Search

Adjoint waveform tomography (AWT) sits at the cutting edge of seismic tomography on local, regional, and global scales. However, the choice in starting model may have a significant impact on the final inversion results. In this paper, we present 3 AWT models of California that are based on different starting models. We chose three models that were inverted at different scales: SPiRaL, a global travel‐time tomography model (Simmons et al., 2021, 10.1093/gji/ggab277), CSEM_NA, a regional adjoint tomography model of North America and the North Atlantic (Krischer et al., 2018, 10.1029/2017JB015289), and WUS256, a regional adjoint tomography model of the western US (Rodgers et al., 2022, https://doi.org/10.1029/2022JB024549). We then inverted three AWT models using the same source and receiver set. We ran each model over three period bands: 30–100 s, 25–100 s, and 20–80 s. Once the iterations were finalized, we used five methods of testing model similarity in both the model and data space. We conclude that the choice of starting model has a minimal impact on long wavelength models if an appropriate multi‐scale inversion approach is used. Plain Language Summary: Seismic tomography uses earthquake records to resolve three‐dimensional models of seismic wavespeeds below Earth's surface. Full waveform inversion seeks to model the physics of wave propagation to create models that can most accurately reproduce the waveform data that are observed from recorded events. In this paper, we sought to study the effect of the choice of starting model on final inversion results in California using AWT, a type of Full waveform inversion technique; to our knowledge, this type of experiment has not been carried out before with real data at this scale. We began with three starting models that were created on different scales with different datasets. We then used five methods to compare the final model results. We focused on comparisons of the visual similarity of the models, the tectonic structures the models resolve, and how well the synthetic waveforms match the observed data. We find that the choice of starting model does not have a large impact when the model domain has azimuthally distributed events and stations over long length‐scales. Key Points: We present three adjoint waveform tomography models of California that each begin with a different starting modelThe three final models are compared using five metrics that span both the model space and data spaceBased on a conservative multi‐scale inversion approach, we conclude that the choice of starting model has minimal effect on the final model [ABSTRACT FROM AUTHOR]

We propose a new approach capable of measuring local seismic anisotropy from 6C (three‐component translation and three‐component rotation) amplitude observations of ambient seismic noise data. Our recent theory demonstrates that the amplitude ratio of 6C cross‐correlation functions (CCFs) enables retrieving the local phase velocity. This differs from conventional velocity extraction methods based on the travel time. Its local sensitivity kernel beneath the 6C seismometer allows us to study anisotropy from azimuth‐dependent CCFs, avoiding path effects. Such point measurements have great potential in planetary exploration, ocean bottom observations, or volcanology. We apply this approach to a small seismic array at Pin˜ $\widetilde{n}$on Flat Observatory (PFO) in southern California, array‐deriving retrieves rotational ground motions from microseismic noise data. The stress‐induced anisotropy is well resolved and compatible with other tomography results, providing constraints on the origin of depth‐dependent seismic anisotropy. Plain Language Summary: In contrast to the well‐known translational displacement of seismic waves, rotation measures the angle of rotation of a point in a medium as it deforms. Traditionally, the travel time of cross‐correlation functions of translational displacement of ambient seismic noise and its corresponding propagation distance is used to calculate the velocity of the subsurface medium. However, our recent theory shows that the amplitude ratio of translational displacement and rotation allows us to reveal the local velocity of the medium beneath the receiver from ambient seismic noise data and to study anisotropy from earthquake events. This paper points out for the first time that the amplitude observations of 6C cross‐correlation functions of ambient seismic noise provide a new approach to measure the azimuth‐dependent velocity in anisotropic media. Key Points: We show a new approach enabling the extraction of local phase velocity from 6C cross‐correlation functions of ambient noise dataAzimuth‐dependent 6C cross‐correlation functions allow us to study local seismic anisotropy and its depth dependenceLocal anisotropy at Pin˜ $\widetilde{n}$on Flat Observatory is compatible with compression stress, providing constraints on stress‐induced anisotropy [ABSTRACT FROM AUTHOR]

Reconstructions of Green's functions by ambient noise interferometry enable the imaging of the Earth's subsurface. Coda waves from reconstructed Green's functions can be utilized to perform time‐dependent imaging of processes that vary substantially at daily to monthly time scales in the crust. Time‐lapse Coda‐Wave Imaging (CWI) can detect tiny changes in seismic velocity with high temporal resolution. While previous studies on CWI have mainly focused on the descriptions of coda waves' propagation, little attention has been paid to choosing effective inversion algorithms that are suitable for CWI. Here we address this issue by developing a pragmatic inversion approach based on Voronoi tessellation with mesh cells adapted to coda‐wave sensitivity kernels. Using seismic stations in Central California, we present both synthetic and real data imaging to demonstrate that this approach stabilizes the inversion, is computationally efficient, and provides spatially adaptive resolution. We further propose a heuristic approach for a quantitative assessment of spatial resolution based on multi‐scale checkerboard tests. Plain Language Summary: Seismologists can extract information about the properties and structures beneath the Earth's surface by calculating the interferometry of background seismic vibrations. In recent years, this idea has been widely employed to monitor the time‐varying underground processes associated with tectonic, volcanic, hydrologic, and anthropogenic activities. Coda waves, the waves that are multiply scattered, are often used in monitoring applications because of their high sensitivity and time resolution, but their spatial paths are complicated. In this paper we introduce a new method for improving the inversion in Coda‐Wave Imaging based on Voronoi tessellation. By using mesh cells adapted to the geometry of seismic stations in Central California, we show that this method helps to stabilize the inversion and provide spatially adaptive resolution. Additionally, we propose a straightforward way to assess quantitatively the resolution of the images obtained through this method. Key Points: We present a pragmatic approach to improving coda‐wave imaging of seismic velocity changes based on Voronoi tessellationThis approach stabilizes the spatial inversion and provides adaptive resolution using seismic stations with irregular geometry in Central CaliforniaWe further propose a heuristic approach for a quantitative assessment of the spatial resolution based on multi‐scale checkerboard tests [ABSTRACT FROM AUTHOR]

I employ an elasticity‐based method to invert a geodetically derived surface velocity field in the western US using for present‐day surface strain rate fields with uncertainties. The method uses distributed body forces in a thin elastic sheet and allows for discontinuities in velocity across creeping faults using the solution for dislocations in a thin elastic plate. I compare the strain rate fields with previously published stress orientations and moment rates from geological slip rate data and previous geodetic studies. Geologic and geodetic moment rates are calculated using slip rate and off‐fault strain rates from the 2023 US National Seismic Hazard Model (NSHM) deformation models. I find that computed total geodetic moment rates are higher than NSHM summed moment rates on faults for all regions of the western US except the highest deforming rate regions including the Western Transverse Ranges and the northern and southern San Andreas Fault (SAF) system in California. Computed geodetic moment rates are comparable to the moment rates derived from the geodetically based NSHM deformation models in all regions. I find systematic differences in orientations of maximum horizontal shortening rate and maximum horizontal compressive stress in the Pacific Northwest region and along much of the SAF system. In the Pacific Northwest, the maximum horizontal stress orientations are rotated counterclockwise 40–90° relative to the maximum horizontal strain rate directions. Along the SAF system, the maximum horizontal stresses are rotated systematically 25–40° clockwise (closer to fault normal) relative to the strain rates. Plain Language Summary: Geodetic measurements of surface velocities can be converted to strain rate which is a measure of the distortion rate of the crust. Strain rate is an important quantity for earthquake studies because it can be related to the rate of elastic stored energy (called moment) available for future earthquakes. In this paper I employ a new method to invert the surface velocity field in the western US for present‐day surface strain rates. I convert the strain rates to crustal moment rates and compare this with moment release rates computed from previous estimates of fault slip rates from geology as well as with model moment rates from the 2023 US National Seismic Hazard Model (NSHM). I find that computed total geodetic moment rates are higher than NSHM model moment rates on faults for many regions of the western US. I also find systematic differences in present‐day strain rates and total stress in the crust indicating that the total stress in the crust is not simply the accrual of present‐day deformation. Key Points: I implement an elasticity‐based inversion of geodetic data for strain rateSystemic differences in orientation of stress and strain rate in Cascadia, San Andreas Fault, and eastern Basin‐RangeThe total geodetic moment rates are higher than geologic moment rates within most regions of the western US [ABSTRACT FROM AUTHOR]

A new time series interferometric synthetic aperture radar method is presented based on a strain model. Traditionally, the interferometric phases are optimized based only on a constant set of statistically homogeneous points throughout the observing period, and thus the decorrelated phase cannot be well recovered and possible incoherent (or coherent) phases are (are not) used to optimize the target phase. In the proposed method, a strain model, representing the geophysical relationship of spatially adjacent points' deformations, is employed to optimize the interferometric phases interferogram‐by‐interferogram. The unwrapped phase time series can be estimated from the wrapped interferograms based on the linear iterative weighted least squares method, which is simple and can effectively suppress the decorrelation noise. Both simulations and analysis of data over The Geysers geothermal field (TG), California are conducted, validating that the proposed method can reduce the decorrelations and simultaneously preserve the deformation signals. The displacement results of TG from 2007–2011 L‐band ascending ALOS‐1 data and 2015–2020 C‐band ascending/descending Sentinel‐1 data reveal that (a) the displacement rate of TG is decreasing from 2007–2011 to 2015–2020, and (b) temporally variable displacement signals in addition to steady displacement can be observed from 2015 to 2020 for both the east‐west and the vertical components. Plain Language Summary: The interferometric synthetic aperture radar (InSAR) technique is capable of accurately measuring ground surface displacements at high spatial resolution and large coverage. This paper proposes a strain‐model based InSAR time series method, here referred to as the SIGMA method. The SIGMA method aims to enhance the capability of InSAR to measure displacements and can achieve higher accuracy and density of InSAR displacement measurements. Both simulations and real‐data experiments are conducted to illuminate the benefits of the SIGMA method in the presence of substantial decorrelation noise. Particularly, the SIGMA method is used to investigate the surface displacements of The Geysers geothermal field, California, during 2007–2011 and 2015–2020. These time periods are complementary to those considered in previous studies, providing an important data set to investigate the long‐term evolution of ground surface deformation, as well as the effects of recent fluid injection and production practices in the geothermal field. The new method should be of value for improved geo‐hazards monitoring. Key Points: A new time series interferometric synthetic aperture radar (InSAR) method based on a strain model and iterative weighted least squares is proposedThe method can effectively suppress decorrelation noise and improves the InSAR accuracy and densityThe surface deformation of The Geysers geothermal field, California is investigated with 2007–2011 ALOS‐1 and 2015–2020 Sentinel‐1 data [ABSTRACT FROM AUTHOR]

