Search

The violent eruption of the Hunga (Tonga) submarine volcano on 15 January 2022 caused a 58 km‐heigh ash plume, catastrophic tsunami, and significant global seismic and infrasound waves. However, the physical mechanism underpinning its multiple‐explosive events remains unclear, and its resolvability relies on the seismic waveform source inversion. The studies of two different point‐source models, the seismic moment tensor (MT) and the single force (SF), have been performed separately for this eruption, which, interestingly, can explain the seismic data adequately. Here, we use a joint inversion of MT and SF to unravel a composite source of an explosion‐like MT and a significant upward force for the first major explosive event. Regarding the direction and magnitude, we propose that the upward force is likely a rebound force in response to the pressure drop on the seafloor because the water body above the volcano was abruptly uplifted by the shallow underwater explosion. The physical process of the violent eruption of the Hunga (Tonga) submarine volcano on 15 January 2022 remains unclear. To date, the common source model for volcano eruptions—a single force (SF) and the common source model for earthquakes and explosions—a moment tensor (MT), have been inferred individually for this eruption. Interestingly, both can explain the recorded seismic signals reasonably well. A question arises whether a combination of sources is a better physical model. Therefore, we combine the MT and SF to represent the eruption process in this study. The source analysis for the first major event of this eruption reveals a possible composite process of a shallow underwater explosion and a significant upward force. The upward force is opposite the common downward‐reaction force to the material jetting. It is likely caused by the abrupt displacement of the water above the volcano resulting from the shallow underwater explosion. When the downward water pressure on the seafloor vanishes, the seafloor responds by an upward‐rebound force. We perform probabilistic inversion of seismic waveform data to study the force equivalent system of the 2022 Hunga (Tonga) eruptionThe first major explosive event of the shallow eruption sequences may consist of an explosion‐like moment tensor and a large upward forceA possible mechanism of the accompanying upward force is the rebound force responding to the sudden pressure drop of uplifted water body We perform probabilistic inversion of seismic waveform data to study the force equivalent system of the 2022 Hunga (Tonga) eruption The first major explosive event of the shallow eruption sequences may consist of an explosion‐like moment tensor and a large upward force A possible mechanism of the accompanying upward force is the rebound force responding to the sudden pressure drop of uplifted water body

A minequake of magnitude ML4.1 occurred on 18 May 2020 early in the morning at the LKAB underground iron ore mine in Kiruna, Sweden. This is the largest mining‐induced earthquake in Scandinavia. It generated acoustic signals observed at three infrasound arrays at 9.3 (KRIS, Sweden), 155 (IS37, Norway), and 286 km (ARCI, Norway) distance. We perform full‐waveform focal mechanism inversion based on regional seismic data and local infrasound data. These independently highlight that this event was dominated by a shallow‐depth collapse in agreement with in‐mine seismic station data. However, regional infrasound data cannot inform the inversion process without an accurate model of atmospheric winds and temperatures. Yet, our numerical simulations demonstrate a potential of using local and regional infrasound data to constrain an event's focal mechanism and depth. The largest mining‐induced earthquake in Scandinavia (ML4.1) occurred on 18 May 2020 early in the morning at the LKAB underground iron ore mine in Kiruna, Sweden. The seismic waves coupled to the atmosphere and propagated large distances as sound waves which were observed at three infrasound arrays at 9.3 (KRIS, Sweden), 155 (IS37, Norway), and 286 km (ARCI, Norway) distance. Our seismic and acoustic modeling results highlight a strong collapse event within the northern section of the mine. The modeling of acoustic and seismic waves across the Earth‐atmosphere suggests that sound wave data can help when determining the location and properties of a seismic source. Seismic and acoustic records indicate a strong collapse at shallow depth of 1  ±  0.5 km for the 2020 Kiruna minequakeFocal mechanisms and depths of shallow seismic sources can be retrieved from local infrasound recordsInversions using regional infrasound data is possible when accurate weather models are available Seismic and acoustic records indicate a strong collapse at shallow depth of 1  ±  0.5 km for the 2020 Kiruna minequake Focal mechanisms and depths of shallow seismic sources can be retrieved from local infrasound records Inversions using regional infrasound data is possible when accurate weather models are available

The variability in obtaining estimates of tsunami inundation and runup on a near‐real‐time tsunami hazard assessment setting is evaluated. To this end, 19 different source models of the Maule Earthquake were considered as if they represented the best available knowledge an early tsunami warning system could consider. Results show that large variability can be observed in both coseismic deformation and tsunami variables such as inundated area and maximum runup. This suggests that using single source model solutions might not be appropriate unless categorical thresholds are used. Nevertheless, the tsunami forecast obtained from aggregating all source models is in good agreement with observed quantities, suggesting that the development of seismic source inversion techniques in a Bayesian framework or generating stochastic finite fault models from a reference inversion solution could be a viable way of dealing with epistemic uncertainties in the framework of nearly‐real‐time tsunami hazard mapping. Owing to recent advancements in rapid seismic source characterization and tsunami simulation, nearly‐real‐time tsunami hazard forecasts in the framework of tsunami early warning systems are starting to be within reach. However, in this study we bring a note of caution regarding its future operational implementation since the level of uncertainty associated to a single rupture inversion is high and thus calls for the use of multiple realizations of seismic inversions to forecast inundation maps and assess their uncertainty bounds. We provide an empirical assessment of the variability of finite fault inversions and its consequences for rapid tsunami hazard mappingThe variability in modeled tsunami maximum amplitudes, runups, and inundated 16 areas is of the same order as their median values evidencing a high inherent uncertaintyCombining the available knowledge to build a distribution of the tsunami source allows to obtain sound predictions of tsunami variables

To explain how fracture roughness affects seismic source and transport response of deformed fractured rock, we developed a microscale fluid flow‐geomechanics‐seismicity model. The modeling method considers comprehensive grains and cement interactions. Fluid flow behavior is obtained through realistic network models of the pore space in the compacted assembly. In addition, forces and displacements in the grains involved in the bond breakage are measured to determine seismic moment tensor. The results of our work show that in addition to stress conditions in the target formation, geological properties of preexisting fractures affect volumetric deformation of the seismic source as well. The results of the model proves that roughness of fractures, applied in a Berea sample, causes deviation of source mechanism of acoustic emission events toward opening tensile part and enhances the permeability of the sample significantly during its failure even under confining pressure. Hydro‐mechanical‐seismic model is applied to study fracture deformationVolumetric deformation of fracture is affected by its geological propertiesFracture roughness affect its transport and seismic source responses

We compared spatiotemporal slip‐rate and high‐frequency (around 1 Hz) radiation distributions from teleseismic Pwave data to infer the seismic rupture process of the 2015 Gorkha, Nepal, earthquake. For these estimates, we applied a novel waveform inversion formulation that mitigates the effect of Green's functions uncertainty and a hybrid backprojection method that mitigates contamination by depth phases. Our model showed that the dynamic rupture front propagated eastward from the hypocenter at 3.0 km/s and triggered a large‐slip event centered about 50 km to the east. It also showed that the large‐slip event included a rapid rupture acceleration event and an irregular deceleration of rupture propagation before the rupture termination. Heterogeneity of the stress drop or fracture energy in the eastern part of the rupture area, where aftershock activity was high, inhibited rupture growth. High‐frequency radiation sources tended to be in the deeper part of the large‐slip area, which suggests that heterogeneity of the stress drop or fracture energy there may have contributed to the damage in and around Kathmandu. Novel formulations were used to obtain the 2015 Nepal earthquake rupture processDynamic rupture propagated eastward from the hypocenter to KathmanduHeterogeneity of the stress drop or fracture energy inhibited rupture growth

