2011年M_W9.0东日本大地震动力学破裂过程的数值模拟

力学上,地震可以看作在应力场作用下由于断层带介质的突然损伤或软化导致的断层带失稳事件.本文基于这个地震动力学模型,利用一种可以模拟断层大位错的有限元方法,研究了2011年MW9.0东日本大地震(Tohoku-Oki)的动力学破裂过程.比较了无障碍体和具有不同刚度障碍体的断层带模型产生的断层位移、位错和应力降.主要结果表明,障碍体的存在并不明显地改变障碍体区域的初始构造应力场.对有障碍体情形,准静态结果显示断层上盘最大逆冲位移和最大剪切位错分别为51m和58m,均发生在海底表面海沟处,与无障碍体的结果(最大剪切位错约55m)相比差别不大;下盘最大倾向位移(-10m)并不与上盘最大值出现在同一位置,而是在障碍体处.障碍体处剪应力降(约11 MPa)大于周围非障碍体区域.障碍体处正应力降的最大值约为3 MPa.模拟结果似乎不支持海山是导致本次地震异乎寻常大位错的原因,而倾向于断层带剪切刚度在地震过程中极度损伤或软化.

Numerical simulation of the dynamic rupture process of the 2011 Tohoku-Oki,Japan MW9.0 earthquake

Earthquakes can be considered as an unstable dynamic process of stress release induced by sudden softening or damage of material on faults. Based on this idea, a finite element method is proposed to simulate the dynamic rupture process of the 2011 Tohoku-Oki, Japan M_W9.0 earthquake to find a mechanical explanation for the extremely huge fault slip.According to the results from JFAST (Japan Trench Fast Drilling Project), the ruptured area of this earthquake is modeled as a 5 m-thick layer, which is 250 km along strike and 130 km along dip direction. An asperity of 70 km by 31 km with different material properties is placed at the middle of the ruptured area. A 3-D linear elastic joint element is used to model the fault zone and corresponding equations of the finite element method are derived. The method consists of two parts: (1) The tectonic initial stress field is induced by boundary displacement. (2) Under the initial stress, the rupture process is modeled by gradually reducing the shear modulus of the ruptured area (rupture propagates as expanding circles with a rupture velocity of 3 km·s^-1). The amount of reduction of shear modulus in the fault zone is constrained by matching maximum slip from the geodesy observation or inversion results of previous studies.The simulated tectonic stress field shows that shear stress increases with depth. Its minimum value is about 9 MPa at the trench and maximum value is about 15 MPa at depth. The existence of an asperity does not change the initial tectonic stress in the ruptured area remarkably. The displacements, slips and stresses predicted by the models with and without asperities are discussed. If an asperity exists, to match the fault displacements observed at the sea floor, the shear modulus of the asperity and that of non-asperity area have to be reduced to 0.5 MPa and 0.1 MPa, respectively. The maximum quasi-static shear displacement on the hanging wall is 51 m at the trench surface, and that on the footwall is 10 m at the asperity. The maximum fault slips with and without the asperity are 58 m and 55 m, respectively. Both of them appear at the trench surface and their difference is not obvious. At the trench, the hanging wall moves about 50 m horizontally towards the trench and uplifts 12 m, which is consistent with the result of repeated multibeam bathymetric surveys. Shear stress drop at the asperity (about 11 MPa) is larger than that in its surrounding area. Maximum normal stress drop (about 3 MPa) appears at the asperity.Based on the initial tectonic stress simulation, given that the Pacific plate moves towards Japan islands at a rate of 80~85 mm·a^-1 and that tectonic shear stress is totally released during this earthquake, the recurrence time of M_W9.0 in this area is estimated to be about 2900 years. The predicted results of this paper do not support the inference that the rupture of an asperity is responsible for the unusually large slip of the Tohoku-Oki M_W9.0 earthquake, instead the huge damage of fault zone material might be the reason.