The generation of strong earthquakes is a long‐debated problem in seismology, and its importance is increased by the possible implications for earthquake forecasting. It is hypothesized that the earthquake generation processes are anticipated by several phenomena occurring within a nucleation region. These phenomena, also defined as preparatory processes, load stress on the fault leading it to reach a critical state. In this paper, we investigate the seismicity preceding 19 moderate (Mw ≥ 3.5) earthquakes at The Geysers, Northern California, aiming to verify the existence of a preparatory phase before their occurrence. We apply an unsupervised K‐means clustering technique to analyze time series of physics‐related features extracted from catalog information and estimated for events occurred before the mainshocks. Specifically, we study the temporal evolution of the b‐value from the Gutenberg‐Richter (b), the magnitude of completeness (Mc), the fractal dimension (Dc), the inter‐event time (dt), and the moment rate (M˙0 ${\dot{M}}_{0}$). Our analysis shows that out of 19 moderate magnitude events considered, a common preparatory phase for 11 events is clearly identified, plus other five events for which we can guess a preparatory phase but with different characteristics from the previous ones. The latter result confirms that even within the same tectonic context different possible activation behaviors may exist. The duration of the preparatory process ranges between about 16 hr and 4 days. We observe that also for the retrieved preparatory process a decrease in b, Mc, and Dc, and an increase of M˙0 ${\dot{M}}_{0}$. Finally, we show a clear correlation between events with a preparation phase and the location of injection's wells, suggesting an important role of fluids in the preparatory process. Plain Language Summary: We investigate the preparatory phase of moderate magnitude‐induced earthquakes at The Geysers geothermal field in California by studying the spatiotemporal evolution and dynamic properties of small magnitude events and using an unsupervised machine learning approach. To this aim, we rely on features extracted from seismic catalog information that are used for a K‐means clustering analysis. Our results highlight changes in the seismicity characteristics before the moderate‐induced earthquakes. We find that most of the analyzed target earthquakes present a preparatory phase, and that most of the cases the latter presents common characteristics in their key features. We show that the seismicity clusters in space and time before moderate events also becoming more energetic. We estimate a duration for the detected preparatory phases that ranges from 16 hr to 4 days. Finally, we show that the presence of a preparatory phase is correlated to the proximity of injection's wells, suggesting a significant role of fluids in the earthquake's nucleation process. Key Points: Moderate‐induced earthquakes at The Geysers can be preceded by a preparatory phase detectable using K‐means clustering on catalog featuresThe detected preparatory phase has common characteristics presenting negative trends of b‐value, Mc, and Dc and a positive trend of M˙0 ${\dot{M}}_{0}$The earthquakes with a preparatory phase are located nearer to the injection's wells in the area than the events without it [ABSTRACT FROM AUTHOR]

Earthquakes can be dynamically triggered by the passing waves of other distant events. The frequent occurrence of dynamic triggering offers tangible hope in revealing earthquake nucleation processes. However, the physical mechanisms behind earthquake dynamic triggering have remained unclear, and contributions of competing hypotheses are challenging to isolate with individual case studies. To gain a systematic understanding of the spatiotemporal patterns of dynamic triggering, we investigate the phenomenon in southern California from 2008 to 2017. We use the Quake Template Matching catalog and an approach that does not assume an earthquake occurrence distribution. We develop a new set of statistics to examine the significance of seismicity‐rate changes as well as moment‐release changes. Our results show that up to 70% of 1,388 global M ≥ 6 events may have triggered earthquakes in southern California. The triggered seismicity often occurred several hours after the passing seismic waves. The Salton Sea Geothermal Field, San Jacinto fault, and Coso Geothermal Field are particularly prone to triggering. Although adjacent fault segments can be triggered by the same earthquakes, the majority of triggered earthquakes seem to be uncorrelated, suggesting that the process is primarily governed by local conditions. Further, the occurrence of dynamic triggering does not seem to correlate with ground motion (e.g., peak ground velocity) at the triggered sites. These observations indicate that nonlinear processes may have primarily regulated the dynamic triggering cases. Plain Language Summary: Earthquakes interact with each other, such as mainshocks triggering nearby aftershocks. Earthquake dynamic triggering is a type of interaction where seismic waves from an earthquake trigger other earthquakes beyond several fault lengths, and sometimes, up to thousands of kilometers away. Triggered earthquakes may occur upon the arrival of the seismic waves but may also be delayed hours after the wave passage, suggesting the involvement of time‐dependent processes. Identifying delayed cases relies on robust measures of seismicity‐rate changes. Here we present a new method that can identify triggering cases without many assumptions. We find that earthquakes in southern California are frequently triggered by distant earthquakes around the globe, and the triggered earthquakes tend to cluster in space and time. We also find that the triggering incidences do not seem to correlate with the seismic wave characteristics of the distant earthquakes. Our findings suggest that dynamically triggered earthquakes in southern California are likely caused by time‐dependent, local complex processes. Key Points: Earthquake dynamic triggering is ubiquitous in southern CaliforniaTriggered earthquakes are frequently associated with significant moment‐release anomalies and are likely controlled by local processesOur procedure does not assume that seismicity follows a Poissonian distribution when identifying dynamic triggering [ABSTRACT FROM AUTHOR]

The rate of fault zone restrengthening between earthquakes can be influenced by both frictional and cohesive healing processes. Friction is dependent on effective normal stress while cohesion is independent of normal stress, potentially explaining—in part—the lack of depth dependence of earthquake stress drops. Although amenable to laboratory testing, few studies have systematically addressed the normal stress dependence of restrengthening rate. This is partially due to difficulty in separating relative contributions of friction and cohesion in recovery of fault strength. We present results from a series of slide‐hold‐slide tests on thin layers (≤10 휇m) of ultrafine quartz gouge that develop during shearing of initially bare‐surface quartzite. Tests were conducted at 10 MPa constant pore pressure, 20–200 MPa constant effective normal stress, and temperatures of 22°–200°C. Restrengthening, defined as the difference between peak shear stress measured after resumption of sliding and steady‐state sliding shear stress, increases with the log of hold duration. The 200°C healing rate, 0.014 per e‐fold increase in time, is comparable to that determined from seismological observations along the Calaveras Fault, California. Construction of Mohr‐Coulomb failure envelopes shows that changes in cohesion are small (<1 MPa) and independent of hold durations to 105 s, indicating that the increased strength is due to changes in the friction coefficient. These experimental results are inconsistent with the hypothesis that cohesive healing explains the depth independence of earthquake stress drop, but higher temperatures, longer time‐scales, and more complex mineralogy could facilitate cohesive healing in natural fault systems. Plain Language Summary: Faults recover strength, or heal, following earthquakes. This occurs due to processes that can increase the frictional or cohesive strength of the fault surface, but the relative contributions of friction and cohesion to fault strength are difficult to constrain. By conducting a series of slide‐hold‐slide friction experiments, where a simulated fault in quartzite alternates between periods of shear displacement and quasi‐stationary contact, at a range of temperatures and effective normal stresses, we can independently determine the magnitude of frictional and cohesive healing. The strength of the simulated faults increases with time that the surfaces are held in contact. We show that cohesive healing is negligible and the rate of frictional healing is comparable to earthquake observations. This study has implications for our understanding of earthquake recurrence and fault stability. Key Points: Strength recovery is similar at 22° and 200°C but minimized at 100°C, indicating the mechanism is not a simple Arrhenius‐type processCohesive healing in these experiments is negligible and strength recovery is predominately due to frictional healing processesExperimentally determined frictional healing rate is comparable to seismological observations [ABSTRACT FROM AUTHOR]

To constrain fault processes and hazard, fault slip rates may be extrapolated over different fault lengths or time intervals. Here, we investigate slip rates for the Cucamonga Fault (CF). The CF is located at the junction of the Transverse Range fault system with the San Andreas and San Jacinto Faults, and it is hypothesized to connect with these faults, promoting the propagation of large, multi‐fault earthquakes. Previous work has shown that CF displacements on late Quaternary alluvial fan surfaces are highly variable along strike. We present two new 10Be surface exposure ages from depth profiles on the alluvial fans. Slip rates are consistent with a rate of 1.4 ± 0.3 m/kyr over time intervals of ∼20, ∼30, and ∼40 kyr. If the CF participates in multi‐fault ruptures, then these earthquakes occur either rarely or with sufficient regularity to maintain apparently steady rates over multiple intervals. We also explore along‐strike fault displacement variability using a calibrated morphological model. The model successfully reproduces scarp profiles and indicates that fault displacement variability can be explained in part by scarp age but not uplift rate. We infer that both erosion by ephemeral gullying and distributed deformation contribute to fault displacement variability, although both are difficult to detect confidently without excavations across the scarp. These investigations show that better characterization of cumulative‐slip variability along strike may improve accuracy and precision of slip rates. Slip rates that do not consider epistemic uncertainties may not be suitable for extrapolation over longer fault sections. Plain Language Summary: Geologic slip rates describe total fault displacement over a specific time interval and are important for earthquake hazard assessment. On the Cucamonga Fault, a component of hazard is the potential to participate in multi‐fault earthquakes, which can be particularly large and damaging. We measure the displacements that have accumulated during three overlapping intervals of time. The slip rate appears to be constant, but our resolution is limited by large uncertainties, especially uncertainties arising from variable displacement along the length of the fault. We model fault scarp evolution to better understand the sources for this variability. Model results indicate that alteration of the scarp by erosion is important, but other sources also contribute that are difficult to identify from topographic analysis alone. Our slip‐rate data suggest that large, multi‐fault earthquakes are either very rare or frequent enough to give the appearance of steady slip rates through time. Key Points: Cucamonga Fault slip rates are consistent with a single rate of 1.4 ± 0.3 m/kyr at intervals of ∼20, 30, and 40 kyrIf the fault participates in multi‐fault earthquakes, then they occur either rarely or frequently enough to maintain apparently steady ratesSpatial slip‐rate variability is difficult to mitigate; local measurements should consider epistemic uncertainty on dip‐slip faults [ABSTRACT FROM AUTHOR]

