Microswarms: a lens into fault structure and aseismic processes deep in Southern California’s crust

地震群是在相对较短的时间内发生在特定地区的地震事件群。它们不同于标准的余震序列——大地震之后的一系列小地震——因为它们似乎是由非构造因素驱动的。最近的分析表明，引发地震群的最有可能的过程是地壳内部的流体迁移和缓慢的断层蠕变事件。在这里，研究人员检查了2008年至2020年在南加州收集的数据。他们使用机器学习算法，从地震噪声中提取小地震信号。他们发现了数百个未知的蜂群，持续了6个月到几年的时间。该团队现在分析这些新发现的序列，以量化它们的集体空间和时间模式。一个目标是确定是否存在可概括的特征。另一个目标是描述潜在的驱动过程，并记录这些集群发生的位置和频率。该研究旨在为非构造因素在驱动地震活动中的作用提供更全面的区域背景。其成果之一是公开搜索南加州近期地震群的高分辨率目录。该项目还为一名研究生提供支持，并向K-12学生和公众提供服务。它由NSF地球物理学和地理信息学项目共同资助。地震群被认为主要是由地震过程驱动的。因此，这些序列的时空演变编码了独特的信息：断裂带的结构和渗透性结构，地震过程的责任，以及断层性质和地震物理之间的耦合。在这里，研究人员利用深度学习算法在地震监测能力方面的最新进展。他们表明，在过去十年中，南加州经历了数百次以前未知的蜂群事件，持续数月至数年。这些地震微群现在被用来绘制发震地壳深处流体的潜在起源图，并量化这些短暂事件的频率和持续时间。将系统地审查微蝗群开始和停止的时间，以便与最大的区域事件联系起来。此外，大量的微群将用于提供断层带的几何、结构和渗透率结构的深度约束。这些数据对于理解地震群在空间和时间上的复杂演化至关重要。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Earthquake swarms are clusters of seismic events occurring in a given area within a relatively short period of time. They are distinct from standard aftershock sequences – series of small earthquakes following large earthquakes - in that they appear to be driven by non-tectonic factors. Recent analysis suggests that the most likely processes triggering earthquake swarms are fluid migration within Earth’s crust, and slow fault creep events. Here, the researchers examined data collected across Southern California between 2008 and 2020. They used machine learning algorithms which allow extracting small earthquake signals from the seismic noise. They discovered hundreds of unknown swarms that lasted 6 months to several years in duration. The team now analyze these newly identified sequences to quantify their collective spatial and temporal patterns. One goal is to identify if there are generalizable characteristics. Another goal is to characterize the underlying driving processes and document where and how often these swarms occur. The study aims to provide a more comprehensive regional context for the role of non-tectonic factors in driving earthquake activity. One of its outcomes is a publicly searchable high-resolution catalog of recent earthquake swarms in Southern California. The project also provide support for one graduate student and outreach to K-12 students and the public. It is funded by both NSF Geophysics and Geoinformatics programs.Earthquake swarms are believed to be driven mainly by aseismic processes. The spatiotemporal evolution of these sequences therefore encodes unique information: the structural and permeability architecture of fault zones, the aseismic processes responsible, and the coupling between fault properties and earthquake physics. Here, the researchers leveraged recent advances in earthquake monitoring capabilities with deep learning algorithms. They showed that during the last decade, Southern California has experienced hundreds of previously unknown swarm episodes that persist for months to years. These earthquake microswarms are now used to map the potential origins of fluids deep within the seismogenic crust and to quantify the frequency and duration of these transient episodes. The timing of the initiation and arrest of microswarms will be examined systematically for connections to the largest regional events. Furthermore, the large collection of microswarms will be used to provide constraints on the geometrical, structural, and permeability architecture of the fault zones at depth. These data are critical to understanding the complex evolution of earthquake swarms in space and time.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.