Search

We present probabilistic centroid-moment tensor solutions inferred from the combination of Hamiltonian Monte Carlo sampling and a 3-D full-waveform inversion Earth model of the Japanese islands. While the former provides complete posterior probability densities, the latter allows us to exploit waveform data with periods as low as 15 s. For the computation of Green's functions, we employ spectral-element simulations through the radially anisotropic and visco-elastic model, leading to substantial improvements of data fit compared to layered models. Focusing on 13 Mw 4.8–Mw 5.3 offshore earthquakes with a significant non-double-couple (non-DC) component, we simultaneously infer the centroid location, time and moment tensor without any a priori constraints on the faulting mechanism. Furthermore, we perform the inversions across several period bands, varying the minimum period between 15 and 50 s. Accounting for 3-D Earth structure at shorter periods can increase the double-couple (DC) component of an event, compared to the GCMT solution, by tens of percent. This suggests that non-DC events in the GCMT catalog may result from unmodeled Earth structure and the related limitation to longer-period data. We also observe that significant changes in source parameters, and the DC component in particular, may be related to only small waveform changes, thereby accentuating the importance of a reliable Earth model. Posterior probability density distributions become increasingly multimodal for shorter-period data that provide tighter constraints on source parameters. This implies, in our specific case, that stochastic approaches to the source inversion problem are required for periods below ∼20 s to avoid trapping in local minima., Journal of Geophysical Research: Solid Earth, 128 (1), ISSN:2169-9313, ISSN:0148-0227, ISSN:2169-9356

SUMMARY A dynamic finite-fault source inversion for stress and frictional parameters of the Mw 6.3 2017 Lesvos earthquake is carried out. The main shock occurred on June 12, offshore the southeastern coast of the Greek island of Lesvos in the north Aegean Sea. It caused 1 fatality, 15 injuries, and extensive damage to the southern part of the island. Dynamic rupture evolution is modelled on an elliptic patch, using the linear slip-weakening friction law. The inversion is posed as a Bayesian problem and the Parallel Tempering Markov Chain Monte Carlo algorithm is used to obtain posterior probability distributions by updating the prior distribution with progressively more constraints. To calculate the first posterior distribution, only the constraint that the model should expand beyond the nucleation patch is used. Then, we add the constraint that the model should reach a moment magnitude similar to that obtained from our centroid moment tensor inversion. For the final posterior distribution, 15 acceleration records from Greek and Turkish strong motion networks at near regional distances (≈ 30–150 km) in the frequency range of 0.05–0.15 Hz are used. The three posterior distributions are compared to understand how much each constraint contributes to resolving different quantities. The most probable values and uncertainties of individual parameters are also calculated, along with their mutual trade-offs. The features best determined by seismograms in the final posterior distribution include the position of the nucleation region, the mean direction of rupture (towards WNW), the mean rupture speed (with 68 per cent of the distribution lying between 1.4 and 2.6 km s–1), radiated energy (12–65 TJ), radiation efficiency (0.09–0.38) and the mean stress drop (2.2–6.5 MPa).

International audience; Moderate-to-large earthquakes in rifts may occur on leading boundary faults or inner antithetic faults. Here we show a rare case of the 2020–2021 seismic sequence in the Corinth rift, that culminated in the shallow rupture of the antithetic fault, neither preceded nor followed by the leading fault rupture. The hypocenter of the largest shock (Mw 5.3 of 17 February 2021) was located at ∼8 km depth. However, seismic waveform data, supported by satellite-geodetic and tide gauge measurements, pointed to rupture at shallow depth (∼3 km), where no earthquakes were previously observed. We show that the earthquake most probably ruptured two orthogonal, conjugate fault segments: a weak nucleation phase occurred in the microseismically highly active sub-horizontal detachment layer, followed – a few seconds later – by a larger, shallow moment release on a high-angle, south-dipping normal fault. The latter is the Mornos offshore fault, antithetic to the leading, north-dipping Psathopyrgos fault. Our study presents the first instrumental/observational evidence of a very shallow Mw 5+ event in this rift – and one of the few reported worldwide. The depth limit of the main shallow slip patch coincides with the expected crossing of the Mornos fault with the Psathopyrgos fault, stressing the importance of fault segmentation and rooting inherited from the rift history. This unusual shallow slip in a depth range with little background seismicity and few aftershocks needs to be further investigated by dynamic modeling as a possible prototype of hazardous events in rift environments.

The finite difference contrast source inversion (FDCSI) method is a nonlinear inversion method under the framework of inverse scattering theory. In this work, we use the FDCSI method for the elasti...