We use a dense seismic array of 1,108 vertical‐component geophones within a 600‐m footprint to detect thousands of small earthquakes near an active strand of the San Jacinto fault zone in southern California during a 26‐day period. We first correct site effects using multichannel cross‐correlations of the P‐waves of 256 cataloged earthquakes, and then perform beamforming analysis on the continuous waveforms in a slowness range from −0.4 to 0.4 s/km in both the east and north directions. At each time step, we identify the beam slowness with maximum amplitude and apply a picking algorithm to identify 13,408 events. These detections include over 55.6% of the events in the Quake Template Matching (QTM) catalog for all of southern California during the same time period and 70% of those within 100 km of the array. In addition, we detect over 10,000 new events, not in the QTM catalog. Many of these events can also be seen in records from nearby borehole seismic stations. Measured slownesses for the catalog and newly‐discovered events group into clusters that can be associated with QTM earthquake locations, but with slowness values considerably distorted from predictions based on a 1‐D velocity model, presumably owing to strong velocity heterogeneity near the San Jacinto Fault. Amplitudes of the detected events obey a Gutenberg‐Richter distribution with a b‐value close to one. Foreshocks are common among these detected events, increasing in rate before mainshocks following an inverse Omori's law. Plain Language Summary: We show that an array of over 1,000 seismometers located on a small patch of the San Jacinto Fault in southern California can detect many times more earthquakes than those in standard earthquake catalogs. Although accurate locations cannot be obtained for most of these newly identified events, their amplitudes indicate they obey similar scaling with size as larger magnitude earthquakes. Many of the larger events are preceded by foreshocks and the foreshock rate increases in the time leading up to the mainshocks. Our results suggest that future deployments of small‐aperture seismometer arrays could be used to complement existing seismic networks to probe earthquake activity in great detail. Key Points: We detect over 13,000 seismic events during 26 days in 2014 using a dense nodal array on the San Jacinto fault zone in southern CaliforniaMost of the detections are newly discovered earthquakes not in existing catalogsForeshocks are common among these earthquakes, increasing in the rate before their mainshocks following an inverse Omori's law [ABSTRACT FROM AUTHOR]

Establishing a baseline of ongoing secular velocity variations at the subsurface can improve the accuracy of detecting and interpreting short‐term velocity changes, and advance the understanding of observed seismic motions and the behavior of subsurface materials. Toward these goals, we develop and apply a deconvolved autocorrelation (DA) method to estimate regional daily and seasonal changes of seismic velocities in southern California. The DA method combines advantages of traditional autocorrelation and Horizontal‐to‐Vertical Spectral Ratio, and is used to analyze over 10 years of data recorded by 50 stations. The results indicate widespread daily and seasonal changes of up to 10% and 4%, respectively, in the top tens of meters of the crust. The thickness of the surface layer, distance from the coast, and topographic variations are important factors controlling the amplitudes of the resolved velocity variations. The results suggest that changes of soil moisture and thermoelastic strain are likely dominant factors affecting the daily and seasonal variations, respectively. The developed DA method can improve the accuracy and robustness of estimated changes of subsurface materials at other locations. Plain Language Summary: Temporal changes of subsurface seismic velocities can be important for many topics including understanding observed seismic motions and interpretation of crustal processes and structures. To accurately estimate temporal changes of subsurface velocities, we develop a new method and apply it to data recorded at 50 sites in southern California. The results reveal widespread daily and seasonal velocity variations with amplitudes up to 10% and 4% in the top tens of meters. We compare the amplitudes of the velocity variations with various attributes including thickness of the surface layer, surface geology, topography, geographical location, and vegetation. The amplitudes are generally larger in areas with reduced thickness of the surface layer. In addition, the amplitudes of daily variations increase with proximity to the coast, while the amplitudes of seasonal variations increase in areas with larger local topographic variations. The results suggest that changes of soil moisture and thermoelastic strain may govern the amplitudes of the daily and seasonal velocity variations at many sites. Key Points: A newly developed deconvolved autocorrelation method can estimate accurately and robustly subsurface variations of seismic velocitiesAnalysis of seismic data in southern California indicates widespread daily and seasonal velocity variations in the top tens of metersChanges of soil moisture and thermoelastic strain are likely key factors affecting the daily and seasonal variations, respectively [ABSTRACT FROM AUTHOR]

Key factors controlling earthquake ruptures include fault geometry, continuity, and seismic velocity structure around the fault. We present a novel tool that better informs deep bimaterial fault geometry embedded in distributed damage and seismicity, associated velocity contrasts across the fault, and their correlations with surface complexities. The method employs fault zone head and direct body waves and is applied to recordings from five spatiotemporally different seismic arrays along the complex San Jacinto fault zone (SJFZ) in southern California. We detect and distinguish these signals based on instantaneous phase coherence and relative energy in a cascading manner from one scale array to another. The analysis reveals a >70‐km long continuous bimaterial interface within the SJFZ with several deep northeast dipping fault segments. The northern SJFZ, for instance, locates ∼7 km northeast of its surface expression at 18‐km depth. P‐wave velocity contrasts range from near 0% to >15%, consistent with other bimaterial faults, and differ by a few % depending on fault‐array azimuth, implying directional‐dependent velocity contrasts. S‐wave head waves and velocity contrasts are also imaged for the first time at the southern SJFZ, averaging to 2.9% in agreement with tomography results. The imaged geometry and continuity suggest the SJFZ initiated along remnant tectonic structures and translates to a rupture potential of M > 7.2, i.e., the sizes of its largest paleo‐earthquakes. The P and S contrasts, and their ratios, have important implications for earthquake rupture speed, mode, directivity, and frictional heating along the SJFZ and other major faults globally. Plain Language Summary: Large earthquakes occur on faults that are structurally complex, leading to earthquakes that are themselves structurally and dynamically complex. More accurate information on fault complexities is key to quantifying the potential size and shaking from large earthquakes along them. Here, we introduce a novel tool that resolves the deep geometry, continuity, and contrast in seismic velocity along major transform fault systems using the complex San Jacinto fault zone (SJFZ) in southern California as an exemplary case. It employs unique signals that travel along, and illuminate, continuous faults contrasting rock volumes of different seismic properties. With this tool, we reveal a >70‐km long continuous fault within the SJFZ that is nonsimple and nonplanar, curving laterally, and in depth. For example, a deep northern section of the SJFZ dips extensively toward the northeast. The seismic velocity contrast across the entire fault varies from 0% to >15% and depends on the angle of observation from the surface. This long continuous fault is capable of an M > 7.2 earthquake and likely initiated along older structures that predate the SJFZ. The observed geometries and velocity contrasts have implications for the direction, speed, distribution, and other key defining features of large SJFZ earthquakes. Key Points: Image the San Jacinto fault zone with direct body and fault zone head waves recorded by several dense arrays along the faultReveal a ∼72‐km long continuous bimaterial interface within the core fault zone with variable geometry and P‐ and S‐wave velocity contrastsEstimated bimaterial properties have important implications for earthquake rupture properties along the San Jacinto and other faults [ABSTRACT FROM AUTHOR]

Seismic velocities in rocks are highly sensitive to changes in permanent deformation and fluid content. The temporal variation of seismic velocity during the preparation phase of earthquakes has been well documented in laboratories but rarely observed in nature. It has been recently found that some anthropogenic, high‐frequency (>1 Hz) seismic sources are powerful enough to generate body waves that travel down to a few kilometers and can be used to monitor fault zones at seismogenic depth. Anthropogenic seismic sources typically have fixed spatial distribution and provide new perspectives for velocity monitoring. In this work, we propose a systematic workflow to seek such powerful seismic sources in a rapid and straightforward manner. We tackle the problem from a statistical point of view, considering that persistent, powerful seismic sources yield highly coherent correlation functions (CFs) between pairs of seismic sensors. The algorithm is tested in California and Japan. Multiple sites close to fault zones show high‐frequency CFs stable for an extended period of time. These findings have great potential for monitoring fault zones, including the San Jacinto Fault and the Ridgecrest area in Southern California, Napa in Northern California, and faults in central Japan. However, extra steps, such as beamforming or polarization analysis, are required to determine the dominant seismic sources and study the source characteristics, which are crucial to interpreting the velocity monitoring results. Train tremors identified by the present approach have been successfully used for seismic velocity monitoring of the San Jacinto Fault in previous studies. Plain Language Summary: Temporal variations of seismic velocities in the vicinity of fault zones are informative for understanding the earthquake preparation phase, but the lack of appropriate methodology and observations hinders actual applications. Human activities generate strong seismic energy that is suitable for monitoring velocity changes at depth. We design an algorithm to search for persistent anthropogenic seismic sources for fault‐zone velocity monitoring, and the workflow can be generalized to other applications. The feasibility of our algorithm is tested on three different data sets in California and Japan. We find stable sources close to several important fault zones that have hosted or have the potential to host large earthquakes. This study paves the way for future works related to fault‐zone velocity monitoring. Key Points: Repeating anthropogenic seismic sources are useful for seismic velocity monitoringWe use the stability of correlation functions over time as a proxy to search for persistent seismic sourcesCase studies in California and Japan demonstrate the potential for fault zone monitoring [ABSTRACT FROM AUTHOR]

Swarms are bursts of earthquakes without an obvious mainshock. Some have been observed to be associated with transient aseismic fault slip, while others are thought to be related to fluids. However, the association is rarely quantitative due to insufficient data quality. We use high‐quality GPS/GNSS, InSAR, and relocated seismicity to study a swarm of >2,000 earthquakes which occurred between 30 September and 6 October 2020, near Westmorland, California. Using 5 min sampled Global Positioning System (GPS) supplemented with InSAR, we document a spontaneous shallow Mw 5.2 slow slip event that preceded the swarm by 2–15 hr. The earthquakes in the early phase were predominantly non‐interacting and driven primarily by the slow slip event resulting in a nonlinear expansion. A stress‐driven model based on the rate‐and‐state friction successfully explains the overall spatial and temporal evolution of earthquakes, including the time lag between the onset of the slow slip event and the swarm. Later, a distinct back front and a square root of time expansion of clustered seismicity on en‐echelon fault structures suggest that fluids helped sustain the swarm. Static stress triggering analysis using Coulomb stress and statistics of interevent times suggest that 45%–65% of seismicity was driven by the slow slip event, 10%–35% by inter‐earthquake interactions, and 10%–30% by fluids. Our model also provides constraints on the friction parameter and the pore pressure and suggests that this swarm behaved like an aftershock sequence but with the mainshock replaced by the slow slip event. Plain Language Summary: Over 2,000 earthquakes were recorded near Westmorland, California, between 30 September and 6 October 2020. Such an increased level of earthquake activity is quite common in this region, though the causes are not well‐understood. Using available seismological data, satellite imagery, and ground‐based GPS/GNSS, we detect ground deformations a few hours before the increased earthquake activity. We map these ground deformations to motion along the faults at a depth shallower than 5 km. We show that this silent fault slip drove the earthquakes at a greater depth of 5–8 km. The overall spatial and temporal evolution of the earthquakes can be largely predicted based on the stress changes imparted by this silent fault slip. Statistical analysis of earthquake activity and the expansion of the zone with no earthquake further suggest that fluids played a significant role in sustaining the earthquake sequence. Though we can explain earthquake activity after the initial ground deformations, what causes the initial fault motion at depth and thus these ground deformations in the first place remains an open question. Key Points: Five minutes sampled Global Positioning System supplemented with InSAR resolves a shallow slow slip event, which preceded the swarm by 2–15 hrSeismicity was driven in the early stage by the slow slip event with nonlinear expansion and later by fluid with propagating back frontA stress‐driven model explains the overall evolution of seismicity and provides constraints on friction and fluid pressure [ABSTRACT FROM AUTHOR]

