Collaborative Research: The Mechanics of Intermediate Depth Earthquakes: a Multiscale Investigation Combining Seismological Analyses, Laboratory Experiments, and Numerical Modeling

合作研究：中深度地震的力学：结合地震分析、实验室实验和数值模拟的多尺度研究

• 批准号: 1925965
• 负责人: Zhigang Peng
• 金额: $ 52.99万
• 依托单位: Georgia Tech Research Corporation
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-12-15 至 2024-11-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1925965&HistoricalAwards=false
• 关键词: Collaborative; Research; Mechanics; Intermediate; Depth

It has been nearly a century since deep earthquakes, below about 50 kilometers, were definitively detected. Due to the difficulty observing such events, however, the mechanisms that control deep earthquakes are still poorly understood compared to shallower events. The pressures at depths greater than roughly 50 kilometers prohibit the frictional sliding and weakening mechanism that is essential to shallow events. This project focuses on the study of intermediate-depth earthquakes, those between approximately 50 and 300 km depth. The collaborative research combines small-scale scale laboratory experiments conducted at high pressure and temperature, seismological analysis of intermediate-depth earthquakes in well instrumented areas, and physics-based computer simulations to bridge the scales between the laboratory experiments and field observations. The hypothesis, generated from preliminary laboratory data, is that the minerals typical in subducting plates that produce deep earthquakes undergo one or more transformations that lead to mechanical instability as pressure and temperature increase. Minerals can densify when changing phase, reducing pressure locally and allowing for fracture propagation. Other reactions may create weak surfaces between or within grains. Under the right conditions, these instabilities can slip, generating heat and causing a runaway reaction leading to an earthquake. Laboratory data can reproduce these reactions at the small scale. Computer models will be created using the laboratory data to reproduce these experiments and determine parameters for the models. The results will then be systematically scaled up to simulate intermediate depth earthquakes in actual subduction zones. These large-scale simulations will be compared to the characteristics of observed earthquakes, with improved sensitivity to detect micro-earthquakes based on novel template matching and machine learning techniques. The combined research will help to demonstrate, for the first time, whether the same processes observed in small-scale laboratory specimens can account for large intermediate-depth earthquakes in subduction zones. Intermediate-depth earthquakes, while less common than shallow earthquakes, do result in casualties and significant damage. Understanding the mechanisms that cause such events will help to better characterize the potential hazard and risks to seismically active areas. The interdisciplinary experimental, numerical, and seismological work has the potential to transform our understanding of deep seismic events and deep Earth interior that are difficult to observe directly. The project will support postdoctoral researchers, graduate students, and undergraduate students further their education. It has been nearly a century since deep earthquakes, defined as below about 50 km depths, were first unequivocally discovered. The pressures at these depths preclude the frictional sliding that dominates shallow earthquakes and mechanisms of deep earthquakes remain poorly understood. Many challenges surround the study of deep earthquakes, including the inability to physically examine the fault structure and directly observe earthquake slip in the deep Earth interior. Here, this project investigate the mechanisms behind intermediate-depth