Connecting long and short-range earthquake triggering with strain

地震触发为研究地震成核提供了一条基本线索。如果能在地震活动性变化和地震触发应变之间建立一种观测关系，那么我们将对成核过程有一个基本的约束。在这里，我们使用一种新颖的统计方法和大量的现代数据来测量由附近和远处地震引发的地震数量。初步结果表明，用地震波振幅测量的触发尺度强度与动应变呈连续的函数关系。无论距离如何，趋势都是一致的。然而，一些地区，如加利福尼亚，比其他地区，如日本，更容易触发地震。在小距离的触发行为中没有明显的交叉，这意味着动态触发足以解释局部余震。这一结果为一种基于观测到的地震波振幅的地震预报新方法指明了道路。本文提出了对这些初步结果进行测试、分析和应用的研究计划。具体而言，我们将：1)通过：a)改变测量方案中参数的选择来评估触发率与应变之间关系的鲁棒性。b)利用观测到的经深度校正的地震记录和测量到的震源机制，改进动静应变的代理。2)通过以下方式约束地震触发的物理性质：a)使用滤波波形和标准化事件间时间的分布确定立即触发还是延迟触发是一个更强的信号。b)比较新西兰和全球不同构造体系的触发趋势。c)将结果与地震触发的标准模型进行比较。3)将结果应用于减灾：a)将地震的插值地面速度与触发率函数相结合，预测下一小时的触发地震活动性。

Earthquake triggering provides a fundamental clue into earthquake nucleation. If an observational relationship could be established between seismicity rate changes and the triggering strain of an earthquake, then we would have a basic constraint on the nucleation process. Here we use a novel statistical approach and an abundance of modern data to measure the number of earthquakes triggered for both nearby and distant earthquakes. The preliminary results indicate that the intensity of triggering scales continuously as a function of dynamic strain as measured by seismic wave amplitude. The trend is consistent regardless of distance. However, some regions, like California, are more prone to triggering than others, like Japan. The lack of discernible crossover in triggering behavior at small distances implies that dynamic triggering is sufficient to explain local aftershocks. This result points the way to a new methodology for earthquake forecasting based on the amplitude of the observed seismic waves.This proposal presents a plan of study to test, analyze and apply these preliminary results. Specifically, we will:1) Evaluate the robustness of the relationship between triggering rate and strain by: a) Varying the choices of parameters in the measurement scheme.b) Improving the proxies for dynamic and static strain by using observed seismograms corrected to depth and measured focal mechanisms.2) Constrain the physics of earthquake triggering by:a) Determining whether immediate triggering or delayed triggering is a stronger signal using filtered waveforms and the distribution of the normalized inter-event times.b) Comparing the tendency to trigger across different tectonic regimes in New Zealand and globally. c) Comparing the results to standard models of earthquake triggering.3) Apply the results to hazard mitigation by:a) Combining the interpolated ground velocity from an earthquake with the triggered rate function to predict triggered seismicity in the next hour.