One of the key goals of microseismic processing is accurate estimation of the source location. Using full-waveform information in passive-source data sets can potentially delineate microseismic sources. The accuracy of the compressional-wave and shear-wave velocities has a strong influence on the estimation of source locations and hence the reliability of the fracture detection. We have adopted a methodology for passive source and velocity inversion, in which the conventional source term of the elastic wave equation is represented by an equivalent source. The equivalent source term is composed of source images and source functions because it is inspired by elastic reflection waveform inversion. Thus, we update the source locations, source functions, and velocities simultaneously by using a waveform inversion scheme. In the 2D isotropic case, the source terms are defined by two source image components and three source function components. They provide an alternative representation of the source mechanism, usually defined by the moment tensor. Waveform inversion of passive events has severe nonlinearity due to the unknown source locations in space and their functions in time. We have thus used a source-independent objective function, based on convolving reference traces with modeled and observed data, to avoid cycle skipping caused by the unknown sources. We first synthetically examined our method on a modified Marmousi model. Then, by applying a nested inversion for these variables, our method also produces good estimation of the source and background velocity for real microseismic monitoring data. We use a ball-drop event to test the accuracy because the inverted source location should match the ball-seat location. For the uncontrolled events, the estimated source distribution using waveform inversion agrees with the local stress potential information. Although our method has a higher computational cost than traveltime- or migration-based methods, the estimated event locations have significantly improved accuracy.

An Mw 7.5 earthquake struck Palu in the northern coast of Sulawesi island, Indonesia, on 28 September 2018. Its focal mechanism was determined to be a left‐lateral strike‐slip fault, which is generally expected to not produce a tsunami. However, a large tsunami with runup heights of more than 6 m was observed along the coast of Palu city. Here, we show a complex triggering supershear source model as determined by teleseismic waveform inversion. Three asperities with different slip characteristics were found on the 120‐kilometer‐long rupture zone. Significant triggering rupture with a supershear speed was observed south of the epicenter, which was just beneath Palu city. This special rupture process can cause a strong directivity effect that produced anomalously large ground shaking with nonlinear effects in Palu area. The coseismic deformation determined from the inverted source model showed large horizontal displacements. These horizontal movements combined with complex bathymetry and topography could have pushed seawater to generate a tsunami even though the Palu earthquake was a strike‐slip event.

International audience; Since February 2019, NASA's InSight lander is recording seismic signals on the planet Mars, which, for the first time, allows to observe ongoing tectonic processes with geophysical methods. A number of Marsquakes have been located in the Cerberus Fossae graben system in Elysium Planitia and further west, in the Orcus Patera depression. We present a first study of the focal mechanisms of three well-recorded events (S0173a, S0183a, S0235b) to determine the processes dominating in the source region. We infer for all three events a predominantly extensional setting. Our method is adapted to the case of a single, multicomponent receiver and based on fitting waveforms of P and S waves against synthetic seismograms computed for the initial crustal velocity model derived by the InSight team. We explore the uncertainty due to the single-station limitation and find that even data recorded by one station constrains the mechanisms (reasonably) well. For the events in the Cerberus Fossae region (S0173a, S0235b) normal faulting with a relatively steep dipping fault plane is inferred, suggesting an extensional regime mainly oriented E-W to NE-SW. The fault regime in the Orcus Patera region is not determined uniquely because only the P wave can be used for the source inversion. However, we find that the P and weak S waves of the S0183a event show similar polarities to the event S0173, which indicates similar fault regimes.

Beamforming and backprojection methods offer a data-driven approach to image noise sources, but provide no opportunity to account for prior information or iterate through an inversion framework. In contrast, recent methods have been developed to locate ambient noise sources based on cross-correlations between stations and the construction of finite-frequency kernels, allowing for inversions over multiple iterations. These kernel-based approaches show great promise, both in mathematical rigour and in results, but are less physically intuitive and interpretable. Here we show that these apparently two different classes of methods, beamforming and kernel-based inversion, are achieving exactly the same result in certain circumstances. This paper begins with a description of a relatively simple beamforming or backprojection algorithm, and walks through a series of modifications or enhancements. By including a rigorously defined physical model for the distribution of noise sources and therefore synthetic correlation functions, we come to a framework resembling the kernel-based iterative approaches. Given the equivalence of these approaches, both communities can benefit from bridging the gap. For example, inversion frameworks can benefit from the numerous image enhancement tools developed by the beamforming community. Additionally, full-waveform inversion schemes that require a window selection for the comparisons of misfits can more effectively target particular sources through a windowing in a beamform slowness domain, or might directly use beamform heatmaps for the calculation of misfits. We discuss a number of such possibilities for the enhancement of both classes of methods, testing with synthetic models where possible., Geophysical Journal International, 224 (3), ISSN:0956-540X, ISSN:1365-246X

Earthquake source parameters such as slip distribution and slip rate are useful information for understanding the physical processes behind earthquakes. The isochrone back-projection method, or isochrone-BPM, is one of the methods used to obtain slip distribution in the fault plane by incorporating the isochrone concept with back-projection method. Isochrone-BPM has advantages in its ease of implementation. However, it appears that the resulted slip distribution images always contain artifacts causing the images to be smeared. The emergence of these artifacts may lead to misinterpretation of the slip distribution, thus becoming a major weakness of isochrone-BPM. In this work, an alternative approach is proposed by utilizing least-square inversion scheme with the addition of damping factor as an alternative to the isochrone-BPM, which is then referred to as damped isochrone inversion since it still utilizes the core formulation of isochrone-BPM. The application of isochrone-BPM slip inversion to synthetic data generated from a test model shows that the quality of the resulted image will highly depend on the set of receiving stations used. In addition, the isochrone-BPM image will also depend on the focal mechanism of the earthquake, which is indicated by the difference in patterns on different mechanisms. Contrary to isochrone-BPM, the damped isochrone inversion produces slip distribution images that do not depend on the set of receiving stations used and do exhibit dependence on the focal mechanism. These results may suggest that as an alternative to the isochrone-BPM, the damped isochrone inversion offers better performance in recovering the slip distribution images.

