Geology Reports⌕ Search

Geology topics

David R. Shelly

Publications and source records attributed to David R. Shelly.

At least 37 records · Page 2Linked to original sources
USGS2024

Distinct yet adjacent earthquake sequences near the Mendocino Triple Junction: 20 December 2021 Mw 6.1 and 6.0 Petrolia, and 20 December 2022 Mw 6.4 Ferndale

Two earthquake sequences occurred a year apart at the Mendocino Triple Junction in northern California: first the 20 December 2021 �w 6.1 and 6.0 Petrolia sequence, then the 20 December 2022 �w 6.4 Ferndale sequence. To delineate active faults and understand the relationship between these sequences, we applied an automated deep‐learning workflow to create enhanced and relocated earthquake catalogs for both the sequences. The enhanced catalog newly identified more than 14,000 M 0–2 earthquakes and also found 852 of 860 already cataloged events. We found that deep‐learning and template‐matching approaches complement each other to improve catalog completeness because deep learning finds more M 0–2 background seismicity, whereas template‐matching finds the smallest M < 0 events near already known events. The enhanced catalog revealed that the 2021 Petrolia and 2022 Ferndale sequences were distinct in space and time, but adjacent in space. Though both the sequences happened in the downgoing Gorda slab, the shallower Ferndale sequence ruptured within the uppermost slab near the subduction interface, while the onshore Petrolia sequence occurred deeper in the mantle. Deep‐learning‐enhanced earthquake catalogs could help monitor evolving earthquake sequences, identify detailed seismogenic fault structures, and understand space–time variations in earthquake rupture and sequence behavior in a complex tectonic setting.

California↗
USGS2024

Making phase-picking neural networks more consistent and interpretable

Improving the interpretability of phase‐picking neural networks remains an important task to facilitate their deployment to routine, real‐time seismic monitoring. The popular phase‐picking neural networks published in the literature lack interpretability because their output prediction scores do not necessarily correspond with the reliability of phase picks and can even be highly inconsistent depending on how we window the waveform data. Here, we show that systematically shifting the waveforms during training and using an antialiasing filter within the neural network architecture can substantially improve the consistency of the output prediction scores and can even make them scale with the signal‐to‐noise ratios of the waveforms. We demonstrate the improvements by applying these approaches to a commonly used phase‐picking neural network architecture and using waveform data from the 2019 Ridgecrest earthquake sequence.

The Seismic Record↗
USGS2024

Examining the connections between earthquake swarms, crustal fluids, and large earthquakes in the context of the 2020-2024 Noto Peninsula, Japan, earthquake sequence

Earthquake swarms are most commonly composed of small-magnitude earthquakes – those that may in some cases be felt but without causing damage. However, a recent study by Yoshida et al. (2023, https://doi.org/10.1029/2023GL106023) analyzed a swarm beneath the Noto Peninsula in Japan that, after more than two years of moderate-magnitude seismicity, triggered the moment magnitude (Mw) 6.2 Suza mainshock. Based on high-precision earthquake locations and a slip inversion of the mainshock, these authors found that the Mw 6.2 Suza earthquake occurred on the updip extension of a fault that was active during the swarm, likely driven by fluid pressure perturbations. After publication of that paper, a much larger and more destructive Mw 7.5 event occurred nearby. These events underscore the potential for swarms to be precursors to large, damaging earthquakes. Forecasting the eventual evolution of swarms is currently very challenging but could be aided in the future by new observations and models.

Noto Peninsula↗
USGS2024

Stress-driven recurrence and precursory moment-rate surge in caldera collapse earthquakes

