2011年3月11日日本宫城MW9.1地震的预滑与黏滑震相研究

Pre-slip and Stick-slip Phases during the Tohoku, Japan,MW9.1 Earthquake on March 11, 2011

摘要:

实验室内的岩石黏滑实验表明：黏滑错动过程一般可分为预滑、黏滑和止滑三个阶段，其中黏滑过程往往不是一次单点错动过程，而是由多次黏滑错动过程组成，表现出在断层的不同部位多点黏滑错动的特征。
2011年3月11日日本宫城近海发生了M_W9.1大地震。对该地震在全球数字地震仪台网（GSN）的波形记录做了分析，在ERM台（Δ=3.8°）记录的Pn震相前约125.5 s处识别出预滑错动震相Xp；在GSN 98个台的长周期波形记录上识别出三次同震黏滑错动过程中激发出的三个同震黏滑错动震相Xs1、Xs2和Xs3，以及止滑过程中激发出的止滑面波震相XsQ和XsR。根据黏滑实验和观测结果，我们认为ERM台所处的地块在主震前约69.1 s时发生了一次临震预滑错动，激发出了预滑错动震相Xp。主震发生后与弹性破裂过程同时发生了第一次黏滑错动，激发出了黏滑错动震相Xs1；在弹性破裂开始后约27.5 s时发生了第二次黏滑错动，激发出了黏滑错动震相Xs2；在弹性破裂开始后约71.0 s时发生了第三次黏滑错动，激发出了黏滑错动震相Xs3；在弹性破裂开始后约93.1 s时黏滑错动幅度达到峰值Xsm。之后进入止滑阶段，止滑过程激发出了勒夫型长周期面波止滑震相XsQ和瑞雷型长周期面波止滑震相XsR。根据XsQ和XsR震相的周期普遍大于75 s的特征，我们认为XsQ和XsR可能是地幔内传播的面波，并给出XsQ和XsR面波的走时关系。根据主震发生后同震伴随有三个子黏滑错动过程的观测证据，认为此次M_W9.1大地震可能是黏滑错动和弹性破裂共同作用的结果。从地震记录图上识别出Xp震相有助于认识主震前的预滑错动过程，且有一定的前兆意义。研究Xs震相以及XsQ和XsR震相有助于认识同震黏滑错动过程并预判震灾损失。

Abstract:

The results of indoor rock stick-slip experiments show that one stick-slip dislocation process often includes three stages:pre-slip, stick-slip, and stop-slip. In addition, the stick-slip process is not a single point dislocation process; rather, it comprises many stick-slip dislocation processes, showing the characteristic of multi-point stick-slip dislocation in different parts of the fault.
An M_W9.1 earthquake occurred near the coast of Tohoku, Japan, on March 11, 2011. After analyzing the waveform records from the Global Seismographic Network (GSN), we identified the Xp phase about 125.5 s prior to the Pn phase from the records of the ERM station (Δ=3.8°). We then identified three stick-slip dislocation phases (Xs1, Xs2, and Xs3) and two stop-slip phases (XsQ and XsR) from the records of 98 stations in GSN. According to the results of stick-slip experiments and observation, we inferred that at the block where the ERM station was located, an impending pre-slip dislocation occurred about 69.1 s prior to the main shock, which excited the Xp phase. Following the main shock, the first stick-slip dislocation occurred during the elastic rupture process. Then, the stick-slip dislocation phase Xs1 was excited. About 27.5 s after the elastic rupture, the second stick-slip dislocation occurred, and the stick-slip dislocation phase Xs2 was excited. About 71.0 s after the elastic rupture, the third stick-slip dislocation occurred, and the stick-slip dislocation phase Xs3 was excited. About 93.1 s after the elastic rupture, the stick-slip dislocation amplitude reached its peak value of Xsm, then entered the stop-slip stage, wherein the stop-slip phase XsQ and XsR were excited. According to the fact that the periods of XsQ and XsR phases are generally greater than 75 s, we suggested that the XsQ and XsR phases may be the surface wave propagating through the mantle. The travel time relationship between XsQ and XsR phases were given. According to the observed evidence that the main shock was accompanied by three sub-stick-slip dislocation events, we suggested that the M_W9.1 earthquake may have resulted from the cooperation of elastic rupture and stick-slip dislocation. Identifying the Xp phase from seismic records is helpful in understanding the pre-slip dislocation process before the main shock, with certain precursory significance. Studying the Xs, XsQ, and XsR phases will be helpful in understanding the coseismic stick-slip dislocation processes, and in judging the earthquake disaster loss.