The 18 March 2020 M w 5.7 Magna earthquake near Salt Lake City, Utah, offers a rare glimpse into the subsurface geometry of the Wasatch fault system-one of the world's longest active normal faults and a major source of seismic hazard in northern Utah. We analyze the Magna earthquake sequence and resolve oblique-normal slip on a shallow (30-35°) west-dipping fault at~9-to 12-km depth. Combined with near-surface geological observations of steep dip (~70°), our results support a curved, or listric, fault shape. High-precision aftershock locations show the activation of multiple, low-angle (<30-35°) structures, indicating the existence of a complicated fault system. Our observations constrain the deep structure of the Wasatch fault system and suggest that ground shaking in the Salt Lake City region in future Wasatch fault earthquakes may be higher than previously estimated. Plain Language Summary On 18 March 2020, a moment magnitude (M w) 5.7 earthquake occurred beneath Magna, Utah, a suburb of Salt Lake City. It was the largest earthquake in the Wasatch fault system in historical times, and shaking was felt throughout northern Utah. The mainshock and its aftershocks were located in the middle of a dense seismic network, generating a rare and valuable data set of strong ground accelerations from normal-faulting earthquakes. We analyzed the first~6 weeks of seismic data following the mainshock and found that the aftershocks formed a shallow-dipping planar structure at depths of 9-12 km, implying that this segment of the Wasatch fault zone has a curved shape. At the surface, it dips steeply to the west at~70°, but as the depth increases, the dip becomes progressively more shallow. Simulations of ground shaking from large earthquakes previously assumed that this fault segment was planar and steeply dipping (~50°) throughout the upper crust. Our observations suggest that the rupture area of future large earthquakes will be closer to the surface than previously thought, which would cause increased ground shaking in the Salt Lake City metropolitan area with its~1.2 million residents. Geological mapping and modeling of the WFZ finds near-surface dips of 45-90°on the SLCS (Bruhn et al., 1992); however, there are competing models for its subsurface structure under the Salt Lake ©2020. American Geophysical Union. All Rights Reserved.

Since 2008, earthquake sequences within the Fort Worth basin (FWB), north Texas, have been linked to wastewater disposal activities related to unconventional shale‐gas production. The North Texas Earthquake Study (NTXES) catalog (2008–2018), described and included herein, uses a combination of local and regional seismic networks to track significant seismic sequences in the basin. The FWB earthquakes occur along discrete faults that are relatively far apart (>30  km), allowing for more detailed study of individual sequence development. The three largest sequences (magnitude 3.6+) are monitored by local seismic networks (<15  km epicentral distances), whereas basinwide seismicity outside these three sequences is monitored using regional distance stations. A regional 1D velocity model for the FWB reflects basinwide well log, receiver function, and regional crustal structure studies and is modified for the larger individual earthquake sequences using local well‐log and geology data. Here, we present an mb_Lg relationship appropriate for Texas and a basin‐specific ML relationship, both calculated using attenuation curves developed with the NTXES catalog. Analysis of the catalog reveals that the earthquakes generally occur within the Precambrian basement formation along steeply dipping normal faults, and although overall seismicity rates have decreased since 2016, new faults have become active. Between 2006 and 2018, more than 2 billion barrels of fluids were injected into the Ellenburger formation within the FWB. We observe strong spatial and temporal correlations between the earthquake locations and wastewater disposal well locations and injection volumes, implying that fluid injection activities may be the main driving force of seismicity in the basin. In addition, we observe seismicity occurring at greater distances from injection wells (>10  km) over time, implying that far‐field stress changes associated with fluid injection activities may be an important component to understanding the seismic hazard of induced seismicity sequences.

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&gt;3–4) well recorded at regional-to-teleseismic distances (&gt;150  km); however, it is unclear whether they can be modified to work for small events (M&lt;3) well recorded only at local distances (&lt;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.

Immediately following the 18 March 2020 Mww 5.7 Magna, Utah, earthquake, work began on installing a network of three-component, 5 Hz geophones throughout the Salt Lake Valley. After six days, 180 geophones had been sited within 35 km of the epicenter. Each geophone recorded 250 samples per second data onsite for ∼40 days. Here, we integrate the geophone data with data from the permanent regional seismic network operated by the University of Utah Seismograph Stations (UUSS). We use machine learning (ML) methods to create a new catalog of arrival time picks, earthquake locations, and P-wave polarities for 18 March 2020–30 April 2020. We train two deep-learning U-Net models to detect P waves and S waves, assigning arrival times to maximal posterior probabilities, followed by a two-step association process that combines deep learning with a grid-based interferometric approach. Our automated workflow results in 142,000 P picks, 188,000 S picks, and over 5000 earthquake locations. We recovered 95% of the events in the UUSS authoritative catalog and more than doubled the total number of events (5000 vs. 2300). The P and S arrival times generated by our ML models have near-zero biases and standard deviations of 0.05 s and 0.09 s, respectively, relative to corresponding analyst times picked at backbone stations. We also use a deep-learning architecture to automatically determine 70,000 P-wave first motions, which agree with 93% of 5876 hand-picked up or down first motions from both the backbone and nodal stations. Overall, the use of ML led to large increases in the number of arrival times, especially S times, that will be useful for future tomographic studies, as well as the discovery of thousands more earthquakes than exist in the UUSS catalog.

We show that ML–Mc is a viable and regionally portable depth discriminant, and therefore may contribute to nuclear test ban treaty verification. A recent study found that the difference between local magnitude (ML) and coda duration magnitude (Mc) discriminates shallow seismic events (mining blasts, mining‐induced earthquakes, and shallow tectonic earthquakes) from deeper tectonic earthquakes in the Utah region. The shallow seismic events had anomalously high Mc values, with increasingly negative ML–Mc values as depth decreased. Here, we evaluate the performance of ML–Mc as a depth discriminant in three new regions and find that ML–Mc increases between 0 and 9 km depth in all cases. Initially, we investigated ML–Mc as a function of depth for naturally occurring earthquakes in the region around Yellowstone National Park, as recorded by the University of Utah Seismograph Stations. For 3358 Yellowstone earthquakes with well‐constrained depths, we found ML–Mc increased 0.030±0.007 magnitude units (m.u.) for each 1 km increase in depth up to 10 km depth. Next, we examined ML–Mc values for anthropogenic seismicity in northern Oklahoma and southern Kansas, as recorded by the National Earthquake Information Center. For 1628 events with well‐constrained depths, we computed a slope for ML–Mc of 0.022±0.010  m.u./km. Finally, we analyzed ML–Mc for 28,722 well‐located earthquakes in Italy, as recorded by the National Institute of Geophysics and Volcanology, and found an ML–Mc slope of 0.018±0.001  m.u./km. In each case, the quoted error bounds represent 95% confidence regions that exclude zero, implying that the depth dependence of ML–Mc is statistically significant. We performed several robustness tests in which we varied the criterion used to define a well‐constrained depth and the depth range used in the linear fit. In nearly all cases, we found a positive slope for ML–Mc versus depth at a confidence level above 95%.