Modeling of Nucleation and Dynamic Rupture on Heterogeneous Frictional Interfaces with Applications to Foreshocks
Modeling of Nucleation and Dynamic Rupture on Heterogeneous Frictional Interfaces with Applications to Foreshocks
复制标题
异质摩擦界面上的成核和动态破裂建模及其在前震中的应用
DOI：
--
复制
发表时间：
2018
期刊：
影响因子：
--
通讯作者：
Natalie Schaal
中科院分区：
文献类型：
--
作者：
Natalie Schaal
虽然许多大地震之前都会发生可观测到的前震，但导致这些小规模地震事件发生的机制仍然不确定。对支持不断增加的前震的一种物理解释是，它们是由于即将到来的主震成核而与不同属性的断层斑块相互作用而产生的。更好地了解地震如何在异质断层上成核将提高我们预测潜在危险事件的能力。
考虑到这一动机，我们试图了解什么条件会在主震成核区域内产生孤立的微震活动，并研究由此产生的事件的机制。受实验室实验建议的启发，即前震发生在粗糙体上，即整体压缩使平面性的局部偏差变平，我们通过在抗震和地震滑动的长期序列上对其滑动进行 3D 数值模拟，探索了嵌入速率和状态断层模型的较大发震区域中的较高压应力的粗糙体型斑块的行为。我们的模型确实在较大地震事件的抗震成核过程中产生了较小尺度的地震活动，并且我们探索了它们的特性以及产生它们所需的长度尺度的分离。这些类似前震的事件具有与实验室和现场观察一致的应力下降，并且近似恒定，尽管分配给源斑块的压缩高度升高。造成合理应力降的两个主要因素是斑块周围区域的破裂程度和地震事件发生前的地震应力释放。我们还使用通常应用于自然微震事件的频谱分析来研究粗糙型事件的地震学特性。我们发现地震学方法无法充分捕捉模拟事件的特性。在某种程度上，其应力降的地震学估计与根据故障应力变化确定的实际应力降有显着不同。这是因为我们的震源比当前地震学方法所依据的标准模型具有更复杂的特征，包括破裂区域的不均匀应力变化（初始应力变化更大）和不均匀破裂速度。我们识别了粗糙型震源的远场地震图中的特征，这些特征与标准模型不同，并且可能是粗糙型震源的潜在特征。
我们的微震源的粗糙型模型可以深入了解有利于在自然和实验室断层上产生前震的条件以及由此产生的事件的特性。本研究中两个观点共同得出的结论——动态模拟非均质断层模型中震源的行为并从地震学角度分析其远场震源谱——具有值得进一步研究的重要意义。未来研究的主题包括较小规模地震事件之间的相互作用及其在主震成核过程中的作用、时间对其震源特性的影响以及与后续主震的所谓地震成核阶段的关系。
While many large earthquakes are preceded by observable foreshocks, the mechanisms responsible for the occurrence of these smaller-scale seismic events remain uncertain. One physical explanation of foreshocks with growing support is that they are produced by the interaction of slow slip, due to the nucleation of the upcoming mainshock, with fault patches of different properties. Having a better understanding of how earthquakes nucleate on heterogeneous faults would increase our capacity to forecast potentially hazardous events.
With this motivation in mind, we seek to understand what conditions produce isolated microseismicity within the nucleating region of the mainshock and to study the mechanics of the resulting events. Inspired by the suggestion from laboratory experiments that foreshocks occur on asperities, i.e., local deviations from planarity that are flattened by the overall compression, we explore the behavior of asperity-type patches of higher compressive stress embedded in the larger seismogenic region of a rate-and-state fault model by conducting 3D numerical simulations of their slip over long-term sequences of aseismic and seismic slip. Our models do produce smaller-scale seismicity during the aseismic nucleation of much larger seismic events, and we explore their properties as well as the separation in length scales needed to produce them. These foreshock-like events have stress drops that are consistent with laboratory and field observations and approximately constant, despite the highly elevated compression assigned to the source patches. Two main factors contributing to the reasonable stress drops are the significant extent of the rupture into the region surrounding the patch and the aseismic stress release just prior to the seismic event. We also investigate the seismologically-derived properties of the asperity-type events using the spectral analysis commonly applied to natural microseismic events. We find that the seismological methods cannot adequately capture the properties of the simulated events. In part, the seismological estimates of their stress drops are significantly different from the actual stress drops determined from the on-fault stress changes. This is because our sources have more complex features than the standard models from which the current seismological methods have been built, including heterogeneous stress change over the rupture area with much larger initial stress change, and heterogeneous rupture speed. We identify features in the far-field seismograms of the asperity-type sources that differ from the standard models and can be potentially characteristic of the asperity-type sources.
Our asperity-type models of microseismicity sources provide insight into the conditions conducive for generating foreshocks on both natural and laboratory faults and the properties of the resulting events. The conclusions provided jointly by the two perspectives in this study -- dynamically simulating the behavior of seismic sources within heterogeneous fault models and seismologically analyzing their far-field source spectra -- have important implications that warrant further study. Topics for future research include the interaction among smaller-scale seismic events and their role in the mainshock nucleation process, the effect of timing on their source properties, and relation to the so-called seismic nucleation phase of the subsequent mainshock.