Fault Zone Rheology and Deformation Mechanisms of Tibetan Plateau Constrained by Space Geodesy Measurements

我们建议利用大地震后测量的大地测量数据来研究青藏高原地壳和断层的流变行为。青藏高原地区近年来发生了两次大地震：2001年发生在高原北部东昆仑断裂带西段的7.8级科科西利地震和2008年发生在高原东缘龙门山断裂带的7.9级汶川地震。这些大地震提供了独特的机会来研究它们的发生所导致的应力变化，在断层带和周围物质中触发各种过程，这些过程反过来又随着时间的推移而放松应力。这项工作将有助于了解地震周期内的应力/应变演化，以及上下地壳之间以及断裂带内外应力之间的应力耦合如何导致系统失效。这种知识还将阐明大陆动力过程，有助于减轻和减少地震灾害。提出的工作将开发新的技术和计算机代码来模拟横向非均匀岩石圈中的粘弹性变形。以前的研究一直在争论哪一种机制主导了震后变形：是断层面的余震，特别是在脆性层和韧性层之间的过渡深度，还是下地壳和上地幔的粘弹性松弛。人们一直在争论，在高原的某些部分下面是否存在粘性地壳通道流动，如果存在，它是如何驱动高原内部和周围的变形的，特别是在其东部边缘的向东扩张/挤压，以及在构造过程中发生的地震。为了解决这些问题，我们计划开发粘弹性变形模型并模拟震后变形，使用空间大地测量，特别是GPS，观测作为约束。为了精确地解释数据，我们将修改代码来模拟层状介质中的粘弹性变形，并结合边界元方法来模拟断层横向属性变化的层状介质中的变形。利用GPS、InSAR、地质和地震资料反演模型的变形源和结构，更好地了解西藏北部和东部断裂带和地壳、上地幔的流变学。考虑到两次地震在空间上的接近性，他们的联合研究具有一定的优势：精确监测汶川地震后的变形将对初始阶段提供严格的约束，而不太精确监测的科科西利地震后的变形将分别对更长的时间跨度的模型提供约束。我们将比较2001年Kokoxili地震和2002年阿拉斯加Denali地震（均以走滑为主）以及2008年汶川地震和1999年台湾Chichi地震（均以逆断层为主）之间的建模结果。地震规模和断层机制的相似性以及地球结构的差异将有助于揭示西藏、阿拉斯加和台湾地震对的发震过程和岩石圈流变学。

We propose to study rheological behavior of the crust and faults of the Tibetan plateau using geodetic data measured following large earthquakes. Two large earthquakes occurred in the Tibetan plateau region in recent years: the 2001 Mw 7.8 Kokoxili earthquake occurred on the western section of the East Kunlun fault in northern plateau and the 2008 Mw 7.9 Wenchuan earthquake occurred on the Longmen Shan fault at the eastern rim of the plateau, respectively. These large quakes provide unique opportunities to study stress changes that result from their occurrences, triggering a variety of processes in the fault zone and surrounding materials that in turn relax the stress over time. The proposed work will help gain understanding about stress/strain evolution within an earthquake cycle, and how the stress coupling between the upper and lower crust and between the stresses inside and outside of a fault zone drives the system to failure. Such knowledge will also shed lights on continental dynamic processes, and be useful for seismic hazard mitigation and reduction. The proposed work will develop new techniques and computer code for modeling visco-elastic deformation in a lateral inhomogeneous lithosphere.Previous studies have debated which mechanism dominates the postseismic deformation: afterslip on the fault plane, particularly in the transition depth between the brittle and ductile layers, or visco-elastic relaxation in the lower crust and upper mantle. It has been debated whether viscous crust channel flow exists underneath certain parts of the plateau, and if so, how that drives deformation in and around the plateau, particularly the eastward expansion/extrusion at its eastern rim, and earthquakes occurred during the tectonic process. To tackle these issues we plan to develop visco-elastic deformation models and simulate postseismic deformation, using space geodetic, particularly GPS, observations as constraints. To precisely interpret the data we will modify a code to model visco-elastic deformation in a layered media, and incorporate a boundary element approach to allow simulating deformation in a layered media with lateral property change across faults. GPS, InSAR, geological, and seismological data will be used to invert for the deformation source and structure of the model, to better understand the rheologies of the fault zone and the crust and upper mantle in northern and eastern rim of Tibet. Given the spatial proximity of the two earthquakes, their joint study has some advantages: Precisely monitored post-Wenchuan deformation will provide tight constraints on the initial phase, and the less precisely monitored post-Kokoxili deformation will provide constraints on a longer time-span of the model, respectively. We will compare the modeling results between the 2001 Kokoxili and the 2002 Mw 7.9 Denali, Alaska (both predominantly strike-slip) earthquakes, and between the 2008 Wenchuan and the 1999 Mw 7.6 Chichi, Taiwan (both predominantly reverse-faulting) earthquakes. Similarities in the earthquake sizes and faulting mechanisms and differences in the Earth structures will shed light on the seismogenic processes for the earthquake pairs, and lithospheric rheologies of Tibet, Alaska, and Taiwan.