Predicting the recurrence times of earthquakes and understanding the physical processes that immediately precede them are two outstanding problems in seismology. Although geodetic measurements record elastic strain accumulation, most faults have recurrence intervals longer than available measurements. Foreshocks provide the principal observations of processes before mainshocks, but variability between sequences limits generalizations of pre-failure behaviour. Here we analyse seismicity and deformation data for highly characteristic caldera collapse earthquakes from 2018 Kīlauea Volcano (Hawaii, USA), with a mean recurrence interval of 1.4 days. These events provide a unique test of stress-induced earthquake recurrence and document processes preceding mainshocks with magnitude greater than five. We show that recurrence intervals are well predicted by stress histories inferred from near-field deformation measurements and that cycle-averaged seismicity reveals a critical phase, minutes before mainshocks, where earthquakes grew larger and seismic moment rate surged dramatically. The average moment rate in the final 15 minutes (0.7% of the mean cycle duration) was 4.75 times the background, a highly significant change. We infer that as the average stress increased, ruptures were more likely to overcome geometric barriers and grow larger, leading to characteristic, whole-fault ruptures. These findings imply that stress heterogeneity influences both earthquake nucleation and growth, including on potentially hazardous tectonic faults.

Nature Geoscience↗
USGS2024

Subduction intraslab-interface fault interactions in the 2022 Mw 6.4 Ferndale, California earthquake sequence

The Mendocino triple junction, the intersection of the Pacific, North American, and Gorda plates, activates a collection of disparate faults that reconcile Cascadia subduction with San Andreas transform motion. The December 20, 2022, Mw 6.4 Ferndale, California earthquake occurred within this complex zone as strike-slip faulting within the subducting Gorda slab. Here, we analyze the seismic and geodetic signatures of the mainshock and aftershock sequence to illuminate its role within complex tectonic surroundings. We find aftershocks on varied fault structures within the uppermost Gorda slab, yet seismicity on the subduction interface itself was notably absent. Nevertheless, we identify small but coherent postseismic deformation that is well modeled by aseismic slip on this interface, likely triggered by stresses generated at the updip limit of coseismic rupture. This sequence demonstrates the potential for interactions between intra-slab earthquakes and slip on the subduction megathrust, highlighting the need to consider this and other subduction zones as coupled systems of interacting faults.

California↗
USGS2023

Rapid Source Characterization of the 2023 Mw 6.8 Al Haouz, Morocco, Earthquake

The U.S. Geological Survey (USGS) National Earthquake Information Center (NEIC) estimates source characteristics of significant damaging earthquakes, aiming to place events within their seismotectonic framework. Contextualizing the 8 September 2023, M w 6.8 Al Haouz, Morocco, earthquake is challenging, because it occurred in an enigmatic region of active surface faulting, and low seismicity yet produced significant damage and loss of life. Here, we present the rapid earthquake source products produced by the USGS NEIC, describing how the source model was derived using both seismic and geodetic observations. Our analysis indicates that the earthquake was the result of oblique‐reverse faulting in the lower crust on either a steeply north‐dipping fault or a moderately south‐dipping fault. Finite‐slip models using seismic and geodetic data reveal a compact source, with slip occurring at depths of 15–35 km. The causative fault is not apparent, because the rupture did not break the surface, and it is not possible to definitively attribute the earthquake to a known structure. The earthquake centroid depth of 25 km is noteworthy, because it shows slip extending beyond common estimates of seismogenic depth. This earthquake highlights that the seismogenic processes associated with mountain building in this wide plate boundary region are poorly understood.

The Seismological Record↗
USGS2023

Rapid characterization of the February 2023 Kahramanmaraş, Turkey, earthquake sequence

The 6 February 2023 M w "> M w 7.8 Pazarcık and subsequent M w "> M w 7.5 Elbistan earthquakes generated strong ground shaking that resulted in catastrophic human and economic loss across south‐central Türkiye and northwest Syria. The rapid characterization of the earthquakes, including their location, size, fault geometries, and slip kinematics, is critical to estimate the impact of significant seismic events. The U.S. Geological Survey National Earthquake Information Center (NEIC) provides real‐time monitoring of earthquakes globally, including rapid source characterization and impact estimates. Here, we describe the seismic characterization products generated and made available by the NEIC over the two weeks following the start of the earthquake sequence in southeast Türkiye, their evolution, and how they inform our understanding of regional seismotectonics and hazards. The kinematics of rupture for the two earthquakes was complex, involving multiple fault segments. Therefore, incorporating observations from rupture mapping was critical for characterizing these events. Dense local datasets facilitated robust source characterization and impact assessment once these observations were obtained and converted to NEIC product input formats. We discuss how we may improve the timeliness of NEIC products for rapid assessment of future seismic hazards, particularly in the case of complex ruptures.