Most seismological analysis methods require knowledge of the geographic location of the stations comprising a seismic network. However, common machine learning tools used in seismology do not account for this spatial information, and so there is an underutilized potential for improving the performance of machine learning models. In this work, we propose a graph neural network (GNN) approach that explicitly incorporates and leverages spatial information for the task of seismic source characterization (specifically, location and magnitude estimation), based on multistation waveform recordings. Even using a modestly‐sized GNN, we achieve model prediction accuracy that outperforms methods that are agnostic to station locations. Moreover, the proposed method is flexible to the number of seismic stations included in the analysis and is invariant to the order in which the stations are arranged, which opens up new applications in the automation of seismological tasks and in earthquake early warning systems. To determine the location and size of earthquakes, seismologists use the geographic locations of the seismic stations that record the ground shaking in their data analysis workflow. By taking the distance between stations and the relative timing of the onset of the shaking, the origin of the seismic waves can be accurately reconstructed. In recent years, machine learning (a subfield of artificial intelligence) has shown great potential to automate seismological tasks, such as earthquake source localization. Most machine learning methods do not take into consideration the geographic locations of the seismic stations, and so the usefulness of these methods could still be improved by providing the locations at which the data were recorded. In this work, we propose a method that accounts for geographic locations of the seismic stations, and we show that this improves the machine learning predictions. We propose a deep learning approach for automated earthquake location and magnitude estimation based on graph neural network theoryThis new approach processes multistation waveforms and incorporates station locations explicitlyIncluding station locations improves the accuracy of epicenter estimation compared to models that are location agnostic

On New Year's Day 2024, a magnitude 7.6 event struck the Noto Peninsula in central Japan. Prior to this event, an intense earthquake swarm had persisted beneath the northeastern peninsula for more than five years. Geophysical evidence provides insight into the upwelling of deep fluids from the uppermost mantle that triggers the seismic swarm activity. The noble gases and their isotopes have been used as geochemical indicators to determine the origin of the fluids associated with the swarms and their upwelling. Gas samples collected from boreholes around the seismic source region are characterized by anomalously high 3He/4He ratios (∼3.9 RAcor), indicating infiltration of mantle fluids from the subcrustal lithosphere. Using a steady‐state advection model, we calculated mantle helium fluxes of 1.1–2.4 × 10−15mol cm−2a−1, similar to those estimated for other representative fault zones, such as the San Andreas and North Anatolian faults. In the Noto Peninsula, central Japan, a long‐lasting, intense earthquake swarm has continued since May 2018. On 1 January 2024, the largest M 7.6 earthquake to date occurred. Geophysical evidence provides insight into the upwelling of deep fluids from the uppermost mantle that triggers the seismic swarm activity. The origin of fluids associated with earthquake swarms and their upward fluxes were determined using noble gases and their isotopes. We collected gas samples from hot spring wells around the seismic source region of earthquake swarms. The observed 3He/4He ratios several times higher than the atmospheric value. This indicates infiltration of mantle fluids derived from the subcrustal lithosphere. In addition, we calculated mantle helium fluxes of ∼2.4 × 10−15mol cm−2a−1, which is consistent with estimates for other major fault zones around the world. Before the M 7.6 earthquake that occurred in the Noto Peninsula on 1 January, there was a persistent earthquake swarm for over five yearsElevated 3He/4He ratios in gas samples from hot spring wells suggest the emission of mantle‐derived helium from the seismic source regionThe mantle fluids infiltrate the seismogenic crust and then diffuse in the fault zone, triggering the earthquake swarm Before the M 7.6 earthquake that occurred in the Noto Peninsula on 1 January, there was a persistent earthquake swarm for over five years Elevated 3He/4He ratios in gas samples from hot spring wells suggest the emission of mantle‐derived helium from the seismic source region The mantle fluids infiltrate the seismogenic crust and then diffuse in the fault zone, triggering the earthquake swarm

Laboratory experiments report that detectable seismic velocity changes should occur in the vicinity of fault zones prior to earthquakes. However, operating permanent active seismic sources to monitor natural faults at seismogenic depth is found to be nearly impossible to achieve. We show that seismic noise generated by vehicle traffic, and especially heavy freight trains, can be turned into a powerful repetitive seismic source to continuously probe the Earth's crust at a few kilometers depth. Results of an exploratory seismic experiment in Southern California demonstrate that correlations of train‐generated seismic signals allow daily reconstruction of direct Pbody waves probing the San Jacinto Fault down to 4‐km depth. This new approach may facilitate monitoring most of the San Andreas Fault system using the railway and highway network of California. Even though laboratory experiments report that they should be preceded by detectable precursors, earthquakes remain unpredictable. Indeed, contrary to the lab, scanning natural faults at a few kilometers depth where earthquakes initiate requires operating high‐energy seismic sources continuously in time, which is found to be nearly impossible. In this study, we show that large freight trains generate sufficient seismic energy to travel down to a few kilometers depth and be detected at tens of kilometers from railways. We demonstrate that we can turn this apparently random source of seismic signal into an impulsive virtual seismic source to monitor active faults. We finally estimate that this new approach can be used for monitoring most of the San Andreas Fault system using the railway and highway network of California. Freight trains in Southern California are locally equivalent to a magnitude 2.2 earthquake every dayWe use train noise to reconstruct repetitive virtual sources of Pwaves crossing the San Jacinto Fault at 4‐km depthPredictions of train noise across California show potential for passive monitoring of most of the San Andreas Fault system

North Korea's 3 September 2017 nuclear test was followed by several small seismic events, with one eight‐and‐a‐half minutes after the test and three on and after 23 September 2017. Seismic analysis reveals that the first event is a near vertical on‐site collapse toward the nuclear test center from 440 ±260 m northwest of the test site, with its seismic source best represented by a single force with a dip angle of 70°–75° and an azimuth of ~150°, and the later events are an earthquake swarm located 8.4 ±1.7 km north of the test site within a region of 520 m, with a focal depth of at least 2.4 km and a focal mechanism of nearly pure strike slip along the north‐south direction with a high dip angle of 50°–90°. The occurrence of the on‐site collapse calls for continued monitoring of any leaks of radioactive materials from the test site. On 3 September 2017, the Democratic People's Republic of Korea (North Korea) announced that it had successfully conducted a thermonuclear (hydrogen bomb) test. The nuclear test was followed by several reported small seismic events occurring in the region. While there have been intense interests and concerns about what has happened after the nuclear test among the scientific community, governmental agencies, and the public, the nature of those events is unknown. In this study, we present high‐precision relative locations and detailed source characteristics of the small events following the test. Seismic analysis reveals that the first event occurring eight‐and‐a‐half minutes after the test is a near vertical on‐site collapse toward the nuclear test center from 440 ±260 m northwest of the nuclear test site, with its seismic source best represented by a near vertical single force, and the later three events on and after 23 September 2017 are an earthquake swarm located 8.4 ±1.7 km north of the nuclear test site within a region of 520 m, with an event depth of at least 2.4 km. The occurrence of the on‐site collapse calls for continued monitoring of any leaks of radioactive materials from the test site. We present high‐precision locations and detailed source information of four small seismic events following North Korea's 2017 nuclear testThe first event is an on‐site collapse 440 m northwest of the test site and best explained by a near vertical single forceThe later events are an earthquake swarm located 8.4 km north of the test site with a focal depth of at least 2.4 km

