| Rapid development of geodetic observation techniques, greatly enriched the crustal deformation observation data, and provided a new perspective for the study of earthquake fault movement and other geophysical issues. To hypocenter parameters inversion, not only an important means of fault movement characteristics and analysising focal mechanisms, but also it provide the basis for analysis of active fault rupture propagation, post-seismic deformation, change of the lithosphere stress and hazard assessment after the earthquake.Therefore, obtaining the earthquake source parameters accurately is significant for studying the mechanism of earthquakes. To determine more precise hypocerter parameters, in this dissertation, the source dislocation model parameters inversion method carried out research.(1)This article first study the total least squares problem with inequality constrained. Analysising the defect of existing method about that problem, quasi newton correction algorithm for Weighted Total Least Squares problem with Inequality Contraints(ICWTLS) is proposed based on ICEIV model that existing. The result show that the new algorithm has good convergence, and avoiding calculation for the Hesse matrix of the objective function, greatly improving the efficiency of the algorithm. Simulation a single fault slip parameters with inequality contraints for study the applicability for source parameter inversion to proposed method.(2)This article then study the the nonlinear algorithm for hypocenter parameters inversion, the MPSO(Multiple peak Particle Swarm Optimization, MPSO)algorithm was used to invert the hypocenter parameter. the improvement and iterative process of MPSO algorithm is Detailed analysis, further, the algorithm is compared to tradit-ional genetic algorithm. The study result for simulation and actual earthquake cases show that the MPSO algorithm has the advantages of less control parameters, fast convergence,stable and reliable. It is means that the algorithm is ideally suited for highly nonlinear source parameters inversion problem.(3)The slip distribution of the source inversion is then studied in this paper, taking into account the error impact of the matrix Green’s function, the least squares and total least squares method are used for slip distribution inversion, the measurement vector’weight is determined according to the observation data precision, and the weight matrix of coefficient matrix elements is determined cccording to the linear relationship that between the coefficient matrix elements and observation data. Then different slip distribution models and actual examples is studied for slip inversion, the research shows that there are some less differences(less than 10-5) between total least squares solution with least squares solution.(4)This article then studies the April 20, 2013 Sichuan Lushan Ms7.0 earthquake and April 6, 2009 L’Aquila(Italy) Mw6.3 earthquake extensively, the nonlinear hypocenter parameters and slip distribution of two earthquake is inverted respectively. For Lushan earthquake, obtained through the GPS co-seismic three-dimensional defor-mation data around the epicenter, indicates that earthquake is a thrust fault with a strike of 206.47 o, a dip of 44.11 o, the focal slip mainly occurs in the depth range 4km-20 km, the average slip is 0.17 m, the maximum slip of 0.72 m is located in the depth range13km-14 km, and the inverted geodetic moment is 1.19 ? 1019 N.m(Mw6.68). For L’Aquila earthquake, seismic deformation field is obtained by Envisat satellite SAR data, the result indicates that a normal movement with a strike of 144.37 o, a dip of 59.06 o, he focal slip mainly occurs in the depth range 4km-15 km, the average slip is 0.22 m, the maximum slip of 0.95 m is located in the depth range6km-8km, and the inverted geodetic moment is 3.63 ? 1018 N.m(Mw6.34). Which are consistent with the study result of seismology. |
PDF Full Text Request