The Seismic Record↗
USGS2023

Dense geophysical observations reveal a triggered, concurrent multi-fault rupture at the Mendocino Triple Junction

A central question of earthquake science is how far ruptures can jump from one fault to another, because cascading ruptures can increase the shaking of a seismic event. Earthquake science relies on earthquake catalogs and therefore how complex ruptures get documented and cataloged has important implications. Recent investments in geophysical instrumentation allow us to resolve increasingly complex, multi-fault ruptures for even moderate-sized earthquakes. We combine dense seismic and geodetic measurements to reveal an enigmatic rupture in late 2021 at the Mendocino Triple Junction in northern California. We show that rupture was dynamically triggered, yet concurrent, on two distinct faults roughly 30 km apart. Thus, this rupture combines features of complex ruptures usually considered to be single earthquakes, and triggered ruptures considered as multiple earthquakes. This event illustrates that moderate-sized earthquakes can exhibit similar complexity to that more commonly documented for large earthquakes.

California↗
USGS2023

Using corrected and imputed polarity measurements to improve focal mechanisms in a regional earthquake catalog near the Mt. Lewis Fault Zone, California

We utilized relative polarity measurements and machine learning techniques to better resolve focal mechanisms and stress orientations considering a catalog of ∼29,000 relocated earthquakes that occurred during 1984–2021 in the southeastern San Francisco Bay Area. Earthquake focal mechanisms are commonly produced using P wave first motion polarities, which traditionally requires events to be well-recorded across a seismic network with good focal sphere coverage. We adapted recently developed approaches that are less dependent on high signal-to-noise records and exploit similar waveforms to produce relative polarity and amplitude measurements between earthquake pairs. These techniques were previously only applied on localized earthquake sequences, and we further developed these approaches so that they can be utilized for regional catalogs. We validated or corrected manually identified polarities by performing polarity consensuses using earthquake pairs. Missing and unreliable polarity measurements were imputed using iterative random forests, an unsupervised ensemble machine learning method. Relative P and S wave amplitude measurements were made between earthquakes, constraining S / P ratios for low signal-to-noise waveforms. Using these techniques, we were able to reduce focal mechanism uncertainties by an average of ∼13° and produced well-constrained focal mechanisms for ∼6 times as many earthquakes than those produced using only the traditionally derived polarities. We performed stress inversions using the focal mechanisms by grouping the focal mechanism results into a quadtree structure. Our stress results are consistent with previous work, albeit at a higher spatial resolution, and demonstrate these techniques can aid our understanding of fault structures and kinematics in more detail than was previously possible.

California↗
USGS2023

Fracture-mesh faulting in the swarm-like 2020 Maacama sequence revealed by high-precision earthquake detection, location, and focal mechanisms

In August of 2020, an earthquake sequence initiated within the Maacama fault zone in northern California, raising questions about its relationship with the larger-scale fault. To investigate the faulting geometry and its implications for physical processes driving seismicity, we applied an integrated, multi-faceted seismic analysis including waveform-correlation-based event detection, relative relocation, and high-precision focal mechanisms. To determine mechanisms for a large population of very small earthquakes (predominantly M < 1), we combined correlation-derived polarity analysis with a recently developed technique to determine S/P amplitude ratios from single seismic components. Finally, we applied an iterative stress inversion to distinguish between the likely fault and auxiliary planes. This analysis reveals that although the sequence initiated on a right-lateral fault, plausibly a strand of the Maacama Fault, it also activated numerous en echelon left-lateral conjugate faults. Together, these interlocking faults form a fracture mesh, consistent with structures associated elsewhere with fluid-induced faulting and earthquake swarms.