Estimating an earthquake's magnitude and location may not be necessary to predict shaking in real time; instead, wavefield‐based approaches predict shaking with few assumptions about the seismic source. Here, we introduce GRAph Prediction of Earthquake Shaking (GRAPES), a deep learning model trained to characterize and propagate earthquake shaking across a seismic network. We show that GRAPES’ internal activations, which we call “seismic vectors”, correspond to the arrival of distinct seismic phases. GRAPES builds upon recent deep learning models applied to earthquake early warning by allowing for continuous ground motion prediction with seismic networks of all sizes. While trained on earthquakes recorded in Japan, we show that GRAPES, without modification, outperforms the ShakeAlert earthquake early warning system on the 2019 M7.1 Ridgecrest, CA earthquake. Have you ever heard something through the grapevine? It often takes you by surprise to hear a message from someone other than the original source. You might have felt an earthquake in a similar way: experiencing shaking (the message) at your location rather than movement along a fault (the source). We apply grapevine‐style communication to earthquake early warning (EEW). The goal of EEW is to warn people to prepare for earthquake shaking before damaging seismic waves arrive at their location. We build on recent work that used deep learning and large earthquake data sets to predict earthquake shaking. We developed a deep learning algorithm called GRAPES which predicts shaking in a manner similar to a game of seismic telephone: when a seismic sensor detects shaking, it sends a message to its neighboring sensors, warning them to expect shaking soon. These sensors then pass on the message to their more distant neighbors along the grapevine. We show that the messages GRAPES learned to send between sensors are like seismic status updates: “I'm seeing this type of seismic wave right now”. We applied GRAPES to the 2019 M7.1 Ridgecrest, CA earthquake and it predicted shaking accurately and quickly. A deep learning network trained to predict ground motion learned an internal representation of the seismic wavefieldIndividual neurons within the network activate with the arrival of P waves, S waves, surface waves, coda waves, and ambient noiseWhile trained on earthquakes in Japan, the model generalizes well to predicting ground motions for the 2019 Ridgecrest, CA earthquake A deep learning network trained to predict ground motion learned an internal representation of the seismic wavefield Individual neurons within the network activate with the arrival of P waves, S waves, surface waves, coda waves, and ambient noise While trained on earthquakes in Japan, the model generalizes well to predicting ground motions for the 2019 Ridgecrest, CA earthquake

Tremors are a type of slow earthquake with long‐duration signals compared to ordinary earthquakes. The long signals have been considered to solely reflect their long source process. However, here, we provide evidence suggesting that the source processes of tremors are not always long. We refer to these observations as short‐duration tremors. They were recorded by ocean‐bottom seismometers placed very close to the source. Although these tremors exhibit a short‐duration signal when recorded near the source, they exhibit a typical long‐duration signal elsewhere. Our numerical simulations demonstrate that the features can be captured by considering a strongly scattering medium around their source. One such structure could be small low‐velocity inclusions distributed around the seismic source. The inclusions may represent the seismic expression of geologically detected aquifers in tremor source regions. Furthermore, this medium could be embedded along the slow earthquake fault zone and play a critical role in their source process. At least two types of earthquakes occur in shallow subduction zones: ordinary earthquakes and tremors. Tremors are known to exhibit long signal duration compared to ordinary earthquakes. To date, tremors' long‐duration signal has been solely interpreted by their source process. Here, we discovered tremors that exhibit short duration signals when recorded close from the source which we referred to as "short‐duration tremors". They suggest that tremors' source process is not always long and structural effects may partially form the typical long‐duration signals. We performed numerical simulations on elastic wave propagation and demonstrated that the observations can be qualitatively reproduced by assuming a strongly scattering material surrounding the seismic source. On the other hand, it has been reported based on ocean bottom drilling project that tremor source region may consist of patchily distributed aquifers. The inclusions in our model may be the seismic expressions of the geologically detected aquifers. Further, such a structure could be embedded along the slow‐earthquake fault zone and play a key role in their source process. Short‐duration tremors in ocean‐bottom seismometer records suggest that the source process of tremors is not always longShort‐duration tremors can be interpreted by placing a strongly scattering medium around their sourceSuch a medium embedded along the slow earthquake fault zone could play a key role in the tremor source process Short‐duration tremors in ocean‐bottom seismometer records suggest that the source process of tremors is not always long Short‐duration tremors can be interpreted by placing a strongly scattering medium around their source Such a medium embedded along the slow earthquake fault zone could play a key role in the tremor source process

As rivers transport water and sediment across Earth's surface, they radiate elastic and acoustic waves. We use seismic and infrasound observations during a controlled flood experiment (CFE) in the Grand Canyon to show that three types of fluvial processes can be monitored from outside the channel. First, bed‐load transport under conditions of evolving bed mobility is identified as the dominant seismic source between 15 and 45 Hz. Two lower‐frequency seismic bands also excited by the CFE exhibited greater power increases and are consistent with source processes related to fluid rather than sediment transport. The second fluvial seismic source is inferred to be fluid tractions on the rough riverbed, which drive the maximum seismic power increase at 0.73 Hz, but do not excite infrasound. Waves at the fluid‐air interface are suggested as a third source, which generates a common 6–7 Hz peak in seismic and infrasound responses to the CFE. Seismic detection of pulsed bed‐load transportIsolation of the local bed‐load seismic power spectrumDetection of seismic power generated by two distinct fluid transport sources

Whether the final properties of large earthquakes can be inferred from initial observations of rupture (deterministic rupture) is valuable for understanding earthquake source processes and is critical for operational earthquake and tsunami early warning. Initial (P‐wave) characteristics of small to moderate earthquakes scale with magnitude, yet observations of large to great earthquakes saturate, resulting in magnitude underestimation. Whether saturation is inherent to earthquake dynamics or rather is due to unreliable observation of long‐period signals with inertial seismic instrumentation is unclear. Seismogeodetic methods are better suited for broadband observation of large events in the near‐field. In this study, we investigate the deterministic potential of seismogeodetically derived P‐wave amplitude using a dataset of 14 medium‐to‐great earthquakes around Japan. Our results indicate that seismogeodetic P‐wave amplitude is not a reliable predictor of magnitude, opposing the notion of strong determinism in the first few seconds of rupture. Rapid estimation of an earthquake's magnitude is key for the effectiveness of earthquake early warning. Large magnitude earthquakes can take minutes to fully evolve; whether information from the first few seconds of a seismogram is indicative of the final magnitude is therefore of significant interest for early warning operations as well as general understanding of seismic source processes. This study investigates these early rupture observations using a combination of seismic and GPS instrumentation. We find that observation of the first few seconds of rupture is not sufficient to reliably estimate the magnitude of an earthquake greater than magnitude 7.5. Seismogeodesy provides a more accurate P‐wave amplitude than high‐pass filtered seismic‐only methods due to its broadband sensitivity.Magnitude saturation from P‐wave characteristics is inherent to large earthquake dynamics and not due to instrumental limitations.P‐wave amplitude measurements are not predictive of final magnitude for medium‐to‐large magnitude earthquakes.

On 12 November 2017, a MW=7.3 earthquake struck near the Iranian town of Ezgeleh, at the Iran‐Iraq border. This event was located within the Zagros fold and thrust belt which delimits the continental collision between the Arabian and Eurasian Plates. Despite a high seismic risk, the seismogenic behavior of the complex network of active faults is not well documented in this area due to the long recurrence interval of large earthquakes. In this study, we jointly invert interferometric synthetic aperture radar and near‐field strong motions to infer a kinematic slip model of the rupture. The incorporation of these near‐field observations enables a fine resolution of the kinematic rupture process. It reveals an impulsive seismic source with a strong southward rupture directivity, consistent with significant damage south of the epicenter. We also show that the slip direction does not match plate convergence, implying that some of the accumulated strain must be partitioned onto other faults. Iran is a very seismically active region. However, the 2017 Ezgeleh earthquake of magnitude 7.3 occurred in a region where large earthquakes have not been documented for several centuries. Our knowledge of fault locations, geometries, and seismic behaviors is therefore limited in this region. We use near‐field seismological and satellite geodetic data to retrieve the spatial and temporal distribution of slip occurring on the fault during the Ezgeleh earthquake. We show that the high slip rate and Southward directivity of the rupture may have worsened the damage south of the epicenter. We also observe that tectonic motion is partitioned between different types of faults. Although the Ezgeleh earthquake did release a significant part of that strain, other seismogenic faults in the region could represent an important hazard for the nearby population. The Ezgeleh earthquake ruptured a flat thrust fault in the Zagros fold and thrust beltKinematic slip modeling reveals a highly impulsive source with southward directivity, possibly causing the large damage in the areaThe direction of coseismic slip suggests a strain partitioning between thrust and unmapped strike‐slip faults