Magnetic field changes as earthquake precursors have been the subject of numerous studies and some controversy. Infrequent large earthquakes and sparse magnetometer coverage along fault zones complicate statistical analysis. We present an analysis of ground‐based magnetic time‐series measurements before 19 earthquakes ≥M4.5 in California drawing from over 330,000 site‐days of measurement spanning a decade. To perform a fair existential test for electromagnetic antecedents we applied a pre‐specified statistical analysis with two key ideas. First, we combine signals from nearby (≤40 km) sites via spectral cross‐power, and then look for large spikes in frequency domain (0.016–25 Hz). The former is only possible with a dense set of sites running over a long period of time. In this statistical case‐control study we used the machine learning concept of rigorously separated train and test sets of earthquakes which were generated via a rule‐based query of the USGS earthquake catalog. Before each declustered earthquake, we constructed one period 24–72 hr before (the "precursor" or "p‐period") and a series of seven equally‐sized preceding periods ("quiescent" or "q‐periods"). We distilled the data in each period to a frequency‐dependent feature—the 98th percentile of spectral cross power. We trained a model based on Linear Discriminant Analysis and applied the discriminator to the test set revealing a modest effect in the days leading up to an earthquake. While the observed effect size is not directly useful for earthquake prediction (long a scientific goal), it suggests a relationship which should be further investigated for a physical link. Plain Language Summary: We identified changes in the magnetic field near intermediate‐large earthquakes in California in the days before the earthquakes happened. The statistical signal is of modest size, which means that we can not directly provide a prediction that can be used to alert the public. This study provides evidence that there is a physical change that can be observed in the days before an earthquake, but further scientific study is needed to understand this process. Key Points: Frequency domain analysis of ground‐based magnetometer data shows a modest change in days leading up to intermediate‐large (M ≥ 4.5) earthquakesOne novel part of the analysis is the use of cross‐power signals, which combines the signals from instruments separated by tens of kmA supplementary analysis of the data with first order global geomagnetic effects subtracted increased the measured effect size significantly [ABSTRACT FROM AUTHOR]

We estimate Brune‐type stress drops for over 70,000 southern California M 1.5–4 earthquakes from 1996 to 2019 using a P‐wave spectral decomposition approach. Based on our recent work documenting hard‐to‐resolve trade‐offs between absolute stress drop, stress drop scaling with moment, high‐frequency falloff rate, and empirical corrections for path and attenuation terms, we adopt a new approach in which the average corner frequency of the smallest earthquakes within a short distance from each target event is fixed to a constant value. This removes any true coherent spatial variations in stress drops among the smallest events but ensures that any spatial variations seen in larger event stress drops are real and not an artifact of inaccurate path corrections. Applying this approach across southern California, we document spatial variations in stress drop that agree with previous work, such as lower‐than‐average stress drops in the Salton Trough, as well as small‐scale stress drop variations along many faults and aftershock sequences. We observe an apparent increase in Brune‐type stress drop with moment for M 3–4 earthquakes, but their spectra can be fit equally well with self‐similar models with a high‐frequency falloff rate shallower than f−2. Plain Language Summary: Earthquakes vary widely in their properties. One key parameter used by seismologists to describe earthquakes is their stress drop, a measure of the forces unleashed during fault rupture. High‐stress drop earthquakes are particularly dangerous because they generate stronger ground motions than low‐stress drop events of the same size. Previous studies have suggested that parts of southern California are more prone to high‐stress drop earthquakes than other regions. However, stress drop estimates have large uncertainties and studies do not always agree very well with each other, making it difficult to be sure that these observed spatial variations are real. Here, we introduce a new approach for earthquake stress drop estimation that is more stable than previous methods for assessing differences in average levels of stress drop across southern California. Applying our method to over 70,000 earthquakes from 1996 to 2019, we find that there are indeed substantial regional variations in stress drop with the Imperial Valley and Salton Sea region near the US‐Mexico border having particularly low earthquake stress drops. Key Points: We estimate Brune‐type stress drops for over 70,000 southern California M 1.5–4 earthquakes using P‐wave spectral decompositionWe stabilize the estimation of empirical attenuation corrections by forcing small‐earthquake corner frequencies to a fixed valueWe observe large‐scale variations in median stress drop across the region and deviations from a self‐similar Brune model for M > 3 events [ABSTRACT FROM AUTHOR]

The magma plumbing in the lower crust beneath the Coso volcanic field (CVF) remains controversial, largely because of the absence of high‐resolution lower crustal velocity models. For the first time, we develop a high‐resolution crustal P‐wave velocity model for the Coso‐Ridgecrest region by jointly inverting 137,992 first P and 8,636 PmP travel‐time data using an eikonal equation‐based seismic reflection tomography method. More than half of the PmP travel times are picked from earthquakes after the 2019 Mw 7.1 Ridgecrest earthquake. Such abundant PmP travel times significantly improve the resolution of the lower crust. Our final velocity model reveals a prominent low‐velocity body sitting right beneath the CVF at 5–20 km depths, which we interpret as a rhyolite magma reservoir that supplies heat flux to the hot springs and also feeds the volcanic activities at Coso. We find that the upper‐middle crustal low‐velocity body dips southwards into the lower crust, extending to regions beneath the Indian Wells Valley and the Garlock Fault at depth greater than 20 km. We ascribe the lower crustal low‐velocity body (more than 4% Vp reduction) to a basaltic magma reservoir that connects the melts in the uppermost mantle with the eruptible rhyolitic reservoir at shallower depths. The basaltic magma reservoir constitutes an important part of a continuous N‐S elongated crustal magma plumbing system beneath the CVF, formed as a combined result of local extension, faulting, and stress distribution. Plain Language Summary: The deep origin of the Coso volcanic field (CVF) in the lower crust is not fully understood. To well constrain the deep origin of magmas, we use a newly developed seismic reflection tomography method to image the entire crustal P‐wave velocity structure beneath the CVF by incorporating the direct P and the Moho reflected waves. Tomographic results reveal a significant low‐velocity anomaly right beneath the CVF in the upper‐middle crust, which dips southwards into the lower crust and extends to the Indian Wells Valley and the Garlock Fault. The low‐velocity anomaly in the upper‐middle crust is interpreted as a rhyolite magma reservoir (contains mostly crystallized silicic magma) directly supplying hot springs at or near the surface, whereas the low‐velocity anomaly in the lower crust is interpreted as a basaltic magma reservoir (contains fewer silica contents) that constitutes a continuous magma plumbing system mobilizing melts from the upper mantle along N‐S direction into the more eruptible shallower Coso reservoir. Key Points: We obtain a new Vp model of the whole crust beneath the Coso volcanic field by inverting the first and Moho reflected P‐wave arrivalsOur model shows a pervasive low‐velocity anomaly in the upper‐middle crust, which dips southwards beneath the Coso volcanic fieldLower crust low Vp may represent a basaltic magma reservoir that ascends northwards and differentiates into a shallower rhyolite reservoir [ABSTRACT FROM AUTHOR]

The Parkfield segment of the San Andreas Fault (SAF) sits at the transition between the locked Cholame segment to the South and the SAF creeping segment to the North. The Parkfield segment hosts regular ∼Mw $\sim {\boldsymbol{M}}_{\boldsymbol{w}}$6 earthquakes followed by postseismic deformation. Recent studies based on geodetic data have highlighted spatial and temporal variations of aseismic slip rate in addition to postseismic slip along this section of the fault. We combine Global Navigation Satellite Systems (GNSS) and seismicity data over the 2006–2018 period to detail a comprehensive picture of transient slip events. We produce a catalog of relocated seismicity and repeating earthquakes. We use a variational Bayesian independent component analysis decomposition on GNSS data to separate geodetic deformation due to non‐tectonic sources from signals of potential tectonic origin. We then reconstruct the temporal evolution of fault slip and detect potential slip transients. Those events, determined as mostly aseismic with the exception of one related to a Mw ${\boldsymbol{M}}_{\boldsymbol{w}}$4.8 earthquake, occur more frequently during the 2004 Mw ${\boldsymbol{M}}_{\boldsymbol{w}}$6 post‐seismic period than during the subsequent inter‐seismic phase. Our study illustrates the rich dynamics of seismic and aseismic slip during both post‐ and inter‐seismic periods along active faults. Plain Language Summary: Faults can slip abruptly, generating earthquakes and seismic waves, or slowly and aseismically. Such a slow motion has now been observed along multiple active faults and is recognized as one of the important processes influencing the Earth's crust stress field. Aseismic slip interacts with earthquakes, and monitoring its time evolution on faults help us in better modeling the seismic cycle. Here we focus our attention on the Parkfield fault segment, sitting along the San Andreas Fault in California in between a section that generated a large earthquake (M7.9) in 1857 and a 150‐km‐long central creeping section that slips aseismically. We dig in noisy time series of surface displacements measured by GNSS (Global Navigation Satellite System) to detect and model small variations of aseismic slip rate (10–100 mm/year) along the fault. We find multiple candidate slow slip events during the period following the 2004 Parkfield earthquake, confirming and expanding previous independent observations. While aseismic slip following earthquakes was first thought to be a steady release of tectonic stress, we show that, in Parkfield, it is evolving with sudden jumps and slowdown. Key Points: Independent Component Analysis of GPS time series highlights slip transients along the San Andreas Fault after the 2004 M6 Parkfield eventWe document the spatio‐temporal distribution of seismic and aseismic slip historySlow Slip Events are more frequent during the 2004 M6 post‐seismic period than the subsequent interseismic period [ABSTRACT FROM AUTHOR]