earthquakes, defined as those between about 50 and 300 km depths, by integrating three key approaches: (1) detailed seismological investigation over a few well-studied tectonic settings, (2) controlled laboratory experiments on candidate mineral/rock groups with potential mechanical instabilities triggered by high-pressure, high-temperature reactions with acoustic emission monitoring and quantitative waveform analyses, and (3) micromechanics-based mathematical and physical models with a multiple scaling scheme to cover rupture processes from mm to km scales. Emerging new seismological tools such as template matching and machine learning allow detection of microevents in subduction zones with unprecedented spatial and amplitude resolution. The more than 10-fold increase in event detection provides much more illumination of fault areas than previously available. With such advances, this study will focus on the subduction zones in Central and Northern Japan, to examine event distribution, frequency magnitude statistics, aftershock productivities, source properties, fault orientation and stress drops. Experimentally, a number of major constituents of subducting slabs such as partially serpentinized olivine, eclogitization of lawsonite blueschist facies rocks, and even harzburgite, are now known to produce mechanical instability. Several physical mechanisms have been proposed for intermediate-depth earthquakes based on these observations. Development of experimental devices have increased sample linear dimensions by a factor of about 10. New developments in broadband acoustic emission technology have permitted quantitative analyses of acoustic emission events ("labquakes") using state-of-the-art seismological tools. Therefore, earthquakes and labquakes can be treated in a unified fashion in seismological analyses, allowing direct and better comparison with observations at very different scales. Thermo-poro-mechanical models will account for phase transformations, and formation of nano-shear or reaction bands as observed in the experiments. Simulations will be conducted in three stages: (1) Simulate the small-scale experiments. Detailed scans of experiments will allow us to mimic the perturbation in material that will initiate the transformation. The experiments at this stage will be used to validate and improve the model, as well as fit model parameters. (2) Mathematically upscale the model, homogenizing the small-scale behavior, so that the model can be used to simulate earthquakes in plates. (3) Simulate the Japan subduction zone. The models will then be compared with seismological observations to valid the results.The study will complete one of the last pieces in the puzzle of intermediate-depth earthquakes: verifying whether observed phenomena in small-scale experiments can quantitatively be the principal mechanism for regular intermediate-depth earthquakes observed at large depths. The combined seismological, experimental, and multiscale numerical work has the potential to truly transform the approach to the study of intermediate-depth earthquakes. The research will also lead to the development of new numerical methods and their application to geophysical processes.The subject of intermediate-depth earthquakes bears enormous societal impact with great scientific significance to the Earth and planetary science community. Large intermediate-depth earthquakes are capable of producing significant damage and casualties. Hence, an improved understanding of their mechanisms helps mitigate seismic hazard from these events. The mechanics of solids under high pressures is also of interest to physicists and material scientists. The interdisciplinary nature of the work has diverse applications throughout science and engineering.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.

