【作者】 周云好；

【导师】 陈运泰；

【作者基本信息】 中国地震局地球物理研究所 ， 固体地球物理学， 2002， 博士

【摘要】 本文第一章简要地回顾了地震矩张量概念产生的历史背景，介绍了地震矩张量反演和震源破裂过程反演的研究进展和研究现状。 第二章对用远场体波反演地震矩张量的基本理论和基本方法作了系统地阐述。并用IRIS全球地震台网的远场体波资料反演了1998年3月25日南极M_w8.1（据IRIS）地震的矩张量和2000年6月4日印度尼西亚苏门答腊南部M_s8.0地震的矩张量。 对1998年3月25日南极地震的研究结果表明：这次地震的矩张量的最佳双力偶分量为3.7×10²¹N·m。矩震级为M_w8.3。补偿线性矢量偶极分量为-1.5×10²⁰N·m，约为最佳双力偶分量的-4％。 爆炸分量为1.1×10²⁰N·m，约为最佳双力偶分量的3％。节面Ⅰ：走向为194°，倾角为89°，滑动角为-177°。节面Ⅱ：走向为104°，倾角为87°，滑动角为-1°。P轴：方位角为59°，倾角为3°。T轴：方位角为329°，倾角为2°。B轴：方位角为206°，倾角为87°。这次地震的震源机制基本上是左旋走滑，带有微小的正断层滑动分量。走向为104°的节面Ⅱ是断层面。总体上说，主震是由东向西破裂的。余震发生在主震的断层面上（或其附近），是由主震断层面附近介质的应力调整造成的。 对2000年6月4日苏门答腊南部地震的研究结果表明：这次地震的震源机制主要是左旋走滑，带有很小的逆冲倾滑分量。地震矩张量的最佳双力偶分量为1.5×10²¹N·m。矩震级M_w=8.1。地震矩张量的补偿线性矢量偶极分量为1.2×10²⁰N·m，约为最佳双力偶分量的8％。爆炸分量为-5.9×10¹⁹N·m，约为最佳双力偶分量的4％。节面Ⅰ：走向为199°，倾角为82°，滑动角为5°。节面Ⅱ：走向为109°，倾角为85°，滑动角为172°。P轴：方位角为154°，倾角为2°。T轴：方位角为64°，倾角为10°。B轴：方位角为256°，倾角为80°。P波显示的方向性效应表明：走向为199°的节面Ⅰ是断层面。破裂传播方向为自东北向西南方向，几乎垂直于爪哇海沟走向。左旋走滑断裂的这次地震，可能是由消减板块上不对称分布的障碍体使左边的消减运动受阻造成的。北北东—南南西向分布的余震发生在主震的断层面上（或其附近）；而西北—东南向分布的余震则发生在明打威断层上（或其附近）。本文提出的一个运动学震源模式，较合理地解释了这次地震的地震源机制、余震分布特征和这次地震没有激发海啸的原因。 第三章系统地阐述了从远场体波提取震源时间函数（STF），用提取的STF反演震源破裂时间—空间过程的基本原理和方法。并用这种方法研究2000年6月4日印度尼西亚苏门答腊南部M_s8.0地震的震源时空过程。结果表明：从不同方位台站的Z分向P波和SH波提取的震源时间函数显示了明显的方向性效应。震源时间函数的这种方向性效应与P波的方向性效应相一致，清楚地表明：走向为199°的节面为断层面，地震是从东北向西南方向单侧破裂的。这次地震破裂持续了约16s。破裂面长约95km，宽约60km。平均静态位错约为11m。最大静态位错达27m。平均静态应力降大约为90MPa。最大静态应力降达220MPa。 第四章发展了一种直接拟合远场体波观测地震图反演震源时一空过程的新方法。并通过正、反演数值模拟实验论证了这种方法的可行性。这种方法是位移表示定理的直接应用。根据有限长数据的离散招积运算法则，将全部于断层造成的位移的积分表达式写成对有限长波形数据做离散榴积运算的矩阵表达式，从而可解出每个子断层的地震矩变化率。本方法除加了断层不会发生反向滑动的物理约束条件外，没有对断层破裂点的起始位置、破裂方式、破裂传播速度和各子断层震源时间函数的形状和持续时间等作任何先验的假定。因而反演结果中很少有人为的闲素。这是本方法的特点。正、反演数值模拟实验结果表明：在有4个以上台站的资料（台站的张角不小于90O）、资料的噪声水平不高于 P波最大振幅的 25％的条件下，用这种反演方法，能够可靠地反演出不同破裂方式（单侧破裂、双侧破裂、圆盘破裂、地震矩在断层面上均匀分布和非均匀分布等）震源的时间一主间过程。在一般情况下，这种方法对观测资料要求的上述条件是能够满足的。更多还原