Water injection and Enhanced Geothermal System (EGS) technologies have been used to exploit heat resources from geothermal reservoirs. Detecting spatial and temporal changes in reservoir physical properties is important for monitoring reservoir condition changes due to water injection and EGS. Here, we determine high‐resolution models of the temporal changes in the three‐dimensional P wave velocity and attenuation (Vp and Qp) structures between the years 2005 and 2011 in the northwestern part of The Geysers geothermal field, California, using double‐difference seismic velocity and attenuation tomography. The northwest Geysers has a shallow normal temperature reservoir (NTR) underlain by a high temperature reservoir (HTR) that has substantial underutilized heat resources but may be more fully utilized in the future through EGS. In the southeastern part of the northwest Geysers, however, EGS has been successfully but unintentionally applied for at least 50 years because the waters injected into the NTR have been flowing into the HTR. Our models are well resolved in this area and show that the NTR and HTR have different seismic responses (seismicity, Vp, and Qp) to water injection, which can be explained by the injection‐induced differences in fracturing and saturation that are likely related to their geological properties. Our results indicate that the joint analysis of changes in seismicity, velocity, and attenuation is valuable for characterizing changes in reservoir fracturing and saturation conditions. Our results suggest that high‐permeability zones and/or pre‐existing permeable fault zones are important for the success of EGS at The Geysers and potentially other geothermal systems. Plain Language Summary: To deal with the declining steam pressure and fluid content, water injection is a commonly employed technology to promote the production of geothermal resources from the naturally permeable geothermal reservoirs. To exploit heat energy from the hot, dry but impermeable rocks, water injection technology can be used to create permeability and increase fluid saturation. Such a man‐made reservoir is called Enhanced Geothermal System (EGS). Monitoring changes in material properties of reservoirs due to water injection and EGS is important for evaluating and adjusting the production and injection strategies. Here, we use seismic tomography to image temporal changes in the three‐dimensional subsurface seismic velocity and attenuation structures of the northwestern part of The Geysers geothermal field, California. The northwest Geysers has a normal temperature reservoir (NTR) at ∼1–2‐km depths underlain by a high temperature reservoir (HTR). The southeastern part of the northwest Geysers has been exploited for >50 years with unintentionally created but successful EGS. Our joint analysis of changes in seismicity and seismic velocity and attenuation clearly reveal different injection‐induced changes in fracturing and saturation conditions in the NTR and HTR in this area that are likely related to their geological properties. Key Points: High‐resolution imaging of temporal changes in Vp and Qp reveals detailed injection‐induced changes in fracturing and saturation conditionsDecreased Vp and Qp and diffuse seismicity suggest diffuse fluid flow, increased fracturing, and increased but still partial saturationDecreased Vp, increased Qp, and concentrated seismicity suggest confined fluid flow, increased fracturing, and full saturation [ABSTRACT FROM AUTHOR]

Recent progresses in artificial intelligence and machine learning make it possible to automatically identify seismic phases from exponentially growing seismic data. Despite some exciting successes in automatic picking of the first P‐ and S‐wave arrivals, auto‐identification of later seismic phases such as the Moho‐reflected PmP waves remains a significant challenge in matching the performance of experienced analysts. The main difficulty of machine‐identifying PmP waves is that the identifiable PmP waves are rare, making the problem of identifying the PmP waves from a massive seismic database inherently unbalanced. In this work, by utilizing a high‐quality PmP data set (10,192 manual picks) in southern California, we develop PmPNet, a deep‐neural‐network‐based algorithm to automatically identify PmP waves efficiently; by doing so, we accelerate the process of identifying the PmP waves. PmPNet applies similar techniques in the machine learning community to address the unbalancement of PmP datasets. The architecture of PmPNet is a residual neural network (ResNet)‐autoencoder with additional predictor block, where encoder, decoder, and predictor are equipped with ResNet connection. We conduct systematic research with field data, concluding that PmPNet can efficiently achieve high precision and high recall simultaneously to automatically identify PmP waves from a massive seismic database. Applying the pre‐trained PmPNet to the seismic database from January 1990 to December 1999 in southern California, we obtain nearly twice more PmP picks than the original PmP data set, providing valuable data for other studies such as mapping the topography of the Moho discontinuity and imaging the lower crust structures of southern California. Plain Language Summary: The success of machine learning in computer sciences, medical sciences, and many other fields has accelerated the implementation and development of machine learning techniques in seismology, making it possible to automatically identify seismic phases from the exponentially growing seismic data. At present, the auto‐identification of later seismic phases, such as the Moho‐reflected PmP waves, remains a significant challenge in matching the performance of experienced analysts. The main difficulty lies in the rare identifiable PmP waves, which makes the identification problem inherently unbalanced. In this work, by utilizing a high‐quality PmP data set in southern California, we develop a deep‐neural‐network‐based algorithm, PmPNet, to accelerate the process of identifying the PmP waves. We conduct systematic research with field data, and conclude that the PmPNet can efficiently achieve high precision and high recall simultaneously for automatically identifying the PmP waves from a massive seismic database. Applying the pre‐trained PmPNet to the seismic database from January 1990 to December 1999, we have tripled the PmP data set in southern California. Key Points: A deep‐neural‐network‐based algorithm, PmPNet, is developed to automatically pick the Moho‐reflected seismic waves PmPPmPNet efficiently achieves high precision and recall simultaneously to automatically identify PmP waves from a massive seismic databasePmPNet creates a large PmP database with a total of 28,093 high‐quality PmP picks in southern California [ABSTRACT FROM AUTHOR]

Global Navigation Satellite System (GNSS) vertical displacements measuring the elastic response of Earth's crust to changes in hydrologic mass have been used to produce terrestrial water storage change (∆TWS) estimates for studying both annual ∆TWS as well as multi‐year trends. However, these estimates require a high observation station density and minimal contamination by nonhydrologic deformation sources. The Gravity Recovery and Climate Experiment (GRACE) is another satellite‐based measurement system that can be used to measure regional TWS fluctuations. The satellites provide highly accurate ∆TWS estimates with global coverage but have a low spatial resolution of ∼400 km. Here, we put forward the mathematical framework for a joint inversion of GNSS vertical displacement time series with GRACE ∆TWS to produce more accurate spatiotemporal maps of ∆TWS, accounting for the observation errors, data gaps, and nonhydrologic signals. We aim to utilize the regional sensitivity to ∆TWS provided by GRACE mascon solutions with higher spatial resolution provided by GNSS observations. Our approach utilizes a continuous wavelet transform to decompose signals into their building blocks and separately invert for long‐term and short‐term mass variations. This allows us to preserve trends, annual, interannual, and multi‐year changes in TWS that were previously challenging to capture by satellite‐based measurement systems or hydrological models, alone. We focus our study in California, USA, which has a dense GNSS network and where recurrent, intense droughts put pressure on freshwater supplies. We highlight the advantages of our joint inversion results for a tectonically active study region by comparing them against inversion results that use only GNSS vertical deformation as well as with maps of ∆TWS from hydrological models and other GRACE solutions. We find that our joint inversion framework results in a solution that is regionally consistent with the GRACE ∆TWS solutions at different temporal scales but has an increased spatial resolution that allows us to differentiate between regions of high and low mass change better than using GRACE alone. Plain Language Summary: Fluctuations in surface water mass cause deformation of the Earth's surface and small changes to the Earth's gravity field. Both of these effects can be observed by satellite‐based observation systems and used to solve for terrestrial water storage change (∆TWS), which is the sum of the changes in groundwater, surface water, soil moisture, ice, and snow over the land surface. These observations are essential for drought monitoring and can be used to inform water management decisions. In this study, we combine deformation data with gravity‐derived ∆TWS to solve for monthly ∆TWS over California and western Nevada, USA. Here, multiyear droughts cause sharp declines in TWS and risk the future productivity of the water‐intensive agricultural sector. Key Points: A new joint inversion framework combining Global Navigation Satellite System and Gravity Recovery and Climate Experiment (GRACE) is put forward to calculate monthly terrestrial water storage change (∆TWS) over California from 2003 to 2016A wavelet‐based approach is used to maintain spectral frequency content, allowing for preservation of multi‐ to sub‐annual signal componentsInversion results show improved isolation of larger magnitude signals over regions where we expect large ∆TWS as compared to GRACE alone [ABSTRACT FROM AUTHOR]

We present new P and S teleseismic tomography models of the upper mantle beneath California and Nevada. Improved tomographic methods and the incorporation of new data increase image resolution and seismic structure definition. Technical advancements include the use of 3‐D ray tracing, ray weighting and finite‐frequency sensitivity kernels. Together with the incorporation of modern crustal velocity models to account for crustal delay time, these tomographic advancements reveal slightly higher amplitudes and narrower structures beneath prominent tectonic features in California and Nevada. The addition of new data from the southern California Borderland and northwestern California are particularly useful. Based on our imaging results, we discuss the effects of temperature and partial melt on upper mantle seismic velocity beneath California and Nevada. We interpret the major seismically fast structures in our study area to be Juan de Fuca slab and its downward continuation beneath central Nevada and foundered Farallon lithosphere beneath the southern Sierra Nevada and Transverse Ranges. The seismically slow structures beneath the slab‐free window of northern California Coast Range, the Salton Trough extensional basin, and the Cascadia back arc, are all consistent with the presence of partial melting of decompressing mantle. The imaged seismically fast and slow anomalies imply buoyancy variations and melt productivity that ultimately act on the overlying crust. Plain Language Summary: We present new images of the structure of the upper mantle beneath California and Nevada based on the travel times of seismic waves from around the world. Modern methods and new data increase image resolution and seismic structure definition, and these advancements reveal stronger and narrower structures beneath prominent geologic features in California and Nevada. Based on our results we discuss the effects of temperature and partial melt on upper mantle seismic velocity beneath this region. We interpret structures in the upper mantle beneath central Nevada, the southern Sierra Nevada, the Transverse Ranges, and the Salton Trough. The imaged regions of faster seismic wave speeds are associated primarily with the Juan de Fuca‐Farallon slab. Structures imaged with slower seismic wave speeds indicate the presence of partial melt in the upper mantle. These seismic anomalies imply variations in the upper mantle that ultimately impact the overlying crust in California and Nevada. Key Points: We image upper mantle structures beneath California and Nevada with improved seismic techniques and an expanded data setThe major seismically fast anomalies likely represent fragments of Farallon oceanic lithosphereSlow anomalies are consistent with elevated temperatures and partial melt of decompressing mantle [ABSTRACT FROM AUTHOR]

