Coseismic and Postseismic Deformation from the 1999 Chi-Chi, Taiwan Earthquake

1999年9月21日，台湾集集发生MW7.5级地震，地震发生在台湾人运营的密集全球定位系统网的中心。在128个台站测量到了同震位移，震级高达12米。还记录了一组引人注目的震后瞬时形变信号。几个连续记录的GPS站在垂直和水平分量上都测量到了幅度超过10厘米的瞬时信号。密集的空间覆盖和非凡的信噪比使集集地震的同震和震后形变场可以说是有记录以来最好的。对同震位移的初步分析表明，地震是由东倾逆冲断层引起的，这与现场观测和地震学观测一致。然而，GPS数据不能与单一断裂面上的滑动进行拟合。地表断裂在其北端经历了一个90度的弯曲，长约15公里的东西走向断裂。余震震源机制表明，这一东西走向分支上有走滑运动。然而，GPS数据最适合于在N-S和E-W两个方向上都有斜滑的弯曲倾向断层。首选模型解释了数据中98%的方差。然而，残差太大了，而且在空间上是一致的，无法用测量误差来解释。车龙堡断裂是褶皱冲断系统的一部分，它将更古老和更坚硬的岩石侵位在沉积物上，相应的弹性性质的空间差异可能会影响建模。研究人员和他们的台湾合作者提出了一项详尽的同震变形分析，考虑了弹性性质的垂直和横向变化、非平面断层几何形状、滑动的空间变化，以及可能的不规则表面地形的影响。与同震位移一样令人兴奋的是，震后变形甚至更显著。震后加速变形早在几十年前就已为人所知，但对这一现象的物理机制却知之甚少。余滑、下地壳和上地幔的粘性流动、孔洞弹性松弛都被认为是解释震后瞬时形变的原因。理论研究比比皆是；问题在于缺乏对数据集的定义。集集地震的数据具有如此高的信噪比，我们希望最终回答这些长期存在的问题。具体地说，研究人员建议进行密集的建模工作，将粘弹性、孔弹性和后滑移模型的预测与数据进行比较。初步结果表明，余滑与主震后前100天的数据符合得很好，尽管还需要更多的分析。GPS数据的反演将揭示震后滑动的时空演化。这些结果将对碰撞带的构造产生重要的影响。许多褶皱冲断带模型显示，断层合并为台湾地下的主要滑脱。通过确定车伦堡断层在孕震和次震深度的几何形状，研究人员将对这一经典碰撞造山带的结构提供重要的见解。研究人员计划开发模型，既符合地震间大地测量数据，又能在综合多个地震周期时产生地质上合理的位移。这项拟议的研究将与台湾中央研究院地球科学研究所的台湾同事广泛合作，支持两个博士项目。

Segall 0106695 The Mw = 7.5, September 21, 1999 Chi-Chi, Taiwan earthquake occurred in the center of a dense GPS network operated by the Taiwanese. Coseismic displacements were measured at 128 stations, with magnitudes of up to 12 meters. A remarkable set of transient postseismic deformation signals were also recorded. Several continuously recording GPS stations measured transient signals with amplitudes of more than 10 cm in both the vertical and horizontal components. The dense spatial coverage and extraordinary signal to noise ratio make the coseismic and postseismic deformation fields of the Chichi earthquake arguably the best ever recorded. Preliminary analysis of the coseismic displacements demonstrate that the earthquake was caused by an east-dipping thrust, consistent with field and seismological observations. However, the GPS data can not be fit with slip on a single fault surface. The surface rupture undergoes a 90 degree bend at its northern end, with a ~15 km long east-west trending break. Aftershock focal mechanisms suggest strike slip motion on this E-W striking branch. The GPS data, however, are best fit with a curved dipping fault with oblique slip on both the N-S trending and E-W trending segments. The preferred model explains 98% of the variance in the data. The residuals, however, are far too large, and spatially coherent to be explained by measurement error. The Chelungpu fault, is part of a fold thrust system, which emplaced older and stiffer rocks over sediments and the corresponding spatial variations in elastic properties may bias the modeling. The investigators and their Taiwanese collaborators propose an exhaustive analysis of the coseismic deformation accounting for vertical and lateral variations in elastic properties, non-planar fault geometry, spatial variations in slip, and possibly effects of irregular surface topography. As exciting as the coseismic displacements are, the postseismic deformations are even more significant. Accelerated post earthquake deformation has been known for decades, however the physics of the phenomenon are poorly understood. Afterslip, viscous flow of the lower crust and upper mantle, and poroelastic relaxation have all been proposed to explain transient postseismic deformation. Theoretical studies abound; the problem has been a lack of defining data sets. The data from the Chichi earthquake is of such high signal to noise ratio that we hope to finally answer these long standing problems. Specifically, the investigators propose an intensive modeling effort to compare predictions from viscoelastic, poroelastic, and afterslip models to the data. Preliminary results suggest that afterslip provides a reasonable fit to the data from the first 100 days after the mainshock, although much more analysis is needed. Inversion of the GPS data will reveal the spatiotemporal evolution of postseismic slip. These results will have important implications for the tectonics of collision zones. Numerous models of fold thrust belts show faults merging into a master decollement beneath Taiwan. By determining the geometry of the Chelungpu fault at seismogenic and subseismogenic depths the investigators will provide important insights into the structure of this classic collisional orogen. The investigators plan to develop models that both fit interseismic geodetic data and yield geologically reasonable displacements when integrated over multiple earthquake cycles. The proposed research will support tow Ph.D projects in an extensive collaboration with Taiwanese colleagues at the Institute of Earth Sciences, Academica Sinica Taiwan.