California↗
USGS2023

Magnitude conversion and earthquake recurrence rate models for the central and eastern United States

Development of Seismic Source Characterization (SSC) models, which is an essential part of Probabilistic Seismic Hazard Analyses (PSHA), can help forecast the temporal and spatial distribution of future damaging earthquakes (𝑀 w ≥ 5) in seismically active regions. Because it is impossible to associate all earthquakes with known faults, seismic source models for PSHA often include sources of diffuse seismicity in which future earthquake scenarios are not localized on mapped faults. These sources of diffuse seismicity are referred to as area source zones, distributed seismicity zones, or just source zones. During the early years of PSHA studies, it was assumed that earthquakes in seismotectonic zones have (1) uniform spatial distribution, (2) Poisson temporal distribution, and (3) exponential magnitude distribution (NRC, 2012). In seismically active regions (e.g., the Western United States), where active faults are readily identified, models of the spatial distribution of earthquakes include both the fault source geometries and the distributed seismicity (background) source zones. Source characterization of active faults is complemented by paleoseismic studies with estimates of earthquake magnitudes, dates of occurrences, and slip rates, which provide important information for PSHA studies. In the Central and Eastern United States (CEUS) very few Quaternary-active faults have the requisite information for use in PSHA (i.e., fault geometry and dimensions, event rates or slip rates, etc.), and we lack knowledge about the causative faults for most observed seismicity in the region. As a result, area source zones are frequently used in site-specific PSHA in the CEUS to represent diffuse seismicity that cannot be associated with faults. However, there are examples of active fault sources in the CEUS, such as the Meers fault, the Cheraw fault, and New Madrid region, where individual faults can be characterized. The source characterization models for background seismicity are based, to a large extent, on an assumption that spatial distribution of historical and recorded seismicity will not change substantially for time periods of interest for PSHA (approximately the next 50-100 years for engineered structures). Furthermore, studies such as those by Kafka (2007, 2009) found a correlation between the locations of small- to moderate-magnitude earthquakes and the locations of large-magnitude earthquakes, indicating that we can, with some level of confidence, use the spatial pattern of smaller earthquakes to forecast the future pattern of damaging earthquakes. Within background seismicity zones, the earthquake rate forecast is developed using spatial smoothing of the small to moderate magnitude events in earthquake catalogs. Different methodologies are used for this purpose and can predict varying distributions of seismicity rates. This in turn affects the results of a seismic hazard analysis. The U.S. Geological Survey (USGS) and Nuclear Regulatory Commission (NRC) use different methods for computing spatially smoothed seismicity rates in the CEUS; the USGS uses kernel-based spatial smoothing methods in developing the National Seismic Hazard Model (NSHM), and the method adopted in the Central and Eastern United States Seismic Source Characterization (CEUS-SSC) project is used when evaluating seismic hazard for nuclear power plant siting. These methods are described and the impact on seismic hazard are evaluated in this Research Information Letter (RIL). Another important input to estimating the rate of distributed seismicity is event magnitudes listed in earthquake catalogs. A substantial source of uncertainty in catalogs is the magnitude assigned to a given earthquake. Numerous different magnitude types exist, with each magnitude type computed in a different way. Therefore, for the sake of consistency, both the CEUS-SSC and the USGS NSHM have attempted to assemble a complete catalog with a uniform magnitude determination. To this end, moment magnitude, 𝑀 w , which is a physics-based measurement, has been adopted as the standard. However, 𝑀 w was not computed routinely until the past few decades. To address this issue, the CEUS-SSC conducted extensive analyses to determine conversion equations from which to take a routinely computed network (e.g., 𝑀 L or 𝑚 bLg ) and convert it into 𝑀 w . Another issue with using 𝑀 w is that it becomes increasingly difficult to compute for earthquakes with 𝑀 less than ~4. This study investigates the effects of moment magnitude estimation and spatial smoothing methods on estimation of the earthquake rate forecast and on seismic hazard. We investigate the validity of the magnitude conversion equations and their associated uncertainties by applying them to a case study for induced earthquakes in southern Kansas and northern Oklahoma, and summarize the use of the decay of the seismic coda to estimate 𝑀 w for small earthquakes (𝑀 w < 4. Furthermore, the study documents a comparison and assessment of background seismicity smoothing methods implemented by the USGS for the NSHM and used by the CEUS-SSC for siting nuclear facilities based on probabilistic seismic hazard estimates from multiple source zones in the CEUS and for multiple sites.