Tectonic, hydrological and industrial processes coexist in the dynamic natural environments. However, our knowledge of ground deformation associated with tectonic, hydrological and anthropogenic processes and their interactions remains limited. California represents a natural laboratory that hosts the San Andreas fault system, Central Valley and other aquifer systems, and extensive human extraction of natural resources. The attendant multi‐scale ground deformation that has been mapped using Copernicus Sentinel‐1 Synthetic Aperture Radar (SAR)‐satellite constellation from four ascending and five descending tracks during 2015–2019. We consider the secular horizontal surface velocities and strain rates, constrained from GNSS measurements and tectonic models, as proxies for tectonic processes, and seasonal displacement amplitudes from interferometric SAR (InSAR) time series as proxies for hydrological processes. We synergize 23 types of multidisciplinary datasets, including ground deformation, sedimentary basins, precipitation, soil moisture, topography, and hydrocarbon production fields, using a machine learning algorithm–random forest, and we succeed in predicting 86%–95% of the representative data sets. High strain rates along the SAF system mainly occur in areas with a low‐to‐moderate vegetation fraction (∼0.3), suggesting a correlation of rough/high‐relief coastal range morphology and topography with the active faulting, seasonal and orographic rainfall, and vegetation growth. Linear discontinuities in the long‐term, seasonal amplitude and phase of the surface displacement fields coincide with some fault strands, the boundary zone between the sediment‐fill Central Valley and bedrock‐dominated Sierra Nevada, and the margins of the inelastically deforming aquifer in the San Joaquin Valley, suggesting groundwater flow interruptions, contrasting elastic properties, and heterogeneous hydrological units. Plain Language Summary: Although scientific advances have been achieved in every individual geoscience discipline with more extensive and accurate observations and more robust models, our knowledge of the Earth complexity remains limited. The tectonic, hydrological and anthropogenic processes interact in the highly populous California, which have contributed to multi‐scale ground deformation. Here we rely on remotely sensed ground deformation products and locations of oil and gas fields as proxies for tectonic, hydrological and anthropogenic processes. The training model of the random forest algorithm has a good performance in predicting the representative multidisciplinary datasets and reveals their intrinsic similarities and relative importance in the predictions. We also note compelling spatial correlation between the long‐term and seasonal displacement discontinuities, the high‐strain‐rate fault zones, and a narrow range of vegetation fraction, as well as the margins of the heterogeneous structures. Key Points: Discontinuities in the long‐term, seasonal amplitude and phase of geodetically measured displacement fields coincide with fault strandsThe belt of high strain rate is characterized by the low‐to‐moderate vegetation fractionPhase discontinuities in seasonal displacement occur near the transition between sedimentary basins and consolidated ranges [ABSTRACT FROM AUTHOR]

Repeating earthquakes, earthquakes that rupture the same fault patch more than once, are playing an important role in the study of source processes and fault behavior, and have the potential to improve hazard assessment and earthquake forecast. They are believed to represent brittle failure of a stuck patch (asperity) on an otherwise creeping fault, and as such may offer insight into slip behavior at depth. We carried out a systematic analysis in search for repeating earthquakes in northern California between 1984 and 2014 by identifying events that are co‐located, are of similar size, and share similar waveforms. We find 27,675 events in 7,713 repeating earthquake sequences (RES) that include between 2 and 32 events with magnitudes between −0.5 and 6.5. Temporal behavior of the events within each RES ranges between random and quasi‐periodic over the 30‐year observation period, and includes temporal clustering (bursts). The RESs occur throughout northern California, with some of the largest and most periodic sequences predominately, but not exclusively, along faults within the San Andreas Fault system that exhibit surface creep. Piecewise periodic RESs are found near large earthquakes where they are modulated by transient stress. Slip rates inferred from the new data, while generally consistent with published estimates from geodetic and surface data, offer a high‐resolution image of subsurface creep along individual fault strands and resolve slip partitioning across wider zones of deformation. Plain Language Summary: Repeating earthquakes are earthquakes that rupture the same fault patch more than once, often numerous times and with surprising periodicity. They are believed to represent stuck patches on a fault surface that otherwise creeps in response to tectonic stress. This makes them ideal for studying processes that cause earthquakes and control fault behavior. Because we can use repeating earthquake sequences as slip meters at depth, they are also playing an important role in the assessment of seismic hazard. In this study we carried out a systematic search for repeating earthquakes in northern California by mining 30 years of earthquake data. We find 27,675 repeating earthquakes included in 7,713 sequences that locate throughout northern California, predominately, but not exclusively, along faults within the San Andreas Fault system that exhibit creep at the surface. The pattern of event timing in each sequence ranges from random to quasi‐periodic. That pattern is disturbed by transient stresses in cases where a sequence locates close to a large earthquake. Slip rates inferred from the new data offer a high‐resolution view of subsurface creep along individual fault strands and resolve slip partitioning across wider zones of deformation. Key Points: We identify 27,675 repeating earthquakes in 7,713 sequences based on waveform similarity, co‐location, and size of the eventsThe sequences are widespread, with temporal behavior of events ranging from random to quasi‐periodic over the 30‐year observation periodThe new data imply aseismic slip partitioning across many major and minor faults within the San Andreas Fault system [ABSTRACT FROM AUTHOR]

We develop finite element models of the coseismic displacement field accounting for the 3D elastic structures surrounding the epicentral area of the 2019 Ridgecrest earthquake sequence containing two major events of Mw7.1 and Mw6.4. The coseismic slip distribution is inferred from the surface displacement field recorded by interferometric synthetic aperture radar. The rupture dip geometry is further optimized using a novel nonlinear‐crossover‐linear inversion approach. It is found that accounting for elastic heterogeneity and fault along‐strike curvilinearity improves the fit to the observed displacement field and yields a more accurate estimate of geodetic moment and Coulomb stress changes. We observe spatial correlations among the locations of aftershocks and patches of high slip, and rock anomalous elastic properties, suggesting that the shallow crust's elastic structures possibly controlled the Ridgecrest earthquake sequence. Most of the coseismic slip with a peak slip of 7.4 m at 3.6 km depth occurred above a zone of reduced S‐wave velocity and significant post‐Mw7.1 afterslip. This implies that viscous materials or fluid presence might have contributed to the low rupture velocity of the mainshock. Moreover, the zone of high slip on the northwest‐trending fault segment is laterally bounded by two aftershock clusters, whose location is characterized by intermediate rock rigidity. Notably, some minor orthogonal faults consistently end above a subsurface rigid body. Overall, these observations of structural controls improve our understandings of the seismogenesis within incipient fault systems. Plain Language Summary: On July 4 and 5, 2019, a magnitude 7.1 and 6.4 earthquake occurred near the town, Ridgecrest, California. The permanent surface movements induced by these events were captured by satellite images. Here, we develop realistic numerical models and innovational optimization algorithms to investigate the displacement fields and image the distribution of the earthquake slip. These events were hosted by a pair of orthogonal faults and their resolved orientations are similar to the nearby known fault structures. We discover that the fault slip and aftershock locations could be related to the local rock materials, suggesting that the variable elastic properties of the shallow crust might have influenced the development of the Ridgecrest earthquake sequence. As also supported by other seismic and geophysical datasets, most of the coseismic slip appeared over a zone of elastic weakness where could be relevant to subsurface viscous or fluid bodies, meanwhile, aftershocks tended to cluster within a shallow rigid body. Overall, the findings of this study provide insights into potential structural controls over the seismic occurrence and help improve our understandings of the seismogenesis within the western Mojave desert. Key Points: Heterogeneous velocity structures in the epicentral area are simulated in elastic modelsPeak slip occurred above a zone of reduced S‐wave velocity and documented afterslipMost aftershocks are located outside slip asperities and within a shallow brittle zone [ABSTRACT FROM AUTHOR]

Strength of the upper brittle part of the Earth's lithosphere controls deformation styles in tectonically active regions, surface topography, seismicity, and the occurrence of plate tectonics, yet it remains one of the most debated quantities in geophysics. Direct measurements of stresses acting at seismogenic depths are largely lacking. Seismic data (in particular, earthquake focal mechanisms) have been used to infer orientation of the principal stress axes. I show that the focal mechanism data can be combined with information from precise earthquake locations to place constraints not only on the orientation, but also on the magnitude of absolute stress at depth. The proposed method uses relative attitudes of conjugate faults to evaluate the amplitude and spatial heterogeneity of the deviatoric stress and frictional strength in the seismogenic zone. Relative fault orientations (dihedral angles) and sense of slip are determined using quasi‐planar clusters of seismicity and their composite focal mechanisms. The observed distribution of dihedral angles between active conjugate faults in the area of Ridgecrest (California, USA) that hosted a recent sequence of strong earthquakes suggests in situ coefficient of friction of 0.4–0.6, and depth‐averaged shear stress on the order of 25–40 MPa, intermediate between predictions of the "strong" and "weak" fault theories. Plain Language Summary: Strength of the Earth's crust is one of the most important yet least known quantities in geophysics. It determines how high the mountains can grow and how large the earthquakes can be, but is notoriously difficult to measure. Only a handful of direct measurements are available from data collected in deep boreholes, and even fewer measurements were made in the vicinity of faults capable of large earthquakes. This study uses clusters of small earthquakes to map out active faults at depth, and determine their relative orientations. Such orientations (more precisely, the angle between faults having an opposite sense of slip "conjugate faults") provide clues about how strong natural faults are, and what is the average level of shear stress supported by the seismogenic crust. Data from the Ridgecrest (California) area that hosted a sequence of large earthquakes in 2019 suggest that the newly formed faults progressively weaken as they mature and accumulate more slip. The average shear stress acting at the seismogenic death is found to be on the order of several tens of Megapascals, lower than predicted by the "strong fault" theory, but high enough to sustain earthquake sequences on the same fault structure (as observed in 2019). Key Points: Distribution of dihedral angles between active conjugate faults is used to estimate the magnitude of deviatoric stress at seismogenic depthDihedral angles are quantified using linear clusters of precisely located earthquakes and focal mechanism dataData from the Ridgecrest (California) area suggests in situ coefficient of friction of 0.4–0.6, and depth‐averaged shear stress of 25–40 MPa [ABSTRACT FROM AUTHOR]