【Abstract】 In this paper,the first chapter,the historical background of how the concept of the seismic moment tensor was proposed and the development of the research work of seismic moment tensor inversion and seismic source process inversion were reviewed briefly.In the second chapter,the basic theory and method of how to use far-field body-waves to invert for the seismic moment tensor were expatiated systematically. The moment tensor solutions of the March 25,1998,Antarctic plate earthquake (Mw=8.1,IRIS) and the June 4,2000,southern Sumatra,Indonesia,earthquake (Ms=8.0) were inverted using teleseismic body waves recorded by long period seismograph stations of the IRIS global seismic network.The moment tensor solution obtained in this research for the March 25,1998,Antarctic plate earthquake shows that:The double-couple (DC) component of this earthquake is 3.7X 1021Nm,the moment magnitude is A/w=8.3,the compensated linear vector dipole (CLVD) component is -1.5 X 1020Nm,about -4% of the DC,and the explosion (EP) component is l.lX1020Nm,about 3% of the DC. NP I:The strike is 194,the dip,89,and the rake,-177;NP II:The strike is 104,the dip,87,and the rake,-1. P axis:The azimuth is 59 and the plunge,8;T axis:The azimuth is 329 and the plunge,2;B axis:The azimuth is 206 and the plunge,87. The focal mechanism is mainly left-lateral strike-slip,with a very small normal fault slip component. The NP II is the fault plane. On the whole,the main shock ruptured from the east to the west. The aftershocks were distributed on (or near) the fault plane and were triggered by the stress redistribution in the medium in the nearby region of the fault plane.The result for the June 4,2000,southern Sumatra earthquake shows that:The focal mechanism is mainly left-lateral strike-slip,with a small thrust component. The double-couple (DC) component of this earthquake is 1.5X 1021Nm,the moment magnitude is A/w=8.1,the compensated linear vector dipole (CLVD) component is 1.2X 1020Nm,about 8% of the DC,and the explosion (EP) component is -5.9 X 1019Nm,about 4% of the DC. NP I:The strike is 199,the dip,82,and the rake,5;NP II:The strike is 109,the dip,85,and the rake,172. P axis:The azimuth is 154 and the plunge,2;T axis:The azimuth is 64 and the plunge,10;B axis:The azimuth is 256 and the plunge,80. The P-waveforms recorded at different stations show prominent directivity. The directivity shows that the NP I is the fault plane,and that the earthquake ruptured unilaterally from the northeast to the southwest,nearly perpendicular to the strike of the Java trench. The strike-slip faulting was caused by asymmetrical distribution of barriers on the subducting plate. The NNE-SSW distributed aftershocks occurred on,or near,the fault plane of the main shock;while the NW-SE distributed aftershocks,on,or near,the Mentawai fault. A kinematic source model proposed in this paper explains well the mechanism of this earthquake,the characteristics of the aftershock distribution,and the reason why no tidal wavesgenerated by this great earthquake.In Chapter 3,the basic theory and method of retrieving source time functions from far-field seismic records to invert for the temporal-spatial source process were expatiated systematically. And the method was used to study the temporal-spatial source process of the June 4,2000,southern Sumatra,Indonesia,MS8.0 earthquake. The result obtained in this study shows that:The source time functions retrieved from P- and S-waves recorded at different stations distributed at different azimuths show prominent directivity. The source time function directivity is in accordance with the P waveform directivity,clearly showing that the nodal plane of strike 199 is the fault plane and that the earthquake ruptured unilaterally from the northeast to the southwest. The source duration is about 16s. The fault area is about 95km in length and 60 km in width. The average static slip is estimated to be about llm. The maximum static (final) slip is 27 m. The average static stress drop is about 90 MPa. The highest static stress drop更多还原