We conducted a time‐lapse seismic experiment utilizing automated active seismic source and sensor arrays to monitor a reactivated fault within the Opalinus clay formation at the Mont Terri Rock Laboratory (Switzerland), an analog caprock for geologic carbon storage. A series of six brine injections were conducted into the so‐called Main Fault to reactivate it. Seismic instrumentation in five monitoring boreholes on either side of the fault was used to continuously probe changes in P‐wave travel‐times associated with fault displacement and leakage. We performed time‐lapse travel‐time tomography on five hundred sequential data sets; this revealed a zone of decreased P‐wave velocity, up to 16 m/s, during each injection cycle, followed by a velocity increase during shut‐in. These results demonstrate varying elastic property perturbations, both spatially and temporally, along the fault plane during reactivation. We then interpreted these velocity changes in terms of fault dilation induced by pressurized fluids along the fault. Faults within clay formation caprocks for CO2storage reservoirs are possible pathways for leakage and loss of containment. Understanding how these faults in clay‐rich rocks reactivate and leak fluids is important for predicting, detecting, and preventing CO2movement. Passive seismic monitoring is challenging because of the lack of observable seismic events in such clay‐rich fault rupture. In this study, we measure changes in P‐wave velocity to monitor a fault reactivated by brine injections directly into the Main Fault at the Mont Terri Rock Laboratory, Switzerland. We use a recently developed time‐lapse seismic technique called Continuous Active‐Source Seismic Monitoring (CASSM), which allows us to make these measurements within a few minutes and observe small changes on the same timescale. We relate the measured changes in P‐wave velocity to the opening of that fault damage zone by using a rock physics model, which helps explain changes in permeability within the fault zone. measuring p‐wave velocity changes during fault reactivationmonitoring fault reactivation in an analog caprock for geologic carbon storagefracture damage zone modeling from p‐wave velocities measuring p‐wave velocity changes during fault reactivation monitoring fault reactivation in an analog caprock for geologic carbon storage fracture damage zone modeling from p‐wave velocities

In geophysical inverse problems, the distribution of physical properties in an Earth model is inferred from a set of measured data. A necessary step is to select data that are best suited to the problem at hand. This step is performed ahead of solving the inverse problem, generally on the basis of expert knowledge. However, expert‐opinion can introduce bias based on pre‐conceptions. Here we apply a trans‐dimensional algorithm to automatically weigh data on the basis of how consistent they are with the fundamental hypotheses made to solve the inverse problem. We demonstrate this approach by inverting arrival times for the location of a seismic source in an elastic half‐space, assuming a point‐source and uniform weights in concentric shells. The key advantage is that the data do no longer need to be selected by an expert, but they are assigned varying weights during the inversion procedure. In the Big data era, automated approaches to data evaluation are needed for two main reasons: to be able to process a large amount of data in a limited time, and to avoid bias introduced by data analysists. In this study we present a novel approach to data analysis, where the data themselves measure their consistency with our hypotheses. The approach is applied to earthquake location in mines, where millions of seismic events occur every year, and automatic processing of seismic data is mandatory. We demonstrate that our approach outperforms standard ones when almost nothing is known about the data and their measurement errors. We develop a novel approach for automatic weighting of data in geophysical inverse problems, based on a trans‐dimensional algorithmWe apply the novel approach to seismic event location in mines, obtaining consistent results compared to a more standard methodOur approach outperforms standard seismic monitoring approaches, when limited information are available on local seismic structure We develop a novel approach for automatic weighting of data in geophysical inverse problems, based on a trans‐dimensional algorithm We apply the novel approach to seismic event location in mines, obtaining consistent results compared to a more standard method Our approach outperforms standard seismic monitoring approaches, when limited information are available on local seismic structure

Primary microseisms are background seismic oscillations recorded everywhere on Earth with typical frequencies 0.05 < f< 0.1 Hz. They appear to be generated by ocean waves of the same frequency f, propagating over shallow bottom topography. Previous quantitative models for the generation of primary microseisms considered wave propagation over topographic features with either large scales, equivalent to a vertical point force, or small scales matching ocean wave wavelengths, equivalent to a horizontal force. While the first requires unrealistic bottom slopes to explain measured Rayleigh wave amplitudes, the second produced Love waves and not enough Rayleigh waves. Here we show how the small scales actually produce comparable horizontal and vertical forces. For example, a realistic rough bottom over an area of 100 km2with depths around 15 m is enough to explain the vertical ground motion observed at a seismic station located 150 km away. Ocean waves propagating over small‐scale topography is thus a plausible explanation for the observed microseisms at frequencies around 0.07 Hz. Microseisms are background oscillations of the solid Earth. Most of these oscillations are caused by ocean waves and can thus be used to study their source, the ocean waves, or the medium in which they propagate, the solid Earth. Several theories have been proposed for how ocean waves going over shallow ocean topography make microseisms in the band of periods 10 to 20 s, but they are not satisfactory because they either require unrealistic large slopes of the ocean floor or they produce a ratio of different types of seismic waves, Love and Rayleigh waves, that is too large. We thus revise these theories to show that a plausible seismic source is the propagation ocean of waves over a wavy bottom, when the bottom has wavelengths that match those of ocean wave. We particularly verify that the predicted Rayleigh wave amplitude is of the order of what is measured at a particular seismometer located in Ireland. Because the necessary details in bottom topography vary a lot between different ocean regions, the new theory suggests that the spatial distribution of seismic sources is more heterogeneous than previously thought. Ocean waves over small‐scale topography produce both horizontal and vertical equivalent point forcesSmall‐scale wavy bottoms are stronger sources of primary microseism than constant slopesThe vertical force, caused by pressure modulations, is generally weaker than the horizontal force

Liu et al. (2022, https://doi.org/10.1029/2021GL093691) used Rayleigh waves extracted from the cross‐correlation of ambient noise recorded by two stations to monitor the seismic velocity variations associated with the 2018 Kı̄lauea eruption. However, their study ignored the fact that the tremors on the Island of Hawai'i were dominated by a source at the Kı̄lauea summit before the eruption. Close inspection of the waveforms of the station pair PAUD‐STCD shows a simple, mistakenly identified wave traveling direction in Liu et al. (2022, https://doi.org/10.1029/2021GL093691). A correct wave traveling direction agrees with the noise source model, where the dominant tremor source should be at the Kı̄lauea summit. Because of the drastic change in the tremor source after the eruption, the cross‐correlation of the tremor records may reflect predominantly changes in the source rather than in the medium properties between the two stations. More and more studies are using seismic records to understand the volcano eruption process. A very recent study monitored the dyke intrusion near the Pu‘u‘ō‘ō right after the start of the 2018 Kı̄lauea eruption, by observing increased seismic wave arrival times. However, after re‐inspecting the data, we find their study understood the source location emitting the seismic waves wrong. Thus, their observed increased seismic wave arrival times can no longer be explained by the proposed dyke intrusion in the rift zone. We further suspect that the temporal variations of the seismic source contributed greatly to the arrival time variations. The tremor signals appeared on the ambient noise cross‐correlation functions (CCFs) should be from the Kı̄lauea summit, not Pu‘u‘ō‘ōThe tremor sources changed after the beginning of the eruptionThe apparent temporal variations of the CCFs related to the tremors reflected the change of noise source rather than medium velocity The tremor signals appeared on the ambient noise cross‐correlation functions (CCFs) should be from the Kı̄lauea summit, not Pu‘u‘ō‘ō The tremor sources changed after the beginning of the eruption The apparent temporal variations of the CCFs related to the tremors reflected the change of noise source rather than medium velocity

