Discrimination of Small Earthquakes and Buried Single-Fired Chemical Explosions at Local Distances (<150  km) in the Western United States from Comparison of Local Magnitude (ML) and Coda Duration Magnitude (MC)

Abstract

Seismologists distinguish underground nuclear explosions from more commonly occurring earthquakes using moment tensor inversion, high-frequency P/S amplitude ratios, mb:Ms comparisons, and P-pP differential travel times. These methods are generally successful for large seismic events (M>3–4) well recorded at regional-to-teleseismic distances (>150 km); however, it is unclear whether they can be modified to work for small events (M<3) well recorded only at local distances (<150 km). Here, we evaluate a recently proposed, local-distance seismic source discriminant—the difference between local magnitude (ML) and coda duration magnitude (MC)—using seismograms of earthquakes and buried, single-fired chemical explosions recorded in three regions of the western United States. The quantity ML–MC was previously found to be sensitive to source depth, effectively discriminating mine blasts, induced earthquakes, and very shallow tectonic earthquakes from deeper crustal earthquakes. In this study, we report the first evaluation of ML–MC as a depth discriminant using data from buried, single-fired explosions that, unlike the seismic sources studied earlier, are good analogs for underground nuclear explosions. We find that even when using generic, uncalibrated methods of assigning magnitudes, ML–MC separates single-fired explosions and earthquakes. The area under the receiver operating characteristic curve is 0.92 for 19 explosions and 14 earthquakes in Washington, 0.90 for 22 explosions and 90 earthquakes in Wyoming, and 0.99 for three explosions and 149 earthquakes in Nevada. ML:MC comparisons have the potential to enhance discrimination based on high-frequency P/S amplitudes ratios—which perform less well at local than regional distances—because the two metrics have complementary sensitivities.

Advancing Local Distance Discrimination of Explosions and Earthquakes With Joint P/S and ML‐MC Classification

“…M L and M C values for all events are calculated from the waveforms; none of the M L values are adopted from the USGS or local monitoring catalogs. To emulate the situation that might exist when applying these calculations to an uncalibrated region, a consistent calibration of both M L and M C (optimized for Utah) is applied to all data sets (e.g., BASE, MSH, and SSIP adopted from Koper et al., 2021). Although magnitude calibrations vary between regions, the priority for this study is that discriminant between source types remains consistent.…”

“…Local‐scale source classification, including machine‐learning based techniques (e.g., Linville et al., 2019), often rely on high‐frequency body wave measurements because long period surface waves are barely excited by small magnitude events (e.g., Kim et al., 1993; O’Rourke & Baker, 2017; Taylor et al., 1989; Walter et al., 2018). Accordingly, depth‐sensitive magnitude‐based screening metrics, like the difference between local and coda duration magnitudes (M L ‐M C ), have been proposed and successfully applied to several regions (Holt et al., 2019; Koper et al., 2021; Voyles et al., 2020; Zeiler & Velasco, 2009), replacing the difference between teleseismic body wave and surface wave magnitudes (m b :M S ) that is commonly used for larger events (Russell, 2006; Selby et al., 2012; Stevens & Day, 1985). Although both methods mentioned above show potential for local‐scale discrimination, empirical corrections from global‐to‐regional studies may be difficult to adapt to local crustal wave propagation (Anderson et al., 2009; Walter et al., 1995; Walter & Taylor, 2001).…”

“…The rapid growth of local‐to‐regional seismic data in recent years motivates revisiting classic waveform‐based discrimination methods at local distances (Koper et al., 2021; O’Rourke et al., 2016; Pyle & Walter, 2019; Wang et al., 2020). In this study, we systematically calculated P/S ratios and M L ‐M C values for earthquakes and borehole explosions recorded at distances <200 km by dozens of seismometers in four distinct regions.…”

Combining Deep Learning With Physics Based Features in Explosion‐Earthquake Discrimination

“…(2020) showed that high‐frequency P/S amplitude ratios can potentially be used for small‐magnitude seismic discrimination by averaging over many stations at local distances. Furthermore, a recently proposed depth discriminant ‐ the difference between local magnitude ( M L ) and coda magnitude ( M C ) ‐ also shows the ability to separate explosions from deeper, naturally occurring earthquakes (Koper et al., 2021; Voyles et al., 2020). These physics‐based discriminants provide a good understanding of the different characteristics between the two types of sources, and generally work very well in different regions and studies.…”

“…For example, O'Rourke et al ( 2016), Walter (2019, 2021) and Wang et al (2020) showed that high-frequency P/S amplitude ratios can potentially be used for small-magnitude seismic discrimination by averaging over many stations at local distances. Furthermore, a recently proposed depth discriminant -the difference between local magnitude (M L ) and coda magnitude (M C ) -also shows the ability to separate explosions from deeper, naturally occurring earthquakes (Koper et al, 2021;Voyles et al, 2020). These physics-based discriminants provide a good understanding of the different characteristics between the two types of sources, and generally work very well in different regions and studies.On the other hand, recent successful applications of machine learning to various areas in seismology (Bergen et al, 2019;Karpatne et al, 2019;Kong et al, 2019) suggest that a data-driven approach might be suitable for source classification problems.…”