自从发现深地震以来，已有将近一个世纪的时间被发现。但是，由于观察到此类事件的困难，与较浅的事件相比，控制深层地震的机制仍然很少了解。深度大于大约50公里的压力禁止摩擦滑动和弱化的机制，这对于浅层事件至关重要。该项目的重点是对中间深度地震的研究，大约50至300 km。该协作研究结合了在高压和温度下进行的小规模实验室实验，对仪器良好区域中中期地震的地震分析以及基于物理的计算机模拟，以弥合实验室实验和现场观测之间的尺度。 该假设是由初步实验室数据产生的，是俯冲板中典型的矿物质会导致深层地震发生一种或多种转化，从而导致随着压力和温度升高的机械不稳定性。矿物质在改变相位时可以致密，从而降低局部压力并允许断裂传播。其他反应可能会在谷物之间或内部产生弱的表面。在正确的条件下，这些不稳定性可以滑落，产生热量并导致失控的反应，导致地震。实验室数据可以小规模重现这些反应。将使用实验室数据创建计算机模型来重现这些实验并确定模型的参数。然后将系统缩放结果以模拟实际俯冲带中的中间深度地震。这些大规模的模拟将与观察到的地震的特征进行比较，并提高了基于新型模板匹配和机器学习技术的微观敏感性。合并的研究将有助于首次证明在小规模实验室标本中观察到的相同过程是否可以解释俯冲带中的大型中间严重地震。中等深度地震虽然不如浅层地震，但确实会导致伤亡和严重损害。了解引起此类事件的机制将有助于更好地表征对地震活跃地区的潜在危害和风险。跨学科的实验，数值和地震学工作有可能改变我们对深层地震事件和深层内部的理解，而深层地球内部很难直接观察。 该项目将支持博士后研究人员，研究生和本科生进一步学业。 自从深地震（定义为低于50公里的深度）以来，已经毫无疑问地发现了将近一个世纪的地震。在这些深处的压力排除了主导地震和深层地震机制的摩擦滑动。许多挑战围绕着深层地震的研究，包括无法物理检查断层结构并直接观察到深层内部的地震滑动。在这里，该项目通过整合三种关键方法来调查中间深度地震的机制，该机制被定义为大约50至300 km的深度：（1）对几个构造构造环境的详细地震学调查，（2）对候选矿物/岩石的较高启动性触发较高的机械启动，高级和较高的机械性启动，并触发了高级启动的候选矿物/岩石组，该实验是对候选型矿物/岩石组的较高的实验。定量波形分析以及（3）具有多个缩放方案的基于微力学的数学和物理模型，以覆盖从MM到KM尺度的破裂过程。 新兴的新地震学工具（例如模板匹配和机器学习）可以检测具有前所未有的空间和振幅分辨率的俯冲带中的微事件。事件检测的增长超过10倍，比以前可用的故障区域提供了更多的照明。有了这样的进步，这项研究将集中在日本中部和北部的俯冲带上，以检查事件分布，频率量统计，余震生产率，源特性，断层方向和应力下降。在实验上，已知许多主要组成板，例如部分蛇形化的橄榄石，Lawsonite Blueschist相岩石甚至哈尔兹堡的岩石化，现在已知可以产生机械不稳定性。基于这些观察结果，已经提出了几种物理机制，用于中间深度地震。实验设备的开发使样品线性尺寸提高了约10倍。宽带声发射技术的新发展允许使用先进的地震学工具对声学事件（“ Labquakes”）进行定量分析。因此，可以在地震分析中以统一的方式处理地震和实验室，从而可以直接与更好的比较与截然不同的观测值进行比较。如实验中所观察到的，热孔机械模型将解释相变的相变和纳米剪切或反应带的形成。模拟将分为三个阶段：（1）模拟小规模实验。对实验的详细扫描将使我们能够模仿启动转换的材料中的扰动。在此阶段的实验将用于验证和改进模型以及拟合模型参数。 （2）在数学上高档模型，使小规模行为均匀，以便该模型可用于模拟板中的地震。 （3）模拟日本俯冲带。然后将将模型与地震学观察结果进行比较，以有效。该研究将完成中间深度地震难题中的最后一部分之一：验证小规模实验中观察到的现象是否可以定量地是常规中间次数的主要机制。合并的地震，实验和多尺度数值工作具有真正改变中间深度地震研究的方法。这项研究还将导致新的数值方法的发展及其在地球物理过程中的应用。中等深度地震的主题对地球和行星科学界具有巨大的科学意义，具有巨大的社会影响。大型中级地震能够造成重大损害和伤亡。因此，对其机制的改进理解有助于减轻这些事件的地震危害。物理学家和物质科学家也很感兴趣固体的力学。这项工作的跨学科性质在整个科学和工程中都具有多种应用。该奖项反映了NSF的法定任务，并使用基金会的知识分子优点和更广泛的影响评估标准，被认为值得通过评估来获得支持。

项目成果

Spatiotemporal Variations of Intermediate‐Depth Earthquakes Before and After 2011 Tohoku Earthquake Revealed by a Template Matching Catalog

• DOI: 10.1029/2023gl104068
• 发表时间: 2023-11
• 期刊: Geophysical Research Letters
• 影响因子: 5.2
• 作者: Qiushi Zhai;Zhigang Peng;Makoto Matsubara;K. Obara;Yanbin Wang

The Indo–Eurasia convergent margin and earthquakes in and around Tibetan Plateau

• DOI: 10.2465/jmps.190927
• 发表时间: 2020
• 期刊: Journal of Mineralogical and Petrological Sciences
• 影响因子: 0.7
• 作者: Yanbin Wang;Yangfan Deng;F. Shi;Zhigang Peng