Using P‐wave seismograms, we trained a seismic source classifier using a Convolutional Neural Network. We trained for three classes: earthquake P‐wave, underground nuclear explosion (UNE) P‐wave, and noise. With the current absence of nuclear testing by countries that have signed the Comprehensive Test Ban Treaty, high quality seismic data from UNEs is limited. Even with limited training data, our model can accurately characterize most events recorded at regional and teleseismic distances, finding over 95% signals in the validation set. We applied the model on holdout datasets of the North Korean test explosions to evaluate the performance on unique region and station‐source pairs, with promising results. Additionally, we tested on the Source Physics Experiment events to investigate the potential for chemical explosions to act as a surrogate for nuclear explosions. We anticipate that machine‐learning models like our classifier system can have broad application for other seismic signals including volcanic and non‐volcanic tremor, anomalous earthquakes, ice‐quakes or landslide‐quakes. We train a global seismic event classifier using machine learning on underground nuclear test explosion seismic data. Our classifier model can successfully discriminate (with over 95% accuracy) between underground nuclear explosion, earthquake, and noise signals from stations both regionally and far‐field. Since this model was trained on a relatively small data set (for machine learning applications) we expect that similar methods can be applied to event or discrimination of other unique seismic sources like those from volcanoes, landslides, or glaciers. We successfully discriminate underground nuclear explosions with a Convolutional Neural network (CNN) trained on P‐wave seismogramsRobust global seismic event discrimination is possible with machine learning trained on regional and teleseismic dataA CNN trained with historical nuclear explosion data can be applied with high accuracy to other regions, like the six Democratic People's Republic of Korea's test explosions We successfully discriminate underground nuclear explosions with a Convolutional Neural network (CNN) trained on P‐wave seismograms Robust global seismic event discrimination is possible with machine learning trained on regional and teleseismic data A CNN trained with historical nuclear explosion data can be applied with high accuracy to other regions, like the six Democratic People's Republic of Korea's test explosions

On 4 May 2022, the seismometer on Mars observed the largest marsquake (S1222a) during its operation. One of the most specific features of S1222a is the long event duration lasting more than 8 hr, in addition to the clear appearance of body and surface waves. As demonstrated on Earth, by modeling a long‐lasting and scattered surface wave with the radiative transfer theory under the isotropic scattering condition, we estimated the scattering and intrinsic quality factors of Mars (Qsand Qi). This study especially focused on the frequency range between 0.05–0.09 Hz, where Qsand Qihave not been constrained yet. Our results revealed that Qi= 1,000–1,500 and Qs= 30–500. By summarizing the Martian Qiand Qsestimated so far and by comparing them with those of other celestial bodies, we found that, overall, the Martian scattering and absorption properties showed Earth‐like values. Since February 2019, NASA's InSight (Interior Exploration using Seismic Investigations, Geodesy, and Heat Transport) has been conducting quasi‐continuous seismic observation for more than three years. The seismic data from Mars has contributed significantly to a better understanding of the interior structure and the seismicity of the red planet. On 4 May 2022 (1222 Martian days after landing), another key event occurred, called S1222a. The event showed the largest seismic moment release (magnitude 4.7) and extremely long duration (>8 hr) with intense seismic scattering. As demonstrated on Earth, the long‐lasting scattered waves are useful for retrieving information about the structural heterogeneity within a planet. In this study, by applying the radiative transfer theory—which considers the energy transportation from the seismic source to the observation point—to Mars, we evaluated the energy decay rate due to seismic scattering and energy absorption by a medium. By comparing our results with those of other solid bodies, we found that the Martian scattering and absorption features were closer to the terrestrial ones than to the lunar ones. We modeled the scattering effect of the largest marsquake (S1222a) using radiative transfer theory on a spherical MarsThe inversion revealed that the intrinsic and scattering quality factors below 0.1 Hz are 1,000–1,500 and 30–500, respectivelyWe summarized the Martian quality factors derived so far and found that they are relatively Earth‐like rather than Moon‐like We modeled the scattering effect of the largest marsquake (S1222a) using radiative transfer theory on a spherical Mars The inversion revealed that the intrinsic and scattering quality factors below 0.1 Hz are 1,000–1,500 and 30–500, respectively We summarized the Martian quality factors derived so far and found that they are relatively Earth‐like rather than Moon‐like

The prompt identification of faults responsible for moderate‐to‐large earthquakes is fundamental for understanding the likelihood of further, potentially damaging events. This is increasingly challenging when the activated fault is an offshore buried thrust, where neither coseismic surface ruptures nor GPS/InSAR deformation data are available after an earthquake. We show that on 9 November 2022, an Mw 5.5 earthquake offshore Pesaro ruptured a portion of the buried Northern Apennines thrust front (the Cornelia thrust system [CTS]). By post‐processing and interpreting the seismic reflection profiles crossing this thrust system, we determined that the activated fault (CTS) is an arcuate 30‐km‐long, NW‐SE striking, SW dipping thrust and that older structures at its footwall possibly influenced its position and geometry. The activation of adjacent segments of the thrust system is a plausible scenario that deserves to be further investigated to understand the full earthquake potential of this offshore seismogenic source. The Northern Apennines chain is characterized by thrust faults running from the Po Plain to the Adriatic Sea on the northeastern side of peninsular Italy. These thrusts are buried below ≈2,000 m cover of Plio‐Pleistocene deposits. Controversies arose about these thrust faults' activity and earthquake potential based on their hidden geological signature and the scanty seismicity that could be associated with them. The earthquake (magnitude 5.5) that occurred on 9 November 2022, offshore Pesaro revived this argument. In this work, we analyze the geological structure of the crustal volume affected by the seismic sequence, exploiting seismic reflection profiles and well‐log data to identify the earthquake causative fault. Our results demonstrate that the earthquake ruptured a well‐known fault of the Northern Apennines' buried thrust front, supporting that it is indeed active and seismogenic. The size and architecture of this thrust front suggest that it could generate even larger earthquakes (Mw > 6.5). This type of geological study is instrumental to understanding the geometry of earthquake faults, particularly in offshore areas, because they constitute reliable inputs for earthquake hazard models and, when done promptly after an earthquake, provide key elements for other studies on the seismic source and the unfolding of the ongoing seismic sequence. 9 November 2022, earthquake consistent with activity of the Cornelia thrust, a fault system running off the central Adriatic coastThe seismic reflection profiles in the area allowed for delineating the thrust and its earthquake potential with a much finer resolutionThe properties of the causative fault suggest that the activation of adjacent segments is a plausible scenario that deserves consideration 9 November 2022, earthquake consistent with activity of the Cornelia thrust, a fault system running off the central Adriatic coast The seismic reflection profiles in the area allowed for delineating the thrust and its earthquake potential with a much finer resolution The properties of the causative fault suggest that the activation of adjacent segments is a plausible scenario that deserves consideration

Forecasting hanging glacier instabilities remain challenging as sensing technology focusing on the ice surface fails to detect englacial damage leading to large‐scale failure. Here, we combine icequake cluster analysis with coda wave interferometry constraining damage growth on Switzerland's Eiger hanging glacier before a 15,000 m3break‐off event. The method focuses on icequake migration within clusters rather than previously proposed “event counting.” Results show that one cluster originated from the glacier front and migrated by 13.9(±1.2) m within 5 weeks before the break‐off event. The corresponding crevasse extension separates unstable and stable ice masses. We use the measured source displacement for damage parametrization and find a 90% agreement between an analytical model based on damage mechanics and frontal flow velocities measured with an interferometric radar. Our analysis provides observational constraints for damage growth, which to date is primarily a theoretical concept for modeling englacial fractures. Predicting the development of ice mass breaking off from unstable glaciers is challenging. Such glaciers are often located in remote places with steep terrain, making it difficult to gather observations. Because of that, our current view on ice damage development on unstable glaciers is incomplete. Here, we propose a new approach to tackle this problem using seismic observations from the Eiger hanging glacier in the Swiss Alps before a moderate 15,000 m3break‐off event. We first group seismic signals according to their similarity. We then use waves scattered by fractures within the ice and at the glacier‐rock boundaries to track displacement of ice damage. Our results suggest that a group of seismic signals is generated by crevasse propagation separating the unstable ice mass and the stable ice uphill. Combined with a simple analytical model, these observations indicate that seismic source displacement can be associated with damage propagation within unstable glacier ice. This is an important step toward a better understanding of unstable ice flow and forecasting glacier instabilities. In the months leading up to a break‐off event, we find thousands of recurring icequakes on the Eiger hanging glacier in SwitzerlandCoda wave interferometry resolves displacements of icequake multipletsOne multiplet displacement represents englacial damage corresponding to crevasse extension between unstable and stable ice masses In the months leading up to a break‐off event, we find thousands of recurring icequakes on the Eiger hanging glacier in Switzerland Coda wave interferometry resolves displacements of icequake multiplets One multiplet displacement represents englacial damage corresponding to crevasse extension between unstable and stable ice masses