central and eastern United States↗
USGS2022

Using machine learning techniques with incomplete polarity datasets to improve earthquake focal mechanism determination

Earthquake focal mechanisms are traditionally produced using P ‐wave first‐motion polarities and commonly require well‐recorded seismicity. A recent approach that is less dependent on high signal‐to‐noise exploits similar waveforms to produce relative polarity measurements between earthquake pairs. Utilizing these relative polarity measurements, it is possible to produce composite focal mechanisms for clusters within microseismic sequences using regional networks. However, missing or low‐confidence polarity measurements still limit our ability to calculate high‐quality composite focal mechanisms. Here, we replaced unreliable polarity measurements with estimates using iterative random forests, an unsupervised ensemble machine learning method. Using the imputed (“replaced”) polarity data, we then categorically clustered the events into families. As a case study, we applied this modified composite mechanism workflow to a multistation template matched catalog of an earthquake swarm that occurred during 2020 near the Maacama fault in northern California. We found that our modified methodology produced higher‐quality earthquake families and improved composite focal mechanisms, with fault‐plane uncertainties <35° for 94% of the families compared with 34% of families using the previous methodology.

Seismological Research Letters↗
USGS2022

S/P amplitude ratios derived from single-component seismograms and their potential use in constraining focal mechanisms for micro-earthquake sequences

Focal mechanisms, which reflect the sense of slip in earthquakes, provide important constraints for understanding crustal tectonics and earthquake source physics, including the interactions among earthquakes during mainshock–aftershock sequences or seismic swarms. Focal mechanisms of small (magnitude ≲3.5) earthquakes are usually determined by first‐motion P ‐wave polarities, sometimes supplemented by the ratio of S ‐wave to P ‐wave amplitudes ( S/P ). However, focal mechanisms of such events can be difficult to reliably constrain, particularly with sparse recording networks or very small magnitude events. Here, we describe a method for deriving S/P amplitude ratios from P/P and S/S amplitude ratios measured on single seismic components between pairs of nearby events, as is often performed during correlation‐based earthquake detection and relocation. These measurements can be transformed into relative S/P amplitude ratios, or they can be combined with a smaller number of traditional S/P amplitude ratios to provide a single‐channel estimation of full S/P ratios, even for low signal‐to‐noise‐ratio events not routinely cataloged and not amenable to traditional S/P ratio processing. This approach has the potential to greatly expand the applicability of S/P amplitude ratios, providing additional constraints for focal mechanisms of small earthquakes, particularly for spatially concentrated seismicity sequences.

California↗
USGS2022

Earthquake-derived seismic velocity changes during the 2018 caldera collapse of Kīlauea volcano

The 2018 Kīlauea caldera collapse produced extraordinary sequences of seismicity and deformation, with 62 episodic collapse events which significantly altered the landscape of the summit region. Despite decades of focused scientific studies at Kīlauea, detailed information about the internal structure of the volcano is limited. Recently developed techniques in seismic interferometry can be used to monitor the internal structure of an active volcano more directly by detecting subtle spatiotemporal changes in seismic wave velocity, but their utility relies on accurate interpretations of the underlying phenomena causing those velocity changes. Here, we retrospectively apply repeating-earthquake-based seismic interferometry to the 2018 Kīlauea eruption sequence. We find that seismic velocities changed over two distinct time scales: a sudden increase followed by a slower decrease in velocity in the hours following each collapse event, and a gradual, long-term decrease in velocity over several weeks that ceased approximately 1 month prior to the end of the eruption. Modeling suggests that short-term changes can be explained by magma reservoir pressurization which specifically closed vertical ring fractures. Long-term changes are related to subsidence of the caldera and likely include the influence of inelastic strain from the formation of new fractures. These observations provide new insights into the evolution of Kīlauea during its progressive collapse and will inform future interpretations for near-real-time monitoring at hazardous volcanoes around the world using similar techniques, especially where a dominant fracture orientation is present.

