Seismic Effect Of Finite Fault Thickness And Asymmetric Moment Tensor

In recent decades, the seismic moment tensor representation and inversion was a great success. In source physics research, the moment tensor as a second-order symmetric tensor has been widely accepted and successfully used, and the asymmetric moment tensor is rarely mentioned. This paper emphasizes the rationality and importance of asymmetric moment tensor in. earthquake source theory and argues that the seismic effect of fault thickness should be represented by asymmetric moment tensors.In Chapter2, the representation theorem and phenomenological description are used to derive the asymmetric moment tensor representation of fault with finite thickness, which shows the symmetry is not necessary. The antisymmetric part in asymmetric moment tensor result in that the single couples along the slip directions on fault and auxiliary planes are no longer equal, and could be used to resolve the fault plane ambiguity. According to the fault model of finite thickness, larger single couple is on fault plane, and smaller single couple is on auxiliary plane.The theory, method and application in earthquake emergency response of symmetric moment tensor inversion in time domain are introduced in Chapter3. All these work are important foundations of asymmetric moment tensor inversion.The theory and method of asymmetric moment tensor inversion are similar to symmetric moment tensor inversion. The asymmetric moment tensor has3more scalars than symmetric moment tensor. Only a few modifications are required to realize the asymmetric moment tensor inversion. The moment tensor’s vector representation could be used to analysis the asymmetric moment tensor and quantitatively describe the difference between moment tensors.By analyzing the relation between Green’s Function and every moment tensor’s component, the author discover that, in asymmetric moment tensor inversion, using only vertical data can not distinguish Mxy and Mvx. The vertical and horizontal data should be used in joint inversion to distinguish Mxv, and Myx. This is the biggest difference between asymmetric moment tensor inversion and symmetric moment tensor inversion. If a different velocity structure model or a different method of calculating the Green’s function were used, the resolutions between any moment tensor components need to be revisited.The synthetic test proves that the asymmetric moment tensor inversion method is feasible. The synthetic test indicates that it is necessary to use P and S waves to joint invert asymmetric moment tensor. The joint inversion of P and S waves can improve the result.The asymmetric moment tensor inversion method is applied to13earthquakes between2008and2010. Their fault planes are known or could be deduced. Inversion results show that the symmetric parts in moment tensor solutions are reliable. The AIC shows that there is no over fitting in asymmetric moment tensor inversion. Considering the fault model of finite thickness and asymmetric moment tensor, the fault planes in6of the13earthquakes are well resolved. The values of A calculated from asymmetric moment tensors indicate, in asymmetric moment tensor solutions, the symmetric part is predominant, the antisymmetric part can provide additional information to resolve the fault plane ambiguity. However, the resolution is low and mainly controlled by the inversion errors.