2004年苏门答腊地震的几个断层滑动模型的全球同震位移对比

2004年苏门答腊大地震后，不同作者根据地震波和/或GPS观测，提出了不同的断层错动模型.在利用同震位移观测资料反演断层滑动模型时，由于使用半无限空间均匀介质模型或半无限空间分层介质模型，一般只能利用近场位移GPS观测约束，无法利用远场资料，这些模型有时差别颇大，如何区别这些模型的优劣是一个仍尚未解决的问题.本文采用等效体力有限元方法，在考虑地球球形和分层的条件下，对四个不同作者提供的2004年苏门答腊地震的断层滑动模型计算全球同震位移.由于采用了球形模型，所以不仅可以把四个模型的近场位移计算结果与GPS数据进行对比，而且可以把远场位移计算结果与GPS数据进行对比.我们发现，垂直位移对断层滑动模型的依赖性小于水平位移.四个模型计算的近场位移与GPS位移符合程度均较好，但是四个模型计算的远场位移与GPS位移符合情况有很大不同，其中Chlieh等（2007）模型在近场与远场符合程度均很好，是四个模型中最好的.另外还探讨了断层反演数据资料、断层几何模型以及地球模型对计算结果的影响.对于特大地震，全球同震位移观测与计算值吻合程度的好坏是衡量断层滑动模型的合理性的一个重要依据.

Comparisons of global coseismic displacements from several fault slip models for the 2004 Sumatra earthquake

After 2004 great Sumatra earthquake, some authors presented different fault slip models from seismic waves and/or GPS data. However, because half-space homogenous or layered medium was adopted, the inversion of the coseismic model from observed coseismic displacements can only use near-field GPS data as the constraint but not far-field data. This can result in great differences among fault slip models. Therefore, how to evaluate those modes is an important problem. This paper uses the equivalent body force finite elements method, considering the earth curvature and layered structure, to compute global coseismic displacements using four fault slip models of the 2004 Sumatra earthquake provided by different authors. Comparisons of calculated coseismic displacements show that the vertical displacement is less sensitive to the fault slip model. Calculated coseismic displacements of all the four models agree well with near-field GPS observations, but significantly different in comparison with far-field GPS observations. In the whole four models, the Chlieh model is the best. The effects of data used in inversion, the fault geometry and the earth model are discussed in this paper. For great earthquakes, the consistency between calculated and observed global coseismic displacement is one important criterion for evaluating the fault slip model.