Collaborative Research: Integrating Observations of Low-Velocity Fault Zones with Models of Spontaneous Dynamic Earthquake Ruptures

通过断层陷波和同震形变的大地测量观测，发现了许多活动断层周围的低速断裂带，其地震波速度相对于围岩降低了几到几十个百分点。这些lvfz在机械上与周围的地壳岩石不同，它们对附近同震破裂的反应证明了这一点。1999年Hector Mine地震的同震破裂导致1992年Landers断裂带愈合过程的时间逆转，2004年M6 Parkfield地震的同震破裂导致San Andreas断裂带内地震速度净下降至少2.5%。由于附近的同震破裂，大地测量也观察到几个lvfz之间几毫米到几厘米的位移。以往研究中用于验证lvfz观测结果的震源模型要么是位错（运动滑动）模型，要么是点源模型，这些模型不能准确表征近源动应力场和变形场。本项目采用最先进的自发动态破裂模型来研究lvfz对附近同震破裂的响应。自发动态破裂模型建立在弹性、弹塑性和摩擦定律等物理原理的基础上，为实际地震的观测提供了物理见解。用于模拟几何复杂断层自发动态破裂和复杂速度结构中波传播的有限元方法（FEM）代码已经在社区范围内的代码验证工作中得到验证，并用于本研究。最近对FEM进行了修改，包括对动态破裂增加了断层外弹塑性响应，使pi能够更准确地计算动应力场，并更好地检查lvfz对同震破裂的响应。该项目还研究了破裂断层周围的LVFZ如何影响断层和近场地面运动的自发动态破裂。将断层外弹塑性响应纳入自发动力破裂模型，pi研究了非均匀材料分布、断层外损伤和自发动力破裂的相互作用以及对近场地震动的影响。该项目将lvfz观测与自发动力破裂模型相结合。与运动学或点源模型相比，自发动态破裂模型更准确地量化了lvfz对地震同震破裂的响应，这对提高我们对断层带力学和断层系统应力应变传递的理解具有重要意义。lvfz的观测为动态破裂模型中的参数提供了约束，包括lvfz的速度结构和材料强度（即黏聚力和内摩擦）。这些受观测约束的模型参数可用于支持其他重要问题的研究，例如预测关键设施地点和地震易发地区的强烈地面运动。本研究的结果有望促进地震科学中观测与理论的整合，并对以观测为导向的地震科学界和以理论为导向的地震科学界产生影响。

Low-velocity fault zones (LVFZs) around many active faults with a reduction in seismic wave velocities of several to several tens percent relative to wall rocks have been detected by fault zone trapped waves and geodetic observations of the coseismic deformation. These LVFZs are mechanically distinct from the ambient crustal rock, as evidenced by their responses to nearby coseismic ruptures. A temporal reversal of the healing process of the 1992 Landers fault zone due to the nearby coseismic rupture of the 1999 Hector Mine earthquake and a net decrease in seismic velocities of at least 2.5% within the San Andreas fault zone due to the coseismic rupture of the 2004 M6 Parkfield earthquake have been documented. Displacements of several millimeters to several centimeters across several LVFZs due to nearby coseismic ruptures have also been observed by geodetic surveys. Earthquake source models used in previous studies to examine these observations of LVFZs are either dislocation (kinematic slip) models or point source models, which cannot accurately characterize the near-source dynamic stress and deformation fields. This project utilizes the most advanced spontaneous dynamic rupture model to investigate responses of LVFZs to nearby coseismic ruptures. Spontaneous dynamic rupture models build upon physical principles such as elasticity, elastoplasticity, and friction laws and provide physical insights to observations from real earthquakes. A finite element method (FEM) code for simulating spontaneous dynamic ruptures along geometrically complex faults and wave propagation in complex velocity structures has been verified in a community-wide code validation effort and is used in this study. Recent modifications to the FEM, including addition of elastoplastic off-fault response to dynamic ruptures, allow the PIs to more accurately calculate dynamic stress field and to better examine responses of LVFZs to coseismic ruptures. This project also investigates how a LVFZ surrounding a rupturing fault affects spontaneous dynamic rupture on the fault and near-field ground motion. With elastoplastic off-fault response included in spontaneous dynamic rupture models, the PIs examine interactions of inhomogeneous material distribution, off-fault damage and spontaneous dynamic rupture and effects on near-field ground motion. This project integrates observations of LVFZs with models of spontaneous dynamic ruptures. Compared with kinematic or point source models, spontaneous dynamic rupture models provide more accurate quantification of how LVFZs respond to earthquake coseismic ruptures, which has significant implications for improving our understanding of fault zone mechanics and stress and strain transfer in fault systems. Observations of LVFZs provide constraints to parameters in dynamic rupture models, including velocity structure of LVFZs and material strength (i.e., cohesion and internal friction). These observation-constrained model parameters can be used to support studies of other significant questions, such as prediction of strong ground motion at critical facility sites and in earthquake-prone areas. Results from this study are expected to promote the integration of observation and theory in earthquake sciences, and to impact both observation-oriented and theory-oriented communities of earthquake sciences.