Multi-means InSAR Research And Its Application In Coseismic And Interseismic Deformation Monitoring

China is the location of some of the most destructive earthquakes, such as the 2008 Wenchuan earthquake, the 2010 Yushu earthquake and the 2013 Lushan earthquake, which caused huge damages to local people. Big earthquakes are mainly tectonic stress loading and releasing processes, which usually cause detectable surface deformation around epicentral area. In this thesis we use Interferometry SAR (InSAR) and its related techniques such as PS-InSAR/CR-InSAR to ananlyze coeseismic and interseismic deformation. We choose two recent Qaidam earthquakes happened in 2008 and 2009, respectively to do the InSAR coseismic application. As for the interseismic deformation study, PS-InSAR and CR-InSAR are used separately and jointly, and we choose Haiyuan fault, Tianjingshan fault and west Qinlin fault since our corner reflectors are installed within this area. The main content of this thesis is as follows:1.We first determine the coseismic deformation of the 2008 and 2009 Qaidam earthquakes, based on which we invert fault slip distribution of these events. Coulomb stress changes are then calculated and possible triggering mechanisms between these earthquakes are analyzed. There are totally four event, one happened in 2008 and three in 2009. Using Envisat descending AS AR data, the coseismic deformation is obtained. Our results show that the 2008 event interferogram has a single deformation pattern whilst the 2009 event interferogram have three. The peak values in these two interferograms are 0.097m and 0.4m, respectively. Using one planar segment fault model we invert the InSAR data and determine the largest slip at 19km depth. For the 2009 interferogram, three events are unseparated. We use one segment, two segments and three segments to test the optimal fault model and find that three segments fault model obtain the smallest residuals. We also calculate the coulomb stress change in duced by the 2008 event and analyze possible triggering mechanism between these events. Our results show that the 2009 events are clearly triggered by the 2008 event, as the 2008 event happened in a deeper part of the source fault, which impart large coulomb stress changes on the shallow part of the same source fault that hosted the 2009 events.2. We analyze the specula difference of the Corner reflectors between low-resolution and high-resolution SAR images. We first compare the magnitude of Corner reflectors. The magnitude on TerraSAR images (high resolution) shows a "+" symbol feature and a lightened spot in the center, which is characterized by the triang ular shape of the Corner reflectors. In the mean while the same corner reflector show a diffused lightened block with over 4-6 pixels range on Envisat ASAR images (low resolution). We also analyze the phase information and they show similar differences between low-resolution and high-resolution SAR images. On TerraSAR images the corner reflector shows a "+" symbol and lightened spot in the center.3. Using CR-InSAR and PS-InSAR separately and jointly, we obtained the intersesimc deformation of the Haiyuan fault. Using PS-InSAR, accompanied with standard deviation method we determine a 2-3mm/yr of left lateral strike slip rate for the Haiyuan fault. Using CR-InSAR we determine a 2.1mm/yr left lateral strike slip rate for the same fault. Our results show consistency between these two methods. In addition using PS-InSAR, we also determine a sinistral strike slip rate of 3.9mm/yr for the Tianjingshan segment, but no corner reflectors available for CR-InSAR application.4.We propose an improved method to construct deformation monitoring network using both CR and PS pixels. With CR-InSAR and PS-InSAR, such an improved deformation monitoring network utilizes CRs as a high-precision controlling points, PS as secondary level of controlling points, and combine CRs and PS to construct a uniformed network. Using LAMBDA unwrapping method, we apply this method to study the middle segment of the western Qinlin fault. Our results show a mean sinistral strike slip rake of 2.3mm/yr and-1.5mm/yr for southern side and northern side of the fault, respectively. The total slip deficit reaches 3.8mm/yr, which shows consistency with other studies, such as GPS measurements and long term geological slip rates. We finally compare the CR and PS features and find that a reliable result could be obtained with the inclusion of high-precision CRs, as CRs are stable, which is crucial for calculation.