Automated Detection of Dynamic Earthquake Triggering by the High‐Frequency Power Integral Ratio

Abstract

Dynamic earthquake triggering can be used to investigate the responses of faults to stress disturbances. We develop a new method to detect dynamic triggering by estimating high‐frequency energy change before and during teleseismic waves using the HIgh‐Frequency power Integral ratio (HiFi). Our method is able to identify local events independent of earthquake catalog or subjective judgements. The significance in energy change is evaluated by a statistical analysis of the background ratio in a large number of days, which can suppress the influence of noise and variations in the background seismicity and yield a confidence level of dynamic triggering (0–1). We apply the HiFi method to the Geysers Geothermal Field in California, and the results are largely consistent with previous reports from the β statistic. By comparing the results of HiFi and β statistic, we select a confidence level range of 0.918–0.947 as the optimum threshold to identify dynamic triggering in the region.

Seismological Characterization of the 2021 Yangbi Foreshock‐Mainshock Sequence, Yunnan, China: More than a Triggered Cascade

“…Moreover, afterslips of M w > 4.5 earthquakes are found to be common in the Anza segment of San Jacinto Fault as indicated by the near‐repeating earthquakes (detected with CC > 0.9) and the correlated borehole strainmeter signal (Shaddox et al., 2021). If the Fault_M is already critically stressed before F 1, it would be susceptible to small static stress changes or even dynamic stresses of F 1 (Freed, 2005; Yun et al., 2019).…”

Characteristics of Frequent Dynamic Triggering of Microearthquakes in Southern California

“…Dynamic triggering has been shown to influence the occurrence of earthquakes, aseismic slow earthquakes, tectonic tremors, landslides, and icequakes (Aiken et al., 2013; Gomberg et al., 1997; Gonzalez‐Huizar et al., 2012; H. P. Johnson et al., 2017; Obara, 2002; Peng et al., 2014; Tymofyeyeva et al., 2019; Wallace et al., 2017). Triggered shallow crustal earthquakes occur at a variety of fault systems, including subduction zones, continental plate boundaries, and geothermal fields (Brodsky et al., 2003; Fan & Shearer, 2016; Kaven, 2020; Nissen et al., 2016; Velasco et al., 2008; Yun et al., 2019). For example, the 2012 M w 8.6 Indian Ocean earthquake dynamically triggered earthquakes on a global scale (Pollitz et al., 2012).…”

“…P. Johnson et al, 2017;Obara, 2002;Peng et al, 2014;Tymofyeyeva et al, 2019;Wallace et al, 2017). Triggered shallow crustal earthquakes occur at a variety of fault systems, including subduction zones, continental plate boundaries, and geothermal fields (Brodsky et al, 2003;Fan & Shearer, 2016;Kaven, 2020;Nissen et al, 2016;Velasco et al, 2008;Yun et al, 2019). For example, the 2012 M w 8.6 Indian Ocean earthquake dynamically triggered earthquakes on a global scale (Pollitz et al, 2012).…”

Rupture propagation on heterogeneous fault: Challenges for predicting earthquake magnitude

“…针对由断层⾮均匀性导致的"震中-震级测不准关系", 对⽬标断层开展地震动⼒学模拟 是⼀个可⾏的⼿段。事实上, 地震动⼒学数值模拟已经被⼴泛应⽤于地震情景构建, 包括现 代及历史地震 [82] 。若与⽬标断层的地质背景、滑移速率、能量积累等结合, 可应⽤于⾮均 , 表明了利⽤地震动⼒学模拟来预测震级的可⾏性。 地震破裂受断层上应⼒⽔平, 摩擦性质, 以及断层结构共同控制。因此开展可靠的动⼒ 学模拟, 需要对上述因素进⾏合理约束。我们可以通过地震学⽅法研究断层结构；⽬前基 于密集台阵的断裂带成像、地震精定位等⼿段都可⽤于刻画断层结构 [54,84,85] , 并发展了⼀ 系列基于密集台阵的新⽅法 [57,86,87] 。约束断层在孕震深度的应⼒分布缺乏直接测量数据, ⽬前最有效的⽅法是利⽤现代⼤地测量⼿段, 反演震间期断层上的闭锁程度分布, 进⽽计算 断层上的应⼒分布, 圈定断层的⾼应⼒块体 [9,10,83] 。由于断层摩擦属性很难进⾏直接约束, ⽬前主要参考实验室摩擦实验结果 [88] , 或利⽤近场观测结合动⼒学参数反演的⽅法约束不 同地质条件下的同震摩擦属性 [89,90] 。结合以上⾮均匀断层性质进⾏动⼒学模拟, 可观察地 震是否能突破块体间的低应⼒区, 并得到未来可能的地震情形 [9,10,91] 。 除动⼒学模拟外, Noda 等⼈ [92] 从能量平衡的⻆度, 计算断层已经积累的弹性能和地震 所消耗的能量(摩擦热和破裂能的总和), 判断块体是否能破裂甚⾄发⽣穿越不同块体的 级联破裂。其原理表述如下: 累积弹性能远⼤于破裂传播所消耗的能量, 则破裂可以持续 传播, 最终产⽣⼤地震。但计算破裂能和摩擦⽣热都基于关键同震摩擦参数的数值。如前 所述, ⽬前估算发震断层的摩擦属性尚存挑战 [90,93] , 参数的估计也存在相当的不确定性。据 此估算未来地震的震级也存在相当的不确定性。 2.2 地⾯运动多变性与灾害评估 除了地震的最终震级, 地⾯运动强度预测是地震灾害评估中⾮常重要的⼀环。⽬前绝 ⼤部分灾害评估采⽤经验预测公式(empirical Ground motion prediction equations, GMPEs) 对地震可能产⽣的地⾯运动强度进⾏预测 [94,95] 。在 GMPEs 中, 地震引起的地⾯运动强度 主要受到地震震级、震中距、传播介质属性(尤其为近地表速度结构, ⽐如地下 30 m 的剪 切波速度结构𝑉 𝑠 30)控制。然⽽, 对于复发周期⻓的⼤地震, 此类经验预测公式通常缺乏近 场数据。⽽近场地⾯运动受到地震破裂过程的影响较为剧烈, 其破裂过程, 包括破裂⽅向 [96] 以及地表破裂分布 [97] 等, 对近场地⾯运动灾害分布起到⾄关重要的作⽤, 这对于经验公式 在⼤地震预测中的应⽤提出了挑战。 基于⾮均匀断层的动⼒学破裂模型未来可⽤于预测⼤地震地⾯运动 [98,99] 。上⽂中提到, 地震的破裂⽅向性可能受到断层物质属性及孕震带结构影响, 需要结合动⼒学模型进⾏判 断。即使断层性质已知, 地震从不同位置起破依然会引起不同的⽅向性和地⾯响应。如在 俯冲带, 当地震从深部成核向海沟处传播时, 容易引起较⼤的浅部滑移和海底隆升, 增加海 A c c e p t e d https://engine.scichina.com/doi/10.1360/TB-2021-1086 啸⻛险 [9] 。⾛滑断层在⾮均匀应⼒条件下, 不同的起破点位置可能会造成不同的浅部滑移 分布和地表破裂分布 [10] [90,100,101] , 尤其是 板块边界的断层。但是, 许多陆内断层的构造不⼀, 孔隙压⼒是否遵循静⽔压或近静岩压的 变化规律则需要更多观测验证。通过研究⼩地震震群活动规律与某些已知应⼒扰动的关系, 如远震动态触发 [102][103][104] 、潮汐…”