Monitoring Ground Deformation With Small Data Set Differential InSAR

High spatial resolution and large area of surface deformation information in time series can be rapidly and accurately obtained by using radar differential interferometry (D-InSAR) technology. However, the coherent of SAR interferogram reduces due to the surface scattering change over time and the long spatial baseline of SAR platform. In addition, the atmospheric delayed phase cannot be fully eliminated due to the uneven atmospheric fluctuations of different observation time, which increases the atmospheric disturbance phase in the deformation measurement and then reduces the measurement accuracy of D-InSAR. In order to solve these problems, the current methods of temporal D-InSAR generally need to input a large number of SAR data. However, the number of SAR data is always limited for a certain area, which also limits the application of the existing low-coherent processing technology.A new method is proposed to resolve the problem of limited data set. Firstly, high-coherent targets are detected by using sub-aperture decomposition and coherent detection methods from a small data set. Secondly, ICA signal decomposition is adopted to separate atmospheric phase from the measured phase to reduce measurement errors caused by atmospheric disturbances. Thus, by combining these methods with small baseline differential interferometry, a new D-InSAR procedure is developed for a small data set. Finally, the spatial pattern and temporal trend of the extracted surface deformation is analyzed via the spatial-temporal statistical method. And this new procedure is applied to monitor land subsidence in a normal regional area and an underground coal fire deformation area. The results show that when applied to the slow deformation of large area, this method can effectively solve the small data set problem in low coherence of deformation monitoring. Besides, introduced D-InSAR spatio-temporal analysis method provides another useful tool for surface deformation analysis.The main innovations of this thesis are:(1) the use of coherent coefficients and sub-aperture decomposition to detect reliable high coherent points from small data sets; (2) the use of ICA signal decomposition and gradual approach means separating atmospheric phase from InSAR phase; (3) the space-time surface deformation information extraction by InSAR using geostatistics analysis.