Distilling my experience in having field mapped in detail the volcanic fields at Laguna del Maule and Long Valley and having worked out their time‐volume‐composition magmatic histories, I compare and contrast the postglacial rhyolites of the former with six multi‐vent eruptive sequences of rhyolite in California. Compilations and discussions are made of volcanic‐field areas and longevities, their compositions, vent distributions, individual batch and total volumes, eruptive episodicities, and tectonic influences. Growth of long‐lived pluton‐scale reservoirs of granitic crystal mush, from which the rhyolite melts separated, are interpreted in terms of conceptual models I published previously—(1) fundamentally basaltic transcrustal magmatism, 1981; (2) the deep‐crustal MASH zone model, 1988; and (3) the rhyolite‐melt crystal‐mush model, 2001. Inferences and speculations are advanced concerning processes and timescales of rhyolite‐melt separation from granitic mush and of prompt or long‐delayed subsequent eruption. Key Points: High‐silica rhyolite melts extremely depleted in Sr, Ba, and Eu had been interstitial liquid extracted from leucogranitic crystal mushCrustal conditioning during million‐year‐long basaltic episodes typically precedes the growth of upper‐crustal granitic mush reservoirsTimescales range widely for rhyolite melt batch ascent from mush and subsequently for roof‐zone storage of such low‐T water‐rich melt [ABSTRACT FROM AUTHOR]

The exploitation of ever increasing Interferometric Synthetic Aperture Radar (InSAR) data sets to monitor the Earth's surface deformation is an important goal of today's geodesy. In this study our observations consist of deformations along the Line‐Of‐Sight direction of the satellite. Our observations are the result of the combination of a multitude of sources (either volcano‐tectonic or nontectonic deformation). In most cases, we are facing a Blind source separation (BSS) problem. Natural approaches to tackle BSS problems are multivariate statistical techniques that attempt to decompose the data set into a limited number of statistically independent sources, under the assumption that the different physical mechanisms contributing to the observations have independent footprints in space and/or time. We show the capabilities of a variational Bayesian independent component analysis (vbICA) algorithm in dealing with synthetic InSAR time series and compare it to the commonly used FastICA algorithm. We explore the effectiveness of the spatial and temporal mode decompositions. We apply vbICA on data relative to the San Joaquin Valley and the Central San Andreas fault (CSAF), California, spanning the time range 2015/03/01–2019/07/14. The proposed approach likely isolates the contribution of shallow and deep aquifers to the surface deformation as well as the elastic and inelastic deformation. We present a 1‐dimensional compaction estimation of the elastic and inelastic storage coefficients adopting a formalism that takes into account the last century water level history. Concerning the CSAF, the algorithm helps separating tectonic loading from seasonal behavior concentrated in the Quaternary sediments of the Salinas Valley. Key Points: Variational Bayesian independent component analysis on Interferometric Synthetic Aperture Radar displacement time series performs better than FastICAApplication to San Joaquin Valley allows to isolate elastic and inelastic compaction and potential hydrocompaction contributionsApplication to Central San Andreas fault zone allows to separate seasonal behavior of Quaternary sediments from creep on the fault [ABSTRACT FROM AUTHOR]

The Hayward fault in California's San Francisco Bay area produces large earthquakes, with the last occurring in 1868. We examine how physics‐based dynamic rupture modeling can be used to numerically simulate large earthquakes on not only the Hayward fault, but also its connected companions to the north and south, the Rodgers Creek and Calaveras faults. Equipped with a wealth of images of this fault system, including those of its 3D geology and 3D geometry, in addition to inferences about its interseismic creep‐rate pattern and rock‐friction behavior, we use a finite‐element computer code to perform 3D dynamic earthquake rupture simulations. We find that the rock properties affect the locations and amount of slip produced in our simulated large earthquakes. Crucial factors that control rupture behavior in our modeling are the earthquake nucleation locations, the fault geometry, and the data that reveal where the fault system is creeping or locked. Our findings suggest that large Rodgers Creek‐Hayward‐Calaveras‐Northern Calaveras (RC‐H‐C‐NC) fault‐system earthquakes may result from dynamic rupture that starts in a locked part of the fault system, but is then stopped by the creeping parts, leading to high‐magnitude‐6 earthquakes; or, from dynamic rupture that starts in a locked part of the fault system, then cascades through some of the creeping parts, leading to magnitude‐7 earthquakes. Plain Language Summary: The San Francisco Bay area has experienced large earthquakes in the past and is likely to experience large earthquakes in the future. Although the San Andreas fault is the best‐known earthquake generator, other San Francisco Bay area faults, including some that slowly move (creep) all of the time are also known to be hazardous. We use computer modeling to simulate large earthquakes on the Hayward fault, and its connected companions to the north and south, the Rodgers Creek and Calaveras faults. In our computer simulations, we include information about the rock properties, rock‐friction, and fault geometry, in addition to inferences about where the faults creep. We find that the earthquake nucleation location, the fault geometry, and the pattern of fault creep, are most important. Our findings suggest that large earthquakes on these faults may result from a fast‐slipping rupture that starts in a locked part of the faults, but is then stopped by the slowly slipping parts of the faults, leading to high‐magnitude‐6 earthquakes, or, from fast‐slipping rupture that starts in a locked part of the faults, then cascades through some of the creeping‐fault parts, leading to magnitude‐7 earthquakes. Key Points: We simulated large earthquakes on the Rodgers Creek‐Hayward‐Calaveras fault system, CaliforniaOur spontaneous rupture (dynamic rupture) simulations included the 3D fault geometry, 3D rock properties, and creeping‐fault patchesThe resulting earthquakes are most affected by which fault they start on, the fault geometry, and the pattern of interseismic fault creep [ABSTRACT FROM AUTHOR]

The 2019 Ridgecrest, California, earthquake sequence represents a complex pattern of seismicity that is characterized by the occurrence of a well‐defined foreshock sequence followed by a mainshock and subsequent aftershocks. In this study, a detailed statistical analysis of the sequence is performed. Particularly, the parametric modeling of the frequency‐magnitude statistics and the earthquake occurrence rate is carried out. It is shown that the clustering of earthquakes plays an important role during the evolution of this sequence. In addition, the problem of constraining the magnitude of the largest expected aftershocks to occur during the evolution of the sequence is addressed. In order to do this, two approaches are considered. The first one is based on the extreme value theory, whereas the second one uses the Bayesian predictive framework. The latter approach has allowed to incorporate the complex earthquake clustering through the Epidemic Type Aftershock Sequence (ETAS) process and the uncertainties associated with the model parameters into the computation of the corresponding probabilities. The results indicate that the inclusion of the foreshock sequence into the analysis produces higher probabilities for the occurrence of the largest expected aftershocks after the M7.1 mainshock compared to the approach based on the extreme value distribution combined with the Omori‐Utsu formula for the earthquake rate. Several statistical tests are applied to verify the forecast. Plain Language Summary: Strong earthquakes typically trigger the subsequent sequence of events known as aftershocks. Among those, the largest aftershocks can pose significant hazard and result in additional damage to infrastructure already weakened by the mainshock. Therefore, the estimation of the magnitude of the largest expected aftershock is of critical importance. This problem can be addressed within the statistical modeling of the occurrence of earthquakes. In this study, the 2019 Ridgecrest, California, earthquake sequence is chosen to illustrate and compare several approaches to constrain the magnitudes of the largest expected aftershocks during the evolution of the sequence. The first approach uses the extreme value theory and the modeling of the earthquake rate based on the Omori‐Utsu formula. Whereas, the second approach uses a recently formulated method based on the Bayesian predictive analysis and the Epidemic Type Aftershock Sequence (ETAS) model to approximate the earthquake rate. The obtained results indicate that the latter approach produces statistically accurate forecast for the magnitudes of the largest expected earthquakes. This is verified by applying several statistical tests. Key Points: Statistical analysis of the 2019 Ridgecrest, California, earthquake sequence is performedThe probabilities for the occurrence of the largest expected aftershocks are computed using the Bayesian predictive frameworkThe aftershock forecast is verified retrospectively using several statistical tests [ABSTRACT FROM AUTHOR]

Accurate ground motion prediction requires detailed site effect assessment, but in urban areas where such assessments are most important, geotechnical surveys are difficult to perform, limiting their availability. Distributed acoustic sensing (DAS) offers an appealing alternative by repurposing existing fiber‐optic cables, normally employed for telecommunication, as an array of seismic sensors. We present a proof‐of‐concept demonstration by using DAS to produce high‐resolution maps of the shallow subsurface with the Stanford DAS array, California. We describe new methods and their assumptions to assess H/V spectral ratio—a technique widely used to estimate the natural frequency of the soil—and to extract Rayleigh wave dispersion curves from ambient seismic field. These measurements are jointly inverted to provide models of shallow seismic velocities and sediment thicknesses above bedrock in central campus. The good agreement with an independent survey validates the methodology and demonstrates the power of DAS for microzonation. Key Points: We demonstrate the potential of DAS for site effect analysisDAS recordings are used to compute dispersion curves and horizontal‐to‐vertical spectral ratio (HVSR)Joint inversions suggest that the crystalline bedrock lies 115 m beneath Stanford University central campus [ABSTRACT FROM AUTHOR]

Our understanding of plate boundary deformation has been enhanced by transient signals observed against the backdrop of time‐independent secular motions. We make use of a new analysis of displacement time series from about 1,000 continuous Global Positioning System (GPS) stations in California from 1999 to 2018 to distinguish tectonic and nontectonic transients from secular motion. A primary objective is to define a high‐resolution three‐dimensional reference frame (datum) for California that can be rapidly maintained with geodetic data to accommodate both secular and time‐dependent motions. To this end, we compare the displacements to those predicted by a horizontal secular fault slip model for the region and construct displacement and strain rate fields. Over the past 19 years, California has experienced 19 geodetically detectable earthquakes and widespread postseismic deformation. We observe postseismic strain rate variations as large as 1,000 nstrain/year with moment releases equivalent up to an Mw6.8 earthquake. We find significant secular differences up to 10 mm/year with the fault slip model, from the Mendocino Triple Junction to the southern Cascadia subduction zone, the northern Basin and Range, and the Santa Barbara channel. Secular vertical uplift is observed across the Transverse Ranges, Coastal Ranges, Sierra Nevada, as well as large‐scale postseismic uplift after the 1999 Mw7.1 Hector Mine and 2010 Mw7.2 El Mayor‐Cucapah earthquakes. We also identify areas of vertical land motions due to anthropogenic, natural, and magmatic processes. Finally, we demonstrate the utility of the kinematic datum by improving the accuracy of high‐spatial‐resolution 12‐day repeat‐cycle Sentinel‐1 Interferometric Synthetic Aperture Radar displacement and velocity maps. Key Points: Displacement fields from 19 years of GPS data in California are preferable to velocity fields in characterizing 3‐D transient motionsMapping of secular and transient motions at high temporal and spatial resolution is required for maintaining a 3‐D kinematic reference frameDisplacement fields derived from continuous GPS and InSAR provide 200‐m resolution with millimeter‐level velocity uncertainties [ABSTRACT FROM AUTHOR]

