| With the development and application of Global Positioning System (GPS), aswell as the gradual increase in accuracy of GPS observations, the application field ofGPS becomes more and more wide. In particular, the applications in the geosciencesare also rapid development. Especially, Maruyama (1964) gave Green’s function inthe semi-infinite homogeneous space. Okada (1985) deduced the relationship betweenthe dislocations of the elastic semi-infinite space within the rectangular fault andsurface deformation, which made rapid development of the GPS data used forinversion underground fault structure. It is a key task of inversion to research on theco-seismic or a-seismic sliding using surface deformation data, in order to understandthe earthquake preparation process and the earthquake preparation mode, and even topredict the occurrence of earthquake. The result of this study is meaningful forreducing seismic disasters.we studied the characteristics of the inversion method of Yabuki&Masu’ura,discussed Maruyama Methods for calculating the Green’s function on the problem ofmesh size, analyzed the impact of the fault geometry parameter deviations in theinverse calculation using this inverse method, explored the ability of data density onthe spatial resolution in the inversion calculation, inversed the Tohoku-Oki earthquakeand the WenChuan earthquake by deformation data from GPS Observations, andfinally carried out some discussion in the thesis.The main research contents in the thesis include the following sections:1. Chapter1discussed the convergence of the Green’s function of digitalcomputing. When we calculate the Green function, we apply the partial derivative ofGreen function to integral during a small area using the approximation method ofLagrange mean value, which require the area is small enough. When we divide theblock size is less than1km, the relative error in calculation on the surface can beachieved within. In the simulation of the inversion process, we have given faultgeometry parameters along the Strike and Dip direction, which is divided into24×12 blocks, when we calculate the displacement of each fault face on the surface, thedivide each block into40×40small patches, so that the relative error brought aboutby the zoning size can be controlled at less than.2. Chapter2verified the feasibility of the Yabuki&Matsu’ura inverse methodthrough Simulation. First through a given plane fault model and slip distributionmodel we calculate the theoretical displacement on the surface observation points.Then, considering the actual observation data errors, we add a noise data whichfollow a normal distribution random error to the displacement. Finally we use themeasurements with a random error to inverse the parameters of the fault slipdistribution. Although the data contain random errors, we can get the slidingdistribution using Yabuki&Matsu’ura (1992) inversion method.3. Chapter3discussed the deviation of the fault geometry parameters on theinversion results by simulation. We analyzed the effect of the deviation from the strikeand dip on the inversion results, and found that these two parameters deviation andMisfit values are nearly linear relationship, which requires that we should reduce thedeviation of the strike and dip as much as possible, when we use Yabuki&Matsu’ura(1992) inversion method to calculate. We can get the true sliding distribution whenthe deviation is less than2°. Meanwhile, we also discussed the effect of the differentdensity of the data on the Inversion. Though above researches we found that thegreater density of the uniform distribution deformation data observed on surface, themore accurately reflect the actual characteristics of the slip distribution.4. In Chapter4we inversed the co-seismic slip distribution of the Tohoku-Okiearthquake using Yabuki&Matsu’ura (1992) inversion method. First, we determine afault model by B-spline interpolation method, which is very close to the real model.The aim of above work is to minimize the uncertainty of the results brought by thefault model error. In the inversion process, we first constructed fault model based onseismic tomography of northeastern Japan. Through simulation, we discussed whetherusing co-seismic deformation data observed on the actual GPS observation points toinverse and get the true slip distribution. Then, we inverse the fault slip distributionusing the actual observed co-seismic deformation data. The results showed that the maximum slips on the fault plane are about35m. The large slips distributed on thefault plane shallower than30km above the epicenter. The maximum rupture tookplace in the20km place. The seismic moment is about. Thecorresponding magnitude is9.0. Based on the results of the inverted slip distribution,we compute the far-field deformations in Northeast China of the2011Tohoku-OkiMw9.0earthquake using spherical dislocation theory of Sun et al.(2009). The resultsshow that the great impact of the earthquake in Northeast China, Jilin and Liaoning,which will change the original strain distribution in the region.5. In Chapter5we inversed the co-seismic slip distribution of the2008Wenchuanearthquake co-seismic slip distribution using Yabuki&Matsu’ura (1992) inversionmethod. Because that the earthquake rupture structure is very complex, we only easilychoose four plane faults to simulate the structure of fault rupture, and finally inversethe slip distribution on the plane faults. The inversion results show the slips caused bythe great earthquake are characterized mainly by thrust motion near Yingxiu, the southsegment of the Yingxiu-Beichuan fault. The dextral components are apparently largerthan the thrust slips at the north segment of the fault. Our estimations agree well withthe results of geologic survey and the characteristic of rupturing. The estimated slipshave two high-slip concentrations up to9.6m and9.3m, located respectively near theYingxiu town and Beichuan city, which suffered the greatest fatalities and structuredamages during the earthquake. The total moment released by the great earthquake isabout, corresponding to an earthquake with magnitude of Mw7.9.Finally, based on the2008Wenchuan earthquake co-seismic deformation data, weanalysis different data distribution how they are effect on the inversion result, andthen evaluate the reliability of thrust fault rupture of the inversion results，such as the2011Tohoku-Oki earthquake. |
PDF Full Text Request