Likelihood testing of induced earthquakes in Oklahoma and Kansas has identified the parameters that optimize the forecasting ability of smoothed seismicity models and quantified the recent temporal stability of the spatial seismicity patterns. Use of the most recent 1 year period of earthquake data and use of 10–20 km smoothing distances produced the greatest likelihood. The likelihood that the locations of January–June 2015 earthquakes were consistent with optimized forecasts decayed with increasing elapsed time between the catalogs used for model development and testing. Likelihood tests with two additional sets of earthquakes from 2014 exhibit a strong sensitivity of the rate of decay to the smoothing distance. Marked reductions in likelihood are caused by the nonstationarity of the induced earthquake locations. Our results indicate a multiple‐fold benefit from smoothed seismicity models in developing short‐term earthquake rate forecasts for induced earthquakes in Oklahoma and Kansas, relative to the use of seismic source zones. Optimized spatial distribution forecasts use epicenter locations from the previous year and kernel smoothing distances of 10–20 kmProbability gains achieved from the forecasts of induced earthquake locations decay over the course of about 6–12 monthsThis work was motivated by the development of a short‐term seismic hazard model that includes the effects of induced earthquakes

Large, compressional earthquakes cause surface uplift as well as widespread mass wasting. Knowledge of their trade‐off is fragmentary. Combining a seismologically consistent model of earthquake‐triggered landsliding and an analytical solution of coseismic surface displacement, we assess how the mass balance of single earthquakes and earthquake sequences depends on fault size and other geophysical parameters. We find that intermediate size earthquakes (Mw6–7.3) may cause more erosion than uplift, controlled primarily by seismic source depth and landscape steepness, and less so by fault dip and rake. Such earthquakes can limit topographic growth, but our model indicates that both smaller and larger earthquakes (Mw< 6, Mw> 7.3) systematically cause mountain building. Earthquake sequences with a Gutenberg‐Richter distribution have a greater tendency to lead to predominant erosion, than repeating earthquakes of the same magnitude, unless a fault can produce earthquakes with Mw> 8 or more. Earthquake mass balance based on a seismologically consistent expression of coseismic landsliding and Okada's surface deformation solutionLarge earthquakes are always constructive, while intermediate ones may be erosive if shallow enough and below a steep enough topographyEarthquake sequences on fault smaller than 120 km may limit topography building especially if they follow a Gutenberg‐Richter distribution

The 25 April 2015 Nepal earthquake (Mw7.8) ruptured a segment of the Himalayan front fault zone. Four absolute gravimetric stations in southern Tibet, surveyed from 2010/2011 to 2013 and corrected for secular variations, recorded up to 22.40 ± 1.11 μGal/yr of gravity increase during this period. The gravity increase is distinct from the long‐wavelength secular trends of gravity decrease over the Tibetan Plateau and may be related to interseismic mass change around the locked plate interface under the Himalayan‐Tibetan Plateau. We modeled the source region as a disk of 580 km in diameter, which is consistent with the notion that much of the southern Tibetan crust is involved in storing strain energy that drives the Himalayan earthquakes. If validated in other regions, high‐precision ground measurements of absolute gravity may provide a useful method for monitoring mass changes in the source regions of potential large earthquakes. Absolute gravimetry in southern Tibet showed clear gravity increase before the 2015 Nepal earthquakeThe gravity increase may be related to strain or mass migration in a broad seismic source regionHigh‐precision absolute gravimetry may become a useful method for monitoring seismicity

The canonical model of fault coupling assumes that slip is partitioned into fixed asperities that display stick‐slip behavior and regions that creep stably. We show that this simple asperity model is inconsistent with GPS‐derived deformation in northern Japan associated with interseismic coupling on the subduction interface and the transient response to Mw6.3–7.2 earthquakes during 2003–2011. Comparisons of GPS data with simulations of earthquakes on asperities and associated velocity‐strengthening afterslip require that afterslip overlaps areas of the fault that ruptured in previous earthquakes, including the 2011 Mw9 Tohoku‐oki earthquake. Whereas about 55% of the plate interface ruptured in earthquakes during 2003–2011, we infer that only 9% of the plate interface was fully locked between earthquakes. Inferred locked asperities are roughly 25% the size of rupture areas determined by seismic source inversions. These smaller asperities are consistent with interseismic strain accumulation in 2009, although more extensive locking is required a decade earlier in 1998. Megathrust modeled as either locked or slipping with rate‐strengthening frictionCoseismic slip and afterslip overlap on the subduction interfaceSimple stationary frictional asperity model is nullified

We propose a tsunami forecasting method based on a data assimilation technique designed for dense tsunameter networks. Rather than using seismic source parameters or initial sea surface height as the initial condition of for a tsunami forecasting, it estimates the current tsunami wavefield (tsunami height and tsunami velocity) in real time by repeatedly assimilating dense tsunami data into a numerical simulation. Numerical experiments were performed using a simple 1‐D station array and the 2‐D layout of the new S‐net tsunameter network around the Japan Trench. Treating a synthetic tsunami calculated by the finite‐difference method as observed data, the data assimilation reproduced the assumed tsunami wavefield before the tsunami struck the coastline. Because the method estimates the full tsunami wavefield, including velocity, these wavefields can be used as initial conditions for other tsunami simulations to calculate inundation or runup for real‐time forecasting. We propose a tsunami forecasting method based on successive data assimilationAssimilation directly estimates tsunami height and velocity without using a sourceSimulation successfully estimates a tsunami using a planned tsunameter network

Seismograms always result from mixing many sources and medium changes that are complex to disentangle, witnessing many physical phenomena within the Earth. With artificial intelligence (AI), we isolate the signature of surface freezing and thawing in continuous seismograms recorded in a noisy urban environment. We perform a hierarchical clustering of the seismograms and identify a pattern that correlates with ground frost periods. We further investigate the fingerprint of this pattern and use it to track the continuous medium change with high accuracy and resolution in time. Our method isolates the effect of the ground frost and describes how it affects the horizontal wavefield. Our findings show how AI‐based strategies can help to identify and understand hidden patterns within seismic data caused either by medium or source changes. Seismic waves, emitted by a seismic source and then traveling through the Earth, contain crucial information about the sources and the medium. However, often multiple sources emit simultaneously, while the elastic properties of the medium can change over time. Unmixing and identifying the different processes in the seismograms is a complex task, which we try to solve with methods of artificial intelligence. In a completely data‐driven fashion, we are able to mute the variation in the seismograms due to anthropogenic seismic sources and reveal a continuous medium change due to freezing and thawing. This approach could reveal hidden information in complex environments such as volcanoes, where many different source and medium processes occur. With methods of unsupervised learning, we identify source and medium processes in seismogramsA data‐driven product of the seismogram tracks a continuous medium change due to freezing and thawing of the surfaceThe data‐driven product can act as a filter and reveal the hidden signature of the medium change With methods of unsupervised learning, we identify source and medium processes in seismograms A data‐driven product of the seismogram tracks a continuous medium change due to freezing and thawing of the surface The data‐driven product can act as a filter and reveal the hidden signature of the medium change

