Joint Inversion Of Fault Slip Distribution And Kinematic Rupture Process Of Big Earthquakes

This thesis focuses on joint inversion of source rupture process of earthquakes. I take three typical events happened on China mainland, including the 2008 Ms8.0 Wenchuan earthquake (dominantly thrusting), the 2008 Ms7.3 Yutian earthquake (normal faulting), and the 2010 Ms7.1 Yushu earthquake (mainly sinistral strike slip), to conduct an integrated study. These three events have some common features. Firstly, they all occurred at the boundaries of the active Bayan Hara secondary block, which has now been regarded as one of the most active and deformed blocks in eastern Tibetan. Secondly, their magnitudes all exceed 7, providing a window for scientists to look into the earth's crust and to study the tectonic process. For the third but not the least thing, they all happened in the current age with rapid development of data collection technology. To predict the future risks of major earthquakes and understand present-day deformation along the major faults of the Bayankala secondary block, these earthquakes need further studies. In addition, their causative faults represent the three main types existing in nature, which allow us to obtain new insights into the problem how major quakes are generated by faulting.The source study attempts to clarify how a faulting event (an earthquake) initiates, spreads and stops. In order to describe its characteristics in space and time, different types of data are used separately or jointly, such as remote sensing images, GPS, InSAR and teleseismic data. Keeping the non-uniqueness of an inversion problem in mind, I also try to access how robust a solution could be and which part or parts could be well resolved by a specific dataset, or which part or parts could not be resolved by the current dataset(s). The main contents of this thesis are described briefly as follows:1) To obtain the coseismic deformation of the 2008 Wenchuan earthquake by SAR interferometry; analyze intracontinental thrust faulting characteristics and its associated surface ruptures; compare the InSAR measurements with GPS data, improving the data sets compatibility; jointly invert the GPS and InSAR measurements; do a comprehensive resolution test evaluating the uncertainties of the inversion; anlyze the resolving ability of each data; get a better acknowledge of the spatial variation of slip distribution along both strike and dipinp. 2) To obtain the coseismic deformation of the 2008 Yutian earthquake by SAR interferometry; delineate the surface rupture by high precise remote sensing images; simplify the earthquake ruptures into a three-segment fault model; invert the slip distribution based on InSAR measurements and the predetermined fault model; perform analysis of the uncertainties caused by the azimuth angle and the incidence angle in InSAR observations.3) To obtain the main faulting characteristics of the 2010 Yushu earthquake in space and time by joint inversion of teleseismic data and InSAR measurements; as a first effort to determine the coseismic deformation measured by InSAR along LOS; map the surface ruptures by offset tracking of SAR pixels; establish a five-segment fault model to represent surface ruptures, based on which the synthetic data inversion and real data inversion are conducted; evaluate the uncertainties through the synthetic data inversion and to analysis the rupture history of the 2010 Yushu earthquake.Through this thesis, I reach some conclusions as follows.As for the 2008 Ms8.0 Wenchuan earthquake, through our resolution test study, and comparisons with other studies as well, we find some common features of slip distribution among all of the schemes of inversion. Three peak-slip areas along the YBF are inverted, at Yingxiu, Yuejiashan and Beichuan areas, respectively; and the inverted rake shows dominantly thrusting at southwest segments and changes progressively to dominantly dextral striking or even pure dextral striking northeastward.As for the 2008 Ms7.3 Yutian earthquake, our inversion results show that the incidence angle has certain effect on the slip distribution, which means a substitution of location-dependant incidence angle with constant mean incidence angle, will lose some useful slip distribution on fault plane. Our inversion also shows the azimuth angle, however, has almost no obvious effect and it is an optimal choice to use the mean azimuth angle. Our most robust slip distribution can be obtained with the two parameters location-dependant. In our favored inversion scheme using both the two angles location dependant, the maximum slip reaches 3.2m. Slip is mainly confined to 0¹4km depth and no evidence of shallow slip deficit. The inverted seismic moment M0 is 3.3×10¹⁹ N.m, equivalent to an event of magnitude M_w 7.0.As for the 2010 Ms7.1 Yushu earthquake, through by resolution test, we confirm that different data sets will give different fault source models and the most robust results can be expected reasonably by joint inversion. We choose to ignore the inverted slip at depth greater than 20km, due to the limited resolution of the data used. Basically the slip near the surface can be well resolved both with InSAR measurements and teleseismic data. While InSAR data has limited resolution at hypocenter and deeper depth, teleseismic data can resolve the slip near hypocenter very well. And only with surface offset constraint, can the second peak slip at the eastern segments be resolved by teleseismic data inversion. From the STF of joint inversion, the essential rupture time of the Yushu earthquake lasts mainly for ²0 seconds. Within the first 10 seconds, the slip generates a peak-slip asperity at depth of hypocenter, but little slip has reached the surface. Within the second 10 seconds, a large peak slip area is generated at eastern segments of the fault model. During the rupture, two peaks of energy release time are occurred, at 8 and 12 seconds, respectively. In overall, the earthquake is mainly propagating unilaterally to the SE along strike.