Modeling slow slip and earthquake nucleation on heterogeneous faults: implications for foreshocks and repeating earthquakes

研究人员开发了复杂的数字模型，探索产生大小地震的断层如何产生缓慢的运动，看起来与典型的地震不同。地震通常是由地球内部先前存在的断层在构造板块运动的作用下快速滑动而产生的。观测表明，快速滑动的平均速度约为1m/S，与慢得多的滑动并存，可以利用全球定位系统进行测量。这些类型的现象往往存在于同一地点，并可以相互影响。这个项目将模拟几次大地震的循环，并通过结合我们所知的地球如何变形、地震波和断层的物理学，观察缓慢事件以及小地震和大地震是如何工作的。他们将进行模拟，并将其与实验实验室和自然界中的观察结果进行比较。这些研究将有助于更好地了解地震成核、断层性质和前震序列的预测价值，从而有助于减轻地震灾害。结合波传播和场地和建筑响应的程序，模拟慢滑和地震序列的方法可以成为基于物理的地震危险性分析模拟能力的重要组成部分。这种模拟可能会对我们对地震及其危险的理解产生革命性的影响。这些类型的模型为我们的危险研究增加了另一种工具。为了研究慢滑动与产生地震活动的断层块的相互作用，研究人员将使用一种独特的数值方法来解决滑动的所有阶段：地震成核期间的加速滑动、具有所有惯性(波)效应的动态破裂、震后滑动和包括瞬变的地震间滑动。慢滑将作为长期断层模拟的一部分，使用几种物理机制进行比较：(I)由标准速度衰减率和状态摩擦控制的断层上的成核过程，(Ii)由于非弹性剪胀而引起的这种过程的修改，以及(Iii)速度增强断层上的慢滑。断层模型将包括一个或多个被定义为具有不同摩擦特性的区域的斑块。将探讨可能导致地震活动性的几种类型的差异，包括有效法向应力、速率弱化特性、扩张性和渗透率的变化。这里的重点是模拟和描述慢滑和地震事件的相互作用，包括它们的辐射谱、应力降、在单个慢滑事件期间在斑块上再次发生地震的可能性、所产生的事件相对于斑块大小的空间范围、由此产生的地震活动对慢滑的修正及其震后影响，以及附近斑块之间相互作用的后果。将这些特征与观察结果(包括实验室研究)进行比较，将能够确定与观察到的现象有关的情景，并澄清自然断层和实验室断层的异质性的性质。开发的慢滑与非均质块相互作用的详细模型将用于创建微震活动的简化表示，用于未来模拟地震和慢滑，包括现实断层环境中的前震序列和震动。

The researchers have developed sophisticated numericals models that explore how faults that produce earthquakes, large and small, also produce movements that are slow, and look different from typical earthquakes. Earthquakes are typically generated by rapid slip on pre-existing faults inside the Earth, loaded by tectonic plate motion. Observations show that the rapid slip, with the average velocity of the order of 1 m/s, coexists with much slower slip that can be measured, e.g. using GPS. These types of phenomena often exist in the same place and can influence each other. This project will look to simulate several cycles of large earthquakes and, by incorporating what we know about how the Earth deforms, the physics of seismic waves and faults, observe how slow events and both small and large earthquakes work. They will take their simulations and compare them with observations made in experimental labs and in nature. The studies will lead to better understanding of earthquake nucleation, fault properties, and the predictive value of foreshocks sequences, and hence help mitigate seismic hazard. Combined with codes for wave propagation and site and building response, the methodology for simulating slow slip and earthquake sequences can become a vital ingredient of physics-based simulation capability for seismic hazard analysis This kind of modeling could have transformative impacts on our understanding of earthquakes and their hazards. These types of models add another tool to our study of hazards.To study the interaction of slow slip with seismicity-producing fault patches, the researchers will use a unique numerical approach capable of resolving all stages of slip: accelerating slip during earthquake nucleation, dynamic rupture with all inertial (wave) effects, postseismic slip, and interseismic slip including transients. Slow slip will be generated as a part of long-term fault simulations using several physical mechanisms to enable comparison between them: (i) nucleation processes on faults governed by standard velocity-weakening rate-and-state friction, (ii) modifications of such processes due to inelastic shear dilatancy, and (iii) slow slip on velocity-strengthening faults. The fault models will include one or multiple patches defined as areas of different frictional properties. Several types of differences that may lead to seismicity will be explored, including variations in effective normal stress, rate-weakening properties, dilatancy, and permeability. The focus here is on simulating and characterizing well-resolved interactions of slow slip and seismic events in terms of their radiated spectra, stress drops, potential for seismicity recurrence on a patch during a single episode of slow slip, spatial extent of the generated events in relation to the patch sizes, modification of slow slip by the resulting seismicity and its postseismic effects, and the consequences of interaction between nearby patches. Comparison of these features with observations (including laboratory studies) will allow identification of scenarios relevant to the observed phenomena as well as clarify the nature of heterogeneity on natural and laboratory faults. The developed detailed models of slow slip interaction with heterogeneous patches will be used to create simplified representations of microseismicity for use in future simulations of earthquakes and slow slip, including foreshock sequences and tremor, in realistic fault settings.