The seismic source spatiotemporal evolution of the Mw6.9 event on 24 May 2014 in Northern Aegean is imaged by backprojection of strong motion envelopes. The results indicate that the event ruptured on two different fault segments. In the first one, rupture propagated from the hypocenter westward for ∼20 km. In the second delayed segment to the east, rupture propagated eastward for ∼65 km with a supershear velocity (∼5.5 km/s). At the end of this rupture the largest stacking amplitudes are imaged, associated with possible stopping phases from the abrupt cessation of a fast slip. Low‐aftershock seismicity on the supershear segment implies a simple and linear fault geometry there. This is the third large event in the Northern Aegean Trough‐Northern Anatolian Fault zones that has ruptured with supershear speed. This characteristic should be taken into account in studying past events and estimating seismic hazard in this area. Two distinct rupture segmentsEastern segment with supershear ruptureHigh frequency mainly radiates from the end of ruptured segments

Permafrost degradation is rapidly increasing in response to a warming Arctic climate, altering landscapes and damaging critical infrastructure. Solutions for monitoring permafrost thaw dynamics are essential to understand biogeochemical feedbacks as well as to issue warnings for hazardous geotechnical conditions. We investigate the feasibility of permafrost monitoring using permanently installed fiber‐optic seismic networks. We conducted a 2‐month seismic monitoring campaign during a controlled thaw experiment using a permanent surface orbital vibrator (SOV) and a 2D‐array of distributed acoustic sensing (DAS) cables, and observed significant (15%) shear‐wave velocity (Vs) reductions and approximately 2 m depression of the permafrost table beneath the heating zone. These observations were validated by time‐lapse horizontal‐to‐vertical spectral ratio (HVSR) analysis from three co‐located broadband seismometers. The combination of SOV and DAS provided unique seismic observations for permafrost monitoring at the field scale, as well as a basis for design and development of early warning systems for permafrost thaw. Increased permafrost thaw, caused by a changing climate, is altering the landscape, increasing carbon dioxide emissions, and causing damage to critical infrastructure. In this context, long‐term monitoring of permafrost thaw processes is vital for hazard prevention and mitigation. Unfortunately, a variety of challenges exist for long‐term time‐lapse monitoring of thaw processes, especially in arctic settings with strong near‐surface rainfall and temperature cycles. To meet the need of inexpensive, rugged, and highly sensitive monitoring technology, we conducted a unique experiment in Fairbanks, Alaska using 4‐km of buried fiber‐optic cables as seismic sensors (called distributed acoustic sensing, DAS) to monitor changes while the permafrost was artificially warmed over a 2‐month period. A permanent seismic source generated seismic waves to repeatedly sample the near‐surface. Our results reveal significant reductions in shear‐wave velocity, indicating a decline in soil rigidity as it warmed, as well as deepening of the top of permafrost. These results suggest permafrost thaw monitoring using DAS may be an effective strategy to track changing arctic landscapes and predict areas of greatest thaw vulnerability. We evaluate a novel approach to seismically monitor permafrost degradation using DAS during a controlled heating experimentJoint interpretation of surface‐wave inversion and HVSR reveals 15% S‐wave velocity reductions due to permafrost degradationTime‐lapse seismic techniques utilizing DAS have the potential to track changes in unfrozen water content at subzero temperatures We evaluate a novel approach to seismically monitor permafrost degradation using DAS during a controlled heating experiment Joint interpretation of surface‐wave inversion and HVSR reveals 15% S‐wave velocity reductions due to permafrost degradation Time‐lapse seismic techniques utilizing DAS have the potential to track changes in unfrozen water content at subzero temperatures

We address geometric and kinematic properties of the Mw6.3 26 November 2019 Durrës earthquake, the strongest earthquake in Albania in the past 40 years. Using coseismic surface displacements from Sentinel‐1 Differential Interferometric Synthetic Aperture Radar (DInSAR) and nearby Global Navigational Satellite System (GNSS) stations, we invert for the geometry and slip of the causative fault. We find that both a steep SW‐dipping fault (dip 71°) and a shallow NE‐dipping fault (dip 15°) can fit the data equally well. However, the slip on the SW‐dipping fault occurs at depths 11–23 km, similar to the depths of the mainshock and aftershock seismicity, and thus, we prefer that model. The location of our preferred fault plane correlates with the mapped SW‐dipping backthrust, the Vore fault. The fault rupture did not reach the surface, which implies that an updip stress propagation onto the unruptured shallow portion of the Vore fault and its secondary structures pose an increased seismic hazard for cities in Albania, including the capital, Tirana. The magnitude 6.3 earthquake near Durrës, Albania, on 26 November 2019 was the largest earthquake in the country for over 40 years. It caused 51 deaths and damaged over 2,000 buildings in Durrës and the capital city Tirana. The earthquake occurred below the surface, and it was not immediately clear in the aftermath which fault it occurred on. We investigated that question using a combination of satellite observation techniques: DInSAR (a radar method that maps movements of the ground in one dimension over the large area) and GNSS (observations of three‐dimensional movements of the ground at specific locations). Out of two possibilities, we prefer a model in which the earthquake occurred on a fault that steeply dips (tilts) to the southwest, between 11‐ and 23‐km depth, agreeing with the depths of the mainshock and aftershocks from seismology. This fault, the Vore fault, is partly mapped at the surface and runs close to Tirana. The upper 11 km of the Vore fault and its hanging wall structures did not move in this earthquake, and therefore, they could still sustain a damaging earthquake in the future, threatening Tirana and other cities in northwestern Albania. Geodetic source modeling of the Mw6.3 2019 Durrës earthquake is performed based on Sentinel‐1 DInSAR and GNSS observationsA SW‐dipping fault‐plane agrees better with the seismic source parameters, depths, and locations of mainshock and aftershocksRupture did not reach the surface, and unruptured part of fault potentially poses elevated seismic hazard for the Albanian capital Tirana

Two repeating compensated linear vector dipole (CLVD) earthquakes occurred in 2009 and 2017 on the Kermadec Ridge near the Curtis submarine volcano. The tsunami and seismic waveforms of both events are almost identical. Simulated tsunami and seismic waveforms are compared with observations of the 2017 event to estimate the location and geometry of the source. A CLVD source model at 8–14 km depth is consistent with the global CMT solution but generates no tsunami. The tsunami waveforms can be well reproduced if the source model is about 1.5 km deep. However, a seismic source at this depth is not consistent with the observed seismic waves. This suggests that two sources were involved with different depths. The main, shallow, tsunami source could be due to hydrofracturing of heated water in a shallow sediment layer triggered by a deeper earthquake. Two different earthquakes that generate almost identical tsunami and seismic waves are extremely rare. However, this occurred in 2009 (Mw6.0) and 2017 (Mw5.9) in the Kermadec Ridge, New Zealand. Both tsunamis were also much larger than expected from their moment tensor solutions. This kind of tsunami is very dangerous because the tsunami observation network is sparse and the tsunami threat could be underpredicted. We use computer codes to simulate the tsunami and seismic waves from the 2017 event and compare a suite of observations to the models to study the source mechanism. First, we use tsunami simulation results to find the exact location of the source. Further tsunami and seismic wave simulations are then used to further estimate the earthquake source model. We find that the observations can best be explained by two sources within the Curtis submarine volcano. The deeper source is responsible for most of the seismic energy while the shallow source appears to be mostly responsible for the tsunami. Tsunami and seismic waveforms suggest that the 2009 and 2017 events are repeating tsunami earthquakes, with similar magnitudes and mechanismThe source of the 2017 tsunami is located at Curtis Island with sea surface displacement peak of 3.0 mThe estimated source model for the 2017 earthquake consists of two sources at 8 km (Mw5.8) and 1.5 km (Mw 5.3) depth