Hawaii↗
USGS2021

A big problem for small earthquakes: Benchmarking routine magnitudes and conversion relationships with coda-envelope-derived Mw in southern Kansas and northern Oklahoma

Earthquake magnitudes are widely relied upon measures of earthquake size. Although moment magnitude ( ⁠ M w "> M w Mw ⁠ ) has become the established standard for moderate and large earthquakes, difficulty in reliably measuring seismic moments for small (generally M w &lt; 4 "> M w < 4 Mw<4 ⁠ ) earthquakes has meant that magnitudes for these events remain plagued by a patchwork of inconsistent measurement scales. Because of this, magnitudes of small earthquakes and statistics derived from them can be biased. Furthermore, because small earthquakes are much more numerous than large ones, many applications, such as seismic hazard modeling, depend critically on analysis of events characterized by magnitudes other than M w "> M w Mw ⁠ . To assess this problem, we apply coda envelope analysis to reliably determine moment magnitudes for a case study of small earthquakes from northern Oklahoma and southern Kansas. Not surprisingly, we find significant differences among M L "> M L ML ⁠ , m b L g "> m b L g mbLg ⁠ , and M w "> M w Mw for M ∼2–4 earthquakes examined here. More troublingly, we find that relations designed to convert other magnitudes to M w "> M w Mw ⁠ , which are relied upon for important applications such as seismic hazard analysis, often increase rather than decrease this bias for our dataset. In our case study, we find that converted magnitudes can result in a systematic bias sometimes exceeding 0.5 magnitude units, a difference that typically corresponds to a factor of ∼3 in seismicity rate. Moreover, we find a correspondingly large bias in Gutenberg–Richter b ‐values, controlled primarily by inaccurate magnitude scaling in the conversion relationships. Although this study focuses on a relatively small geographic area, we can expect that similar issues exist with varying severity in other regions. Therefore, magnitudes of small earthquakes and their associated statistics, including seismicity rates and b ‐values, should be treated with caution.

Kansas, Oklahoma↗
USGS2021

Identification of low-frequency earthquakes on the San Andreas fault with deep learning

Low-frequency earthquakes are a seismic manifestation of slow fault slip. Their emergent onsets, low amplitudes, and unique frequency characteristics make these events difficult to detect in continuous seismic data. Here, we train a convolutional neural network to detect low-frequency earthquakes near Parkfield, CA using the catalog of Shelly (2017), https://doi.org/10.1002/2017jb014047 as training data. We explore how varying model size and targets influence the performance of the resulting network. Our preferred network has a peak accuracy of 85% and can reliably pick low-frequency earthquake (LFE) S-wave arrival times on single station records. We demonstrate the abilities of the network using data from permanent and temporary stations near Parkfield, and show that it detects new LFEs that are not part of the Shelly (2017), https://doi.org/10.1002/2017jb014047 catalog. Overall, machine-learning approaches show great promise for identifying additional low-frequency earthquake sources. The technique is fast, generalizable, and does not require sources to repeat.

Geophysical Research Letters↗
USGS2021

Geological constraints on the mechanisms of slow earthquakes

The recognition of slow earthquakes in geodetic and seismological data has transformed the understanding of how plate motions are accommodated at major plate boundaries. Slow earthquakes, which slip more slowly than regular earthquakes but faster than plate motion velocities, occur in a range of tectonic and metamorphic settings. They exhibit spatiotemporal associations with large seismic events that indicate a causal relation between modes of slip at different slip rates. Defining the physical controls on slow earthquakes is, therefore, critical for understanding fault and shear zone mechanics. In this Review, we synthesize geological observations of a suite of ancient structures that were active in tectonic settings comparable to where slow earthquakes are observed today. At inferred slow earthquake regions, a range of grain-scale deformation mechanisms accommodated slip at low effective stresses. Material heterogeneity and the geometric complexity of structures that formed at different inferred strain rates are common to faults and shear zones in multiple tectonic environments, and might represent key limiting factors of slow earthquake slip rates. Further geological work is needed to resolve how the spectrum of slow earthquake slip rates can arise from different grain-scale deformation mechanisms and whether there is one universal rate-limiting mechanism that defines slow earthquake slip.

Nature Reviews Earth and Environment↗