Detecting Real Earthquakes Using Artificial Earthquakes: On the Use of Synthetic Waveforms in Matched‐Filter Earthquake Detection

Abstract

Matched‐filters are an increasingly popular tool for earthquake detection, but their reliance on a priori knowledge of the targets of interest limits their application to regions with previously documented seismicity. We explore an extension to the matched‐filter method to detect earthquakes and low‐frequency earthquakes on local to regional scales. We show that it is possible to increase the number of detections compared with standard energy‐based methods, with low false‐detection rates, using suites of synthetic waveforms as templates. We apply this to a microearthquake swarm and an aftershock sequence, and to detect low‐frequency earthquakes. We also explore the sensitivity of detections to the synthetic source's location and focal mechanism. Source‐receiver geometry has a first‐order control on how sensitive matched‐filter detectors are to variations in source location and focal mechanism, and this likely applies to detections made using both synthetic and real templates.

Optimization of the Match‐Filtering Method for Robust Repeating Earthquake Detection: The Multisegment Cross‐Correlation Approach

“…For horizontal separations in particular, the resolution is approximately the same for both conventional MF and MFMC independent of station locations ( Figure S9). Therefore, station distribution can be an important factor that affects the depth resolution of the conventional multistation MF method, consistent with the finding of Chamberlain and Townend (2018). In contrast, the performance of MFMC is independent of the setup of seismic stations.…”

“…The match-filtering (MF) method uses waveform cross-correlation to determine the similarity between a pair of events. It is a powerful tool in modern seismology to identify repeating earthquakes (e.g., Huang & Meng, 2018;Igarashi et al, 2003;Matsuzawa et al, 2004;Meng et al, 2015;Nadeau et al, 1995;Naoi et al, 2015;Schaff & Richards, 2004Schmittbuhl et al, 2016;Uchida et al, 2003;Yamashita et al, 2012) and to detect events that can be easily missed by conventional phase arrival-based methods (e.g., Chamberlain & Townend, 2018;Gibbons & Ringdal, 2006;Peng & Zhao, 2009;Ross et al, 2019;Shelly et al, 2007;Schultz et al, 2014Schultz et al, , 2017Skoumal et al, 2014Skoumal et al, , 2015Skoumal et al, , 2019Warren-Smith et al, 2017Zhang & Wen, 2015). The extensive applications of this technique have led to major observational breakthroughs (e.g., Shelly et al, 2007).…”

Misconception of Waveform Similarity in the Identification of Repeating Earthquakes

“…In such a case, the computed CC can be extremely sensitive to hypocenter difference (Gao & Kao, 2020). We refrain from investigating the multistation scenario as the CC sensitivity is known to be strongly affected by network geometry (Chamberlain & Townend, 2018;Gao & Kao, 2020), and thus no general/common rules can be inferred objectively.…”

Recent advances in earthquake seismology using machine learning

“…Even for the template-matching problem, large-scale dataset requires the parallel computation of many GPUs (Ross et al 2019a). Furthermore, template matching using synthetic or empirical synthetic waveforms (Chamberlain and Townend 2018;Ide 2021) potentially contributes to better catalog development. Hence, an efficient similar waveform search is an important technique in catalog development.…”