Recent development of dense strong‐motion networks and seismic array processing enables rapid tsunami predictions based on the back‐projection (BP) approach. We develop a multi‐array local BP method (MLBP) using seismic networks with epicentral distance from 0.7° to 3.5°. The local BPs using individual arrays are first calculated and are then merged into a single image of the rupture process. This multi‐array approach circumvents the issue of artifact in single‐array BP caused by the overlapping of multiple phases and coda waves. Based on the local BP approach, tsunami predictions are available 7 min after the origin time. Case studies of the 2003 Mw 8.1 Tokachi‐oki and the 2011 Mw 9.0 Tohoku earthquakes show that their rupture zones are well resolved and are comparable with principal slip areas inferred from tsunami observations. The amplitude and arrival time errors of the predicted tsunami waves are within −1.59 to 3.74 m and −10.0 to 10.0 min, which are sufficiently small for warning purposes. Accurate prediction of tsunami waves for residents who are close to the tsunami source requires fast characterization of tsunami‐genic earthquakes. In this research, we develop a tsunami warning approach based on earthquake source observations using multiple dense clusters of seismic stations installed near the coastal region. At a given stage during an earthquake, the recorded seismic wavefield is back‐propagated to retrieve the direction of the seismic source with respect to each cluster of stations. By combining direction information determined at each station group, we retrieve the principle slip fault area that is most responsible for tsunami generation. A simplified earthquake source model is then constructed to predict the tsunami arrival time and height. By adopting local seismic stations within 80–400 km away from the earthquake sources, we show that the tsunami warning can be issued 7 min after the earthquake origin time. Our case studies of the 2003 Mw 8.1 Tokachi‐oki and the 2011 Mw 9.0 Tohoku earthquakes show that our approach adequately reproduces the tsunami observations. The height errors of the predicted tsunami waves are within −1.59 to 3.74 m; the errors of the arrival times are within −10.0 to 10.0 min, which are sufficiently small for warning purposes. We develop a tsunami warning approach based on a multi‐array back‐projection (BP) method using local seismic networks (0.7°<Δ<3.5°)The multi‐array BP circumvents the issue of artifacts in single‐array BP caused by the overlapping of multiple phases and coda wavesCase studies of the 2003 M8.1 Tokachi and the 2011 M9.0 Tohoku earthquakes demonstrate fast and accurate predictions for tsunami warning

The 4 July 2019 Mw6.4 and subsequent 6 July 2019 Mw7.1 Ridgecrest sequence earthquakes (CA, USA) ruptured orthogonal fault planes in a low slip rate (1 mm/year) dextral fault zone in the area linking the Eastern California Shear Zone and Walker Lane. This region accommodates nearly one fourth of plate boundary motion and has been proposed to be an incipient transform fault system that could eventually become the main tectonic boundary, replacing the San Andreas Fault. We investigate the rupture process of these events using a novel simultaneous kinematic slip method with joint inversion of high‐rate GNSS, strong motion, GNSS static offset, and Interferometric Synthetic Aperture Radar data. We model the Coulomb stress change to evaluate how the Mw6.4 earthquake may have affected the subsequent Mw7.1 event. Our findings suggest complex interactions between several fault structures, including dynamic and static triggering, and provide important context for regional seismic source characterization and hazard models. The San Andreas is a right‐lateral strike‐slip fault marking the main tectonic boundary between the North American and Pacific Plates. East of the San Andreas, a diffuse region of right‐lateral shear known as the Eastern California Shear Zone accommodates roughly one quarter of the motion between the two tectonic plates. The Eastern California Shear Zone has been proposed to be an immature fault system that in the future will accommodate a greater portion of the region's tectonic forces, coalesce into a simple, throughgoing fault, and eventually replace the San Andreas as the major tectonic boundary. The July 2019 Ridgecrest sequence therefore provides the opportunity to gain valuable insights into the development of tectonic boundaries. We analyze the spatial and temporal history of slip during the two largest events in the sequence, the 4 July Mw6.4 and 6 July Mw7.1 earthquakes, and demonstrate a novel approach to evaluate seismic and geodetic observations for both earthquakes simultaneously. We identify complex interactions between discrete fault segments, including dynamic triggering of one fault by slip on another, and present evidence for static triggering of the 6 July Mw7.1 event by the 4 July Mw6.4 earthquake. We present a novel kinematic slip inversion method to model the two largest 2019 Ridgecrest events in a single simultaneous inversion stepWe demonstrate evidence for both static and dynamic triggering within the Ridgecrest sequenceLow rupture velocity and complex fault geometry support the view of this region as an incipient transform fault system

A seismic source model for the Mw8.1 2017 Chiapas, Mexico, earthquake was constructed by kinematic waveform inversion using globally observed teleseismic waveforms, suggesting that the earthquake was a normal‐faulting event on a steeply dipping plane, with the major slip concentrated around a relatively shallow depth of 28 km. The modeled rupture evolution showed unilateral, downdip propagation northwestward from the hypocenter, and the downdip width of the main rupture was restricted to less than 30 km below the slab interface, suggesting that the downdip extensional stresses due to the slab bending were the primary cause of the earthquake. The rupture front abruptly decelerated at the northwestern end of the main rupture where it intersected the subducting Tehuantepec Fracture Zone, suggesting that the fracture zone may have inhibited further rupture propagation. A source model of the 2017 Chiapas earthquake was constructed by kinematic waveform inversion using teleseismic PwaveformsThe model favors a shallow rupture area in the upper part of the slab centered at 28 km depthThe rupture process was driven by downdip extensional stresses due to slab bending and terminated at a subducting fracture zone

The Upper Geyser Basin in Yellowstone National Park contains one of the highest concentrations of hydrothermal features on Earth including the iconic Old Faithful geyser. Although this system has been the focus of many geological, geochemical, and geophysical studies for decades, the shallow (<200 m) subsurface structure remains poorly characterized. To investigate the detailed subsurface geologic structure including the hydrothermal plumbing of the Upper Geyser Basin, we deployed an array of densely spaced three‐component nodal seismographs in November of 2015. In this study, we extract Rayleigh wave seismic signals between 1 and 10 Hz utilizing nondiffusive seismic waves excited by nearby active hydrothermal features with the following results: (1) imaging the shallow subsurface structure by utilizing stationary hydrothermal activity as a seismic source, (2) characterizing how local geologic conditions control the formation and location of the Old Faithful hydrothermal system, and (3) resolving a relatively shallow (10–60 m) and large reservoir located ~100 m southwest of Old Faithful geyser. We seismically image Yellowstone's iconic hydrothermal system utilizing stationary hydrothermal activity as a seismic sourceWe constrain how shallow geologic structure and fluid pathways control the formation and location of Old Faithful geyser and the Upper Geyser Basin hydrothermal systemWe image a relatively large and shallow, 10‐60 m, 200 m wide hydrothermal reservoir centered ~100 m southwest of Old Faithful geyser

Tornadoes generate seismic signals when they contact the ground. Here we examine the signals excited by the Joplin tornado, which passed within 2 km of a station in the Earthscope Transportable Array. We model the tornado‐generated vertical seismic signal at low frequencies (0.01–0.03 Hz) and solve for the strength of the seismic source. The resulting source amplitude is largest when the tornado was reported to be strongest (EF 4–5), and the amplitude is smallest when the tornado was weak (EF 0–2). A further understanding of the relationship between source amplitude and tornado intensity could open up new ways to study tornadoes from the ground. When a tornado touches down, its winds and air pressure shake the ground, creating small, but measurable, seismic waves. We look at these waves created by the Joplin tornado in Missouri (2011), which happened to pass close to a seismometer. We modeled these waves to find out how large of a force the tornado is exerting on the ground. We found that the force is largest during the time when the tornado was reported to be strongest (EF 4–5 on the Enhanced Fujita Scale) and smallest when the tornado was weak (EF 0–2). Further understanding this relationship could open a new way to study the strength of tornadoes by seismic waves. Joplin tornado generated seismic waves when in contact with the groundColocated barometers are required to remove air waves in the vertical component seismogramModeling seismic signals shows correlation between source amplitudes and tornado intensity on EF scale