divided into three groups and processed by hierarchical analysis strategy.The phase quality of pixels with moderate temporal coherence is determined by Pearson correlation coefficient.The subsidence information of Tianjin area is obtained by using the two methods before and after improvement.The results show that the DS point density obtained by hierarchical analysis strategy is increased by 60%.The correlation coefficient and the root mean square error of the subsidence rates of completely coincident points obtained by the two methods is 0.948 and 1.1mm/year,respectively.Compared with the monitoring results of leveling,it is found that the standard deviation of the subsidence results of the two methods are 1.7 mm and 2.3 mm respectively,and the measurement accuracy is relatively high.It is verified that the hierarchical analysis strategy is an effective way to encrypt DS pixels.(4)In light of the problem that how to select more reliable PS and DS,the intensity dispersion index is used to select PS,and DS extraction and phase optimization are conducted by the IEVD-PT algorithm.,a combined PS and DS InSAR(CPDS-InSAR)method is developed based on the subsidence model of Multiple-master Coherent Targets Small Baselines InSAR(MCTSB-InSAR).MCTSB-InSAR and CPDS-InSAR are used to obtain the subsidence information of Beijing area.The results show that the CPDS-InSAR method significantly increases the number of measurement points.and the point density is about 2.1 times that of MCTSB-InSAR.The correlation coefficient and the root mean square error of the subsidence rates of completely coincident points obtained by the two methods is 0.962 and 3.1 mm/year respectively,which indicates the monitoring results of the two methods are relatively close.It is verified that CPDS-InSAR is an effective method to improve the spatial density of subsidence measurement.There are 72 figures,13 tables and 179 references in this paper.