Long period (LP) seismicity is a ubiquitous feature at Mammoth Mountain, California. Phase‐based locations exhibit source depths ranging from ∼8 to 25 km. Between September 2011 and October 2013, over 200 LP events were added to the Northern California Earthquake Data Center catalog. Three typical LP events were used as master events, and continuous waveform correlation was performed over the 2‐year period, with the results indicating that thousands of LP events were present in the continuous seismic data. Using phase‐weighted stacking, representative network stacks for the three events were produced. Full waveform inversion of the representative events using Green functions that adequately represent the topography of Mammoth Mountain and the high‐resolution velocity structure produced through local earthquake tomography suggest that all three event types are located within ∼100 m of the surface in the vicinity of the Chair 12 tree kill and CO2 emission area and within ∼1 km of the Mammoth Mountain Ski Area base lodge. The source mechanisms are purely volumetric and can be attributed to a subhorizontal hydrothermal crack excited into acoustic resonance by the rapid discharge of CO2 gases trapped in the crack. An accompanying vertical force that is opposite in phase to the volume change in the crack can be attributed to pressure and momentum changes in an adjacent fluid‐filled conduit linking the crack to the surface. The implication of this work is that LP seismicity at Mammoth Mountain occurs almost continuously and in direct contradiction to the traditional view—at very shallow depths. Key Points: Long period seismicity at Mammoth Mountain between 2011 and 2013 occurred at shallow depthsLong period seismicity is attributed to the excitation of shallow subhorizontal hydrothermal cracksDriving force is the discharge of CO2 from the underlying gas reservoir [ABSTRACT FROM AUTHOR]

Adjoint tomography has recently been applied to ambient noise data as a new and promising tomographic method that utilizes simulation‐based 3‐D sensitivity kernels rather than ray theory used in traditional ambient noise tomography. However, to date, most studies of ambient noise adjoint tomography only use vertical‐component Rayleigh waves. In this study, we develop a theoretical framework for calculating sensitivity kernels for multicomponent empirical Green's functions extracted from ambient noise data. Under the framework of the adjoint method, we demonstrate that a horizontal component (transverse‐transverse or radial‐radial) kernel can be constructed from the interaction of wave fields generated by point‐force sources acting in the north and east directions based on rotation relationships. Our method is benchmarked for a 3‐D heterogeneous isotropic model by comparing rotated seismograms, individual, and event traveltime misfit kernels with corresponding references computed by numerical simulations with sources directly placed in the radial or transverse directions. Based on our new method, we perform the first Love‐wave ambient noise adjoint tomography in southern California and construct an improved VSH model. Our method for computing sensitivity kernels of multicomponent empirical Green's functions provides the basis for multicomponent ambient noise adjoint tomography in imaging radially anisotropic shear‐wave velocity structures. Key Points: A theoretical framework for calculating 3‐D individual and event kernels for multicomponent ambient noise adjoint tomography is developedThe kernels are validated in a 3‐D heterogeneous isotropic medium and can be efficiently computed for adjoint tomographic applicationThe method can be extended to construct kernels for radially anisotropic model parameters and Rayleigh wave ellipticity [ABSTRACT FROM AUTHOR]

We use GPS data to show synchronization between the 2011 and 2016 drought cycle in California, accelerated uplift of the Sierra Nevada Mountains, and enhanced magmatic inflation of the Long Valley Caldera (LVC) magmatic system. The drought period coincided with faster uplift rate, changes in gravity seen in the Gravity Recovery and Climate Experiment (GRACE), and changes in standardized relative climate dryness index. These observations together suggest that the Sierra Nevada elevation is sensitive to changes in hydrological loading conditions, which subsequently influences the LVC magmatic system. We use robust imaging of horizontal GPS velocities to derive time‐variable shear and dilatational strain rates in a region with highly variable station distribution. The results show that the highest strain rates are near the eastern margin of the Sierra Nevada and western edge of the Central Walker Lane (CWL) passing directly through LVC. The drought period saw geographic shifts in the distribution in active shear strain in the CWL more than 60 km from the LVC, delineating the minimum extent over which the active magmatic system affects the CWL tectonic environment. We analyze declustered seismicity data to show that locations with higher seismicity rates tend to be (1) areas with higher strain rates and (2) areas in which strain rates increased during drought‐enhanced inflation. We hypothesize that drought conditions reduce vertical surface mass loading, which decreases pressure at depth in the LVC system, in turn enhances magmatic inflation, and drives horizontal elastic stress changes that redistribute active CWL strain and modulate seismicity. Plain Language Summary: We show that there were connections between the recent drought in California to changes in uplift rate of the Sierra Nevada, magmatic inflation at the Long Valley Caldera near Mammoth, CA, and earthquakes in the region. We use GPS, seismic, gravity, and climate data to show that the speed of uplift of the Sierra Nevada is sensitive to big changes in surface and groundwater (e.g., in snow, streams, lakes, and aquifers) and that these changes correlate with movements within the large volcano. A plausible explanation is that the weight of snow and water loads the surface and changes pressures deep in the Earth. As the Sierra Nevada was unloaded during recent drought, there was less pressure on the magmatic system, allowing fluids to rise or expand more easily. When that occurred, it provided an outward horizontal push on the crust, influencing small‐ to moderate‐sized earthquakes up to 60 km away. We show where and when these changes occurred and specific examples of earthquakes that may have been triggered by this inflation. Key Points: Faster Sierra Nevada uplift during the 2011–2016 drought in California coincided with faster magmatic inflation at the Long Valley CalderaFaster magmatic inflation changed the rate and distribution of crustal strain in the Central Walker Lane, >60 km from Long ValleySeismicity patterns in the Central Walker Lane suggest that long‐term crustal strain may be driven in part by Long Valley magmatic inflation [ABSTRACT FROM AUTHOR]

The driving concept behind one of the most successful statistical forecasting models, the Epidemic Type Aftershock Sequence (ETAS) model, has been that the seismicity is driven by spontaneously occurring background earthquakes that cascade into multitudes of triggered earthquakes. In nearly all generalizations of the ETAS model, the magnitudes of the background and the triggered earthquakes are assumed to follow Gutenberg‐Richter law with the same exponent (β value). Furthermore, the magnitudes of the triggered earthquakes are always assumed to be independent of the magnitude of the triggering earthquake. Using an expectation maximization algorithm applied to the Californian earthquake catalog, we show that the distribution of earthquake magnitudes exhibits three distinct β values: βb for background events and βa − δ and βa + δ, respectively, for triggered events below and above the magnitude of the triggering earthquake; the two last values express a correlation between the magnitudes of triggered events with that of the triggering earthquake, a feature so far absent in all proposed operational generalizations of the ETAS model. The ETAS model incorporating this kinked magnitude distribution provides by far the best description of seismic catalogs and could thus have the best forecasting potential. We speculate that the kinked magnitude distribution may result from the system tending to restore the symmetry of the regional displacement gradient tensor that has been broken by the initiating event. The general emerging concept could be that while the background events occur primarily to accommodate the symmetric stress tensor at the boundaries of the system, the triggered earthquakes are quasi‐Goldstone fluctuations of a self‐organized critical deformation state. Key Points: Magnitudes of background earthquakes are distributed according to Gutenberg‐Richter (GR) law The magnitudes of earthquakes directly triggered by a given earthquake follow a kinked GR law with two branches The location of the kink coincides with the magnitude of the triggering event [ABSTRACT FROM AUTHOR]

A total of seven independent ML ≥ 4.0 earthquakes occurred in the Los Angeles, California, basin, during the early instrumental period between 1932 and 1952, the largest of which was the 1933 Long Beach earthquake. Revising available macroseismic and instrumental data for a total of 6 4.0 ≤ ML ≤ 5.1 events between 1938 and 1944, we conclude that early instrumental locations can be grossly inconsistent with detailed macroseismic data. We use available macroseismic data to revisit event locations. We further present evidence that most if not all of these moderate earthquakes may have been induced by oil production. We quantify the predicted stress change associated with production from eight oil fields in the southwestern Los Angeles basin and show that frictional failure would have been encouraged beneath and at the periphery of high‐volume fields, with stress changes upward of 0.1 MPa at 5‐km depth. The results suggest that if earthquakes are induced by stress changes associated with production, the magnitudes of events might tend to be limited by the limited spatial extent of lobes of increased Coulomb failure stress. It further appears that the advent of fluid injection recovery methods (water‐flooding) around 1960 mitigated induced earthquake risk considerably. Plain Language Summary: We reconsider the locations and magnitudes of six moderate earthquakes that occurred in the Los Angeles region between 1935 and 1950. We show that, while these earthquakes were recorded by early seismometers, detailed damage observations provide a better indication of locations than do limited instrumental data. We further present evidence for an association between these earthquakes and oil production during the oil boom that began in the late 1930s. Using available industry data and established computation methods, we show that by the 1930s, oil production would have perturbed significantly the stress and faults in proximity to major fields at depths of 3–5 km. We suggest that the extent and therefore magnitude of induced earthquakes would have been generally limited by the spatially complex patterns of stress change. We further suggest that induced earthquake risk was mitigated significantly by the advent of so‐called secondary recovery methods, whereby water was reinjected into fields to compensate for oil extraction. Key Points: We discuss a possible association between early twentieth‐century earthquakes and oil production in CaliforniaCoulomb stress modeling reveals significant stress changes at depths of ~5 km associated with oil productionRupture lengths of production‐induced earthquakes might be limited by the small scale of stress‐change lobes [ABSTRACT FROM AUTHOR]