We have previously developed a tsunami source inversion method based on “Time Reverse Imaging” and demonstrated that it is computationally very efficient and has the ability to reproduce the tsunami source model with good accuracy using tsunami data of the 2011 Tohoku earthquake tsunami. In this paper, we implemented this approach in the 2009 Samoa earthquake tsunami triggered by a doublet earthquake consisting of both normal and thrust faulting. Our result showed that the method is quite capable of recovering the source model associated with normal and thrust faulting. We found that the inversion result is highly sensitive to some stations that must be removed from the inversion. We applied an adjoint sensitivity method to find the optimal set of stations in order to estimate a realistic source model. We found that the inversion result is improved significantly once the optimal set of stations is used. In addition, from the reconstructed source model we estimated the slip distribution of the fault from which we successfully determined the dipping orientation of the fault plane for the normal fault earthquake. Our result suggests that the fault plane dip toward the northeast.

In recent studies, the Time Reverse Imaging (TRI) method has been implemented in tsunami science to estimate tsunami source models. A TRI algorithm called Green's Function Based Time Reverse Imaging (GFTRI) was previously developed to reconstruct the source model by using source inversion as a guide to appropriately scale time-reversed images of the tsunami source. In this article, we consider a more complete approach to source inversion using reversed time images of the tsunami source by considering cross-correlations among Green's functions. As a consequence, the performance of the method improves significantly. We tested this new algorithm using data from the 2011 Japan tsunami. Our results show that the method is capable of extracting more details of the source and providing excellent waveform fits at all stations, including those not used in the source imaging. We have also studied the sensitivity analysis of reversed-time tsunami source inversion and found that the method is less sensitive to the number of stations, once a minimum number of stations is utilized. Moreover, this new approach is able to estimate the tsunami source with reasonable accuracy using data available soon after an earthquake, which indicates that it has potential to be used in both near- and far-field tsunami forecasting.

We propose a test case for the Source Inversion Validation (SIV) community and other interested seismologists using high‐quality waveforms of the M L 5.5 Orkney, South Africa, earthquake. This rupture initiated ∼5 km below the surface and terminated several hundred meters below the deepest levels of a gold mine. The Orkney earthquake and its aftershock sequence were well recorded by a dense network of surface and in‐mine seismometers. In situ stress measurements and fault core through the rupture surface are anticipated to be available as part of the recently funded International Continental Scientific Drilling Program project to drill from within the mine into the Orkney earthquake rupture zone through a project titled “Drilling into Seismogenic Zones of M2.0‐M5.5 Earthquakes in South African gold mines.” The remarkable quality and range of geophysical data available for the M L 5.5 Orkney earthquake make this event an ideal test case for the SIV community using actual seismic data to determine the spatial and temporal evolution of earthquake rupture.

Seismic and tsunami waves were the most destructive phenomena during the 2011 Tohoku earthquake. Here we ask whether the same source model can be used to obtain equally good simulations of the observed strong ground motions and the tsunami waves. If so, we could use the same source modeling technique for predictions of both tsunami waves and strong ground motions. Long-period (20–200 s) strong ground motion data were inverted to determine the rupture process of the 2011 Tohoku earthquake. The inverted slip distribution shows that large-slip area was extended around 150 km to north and 120 km to south from hypocenter and stretched between hypocenter and the trench. We estimated two slip models: without and with additional constraints. The first relatively rough slip model reproduced short-period tsunami wave (4–5 min). Due to better constraint by data of other kind and being relatively smooth the second model reproduces long-period tsunami wave (15–25 min), as well as keeps good fit of the short-period tsunami, although short-period tsunami becomes underestimated. We used a fully nonlinear Boussinesq water wave model to model tsunami propagation, and used the results to tune the slip model. Finally, we developed a tuned slip model that combines the features of both the first and the second slip models and reproduced both the short- and long-period tsunami waves, while maintaining good fit of the seismic waveforms.

Recent studies in dynamic source modeling and kinematic source inversion show that earthquake rupture may contain greater complexity than we previously anticipated, including multiple slipping at a given point on a fault. Finite source inversion methods suffer from the nonuniqueness of solutions, and it may become more serious if we aim to resolve more complex rupture models. In this study, we perform synthetic inversion tests with dynamically generated complex rupture models, including both supershear rupture and slip reactivation, to understand the possibility of resolving complex rupture processes by inverting seismic waveform data. We adopt a linear source inversion method with multiple windows, allowing for slipping from the nucleation of rupture to the termination at all locations along a fault. We regularize the model space effectively in the Bayesian framework and perform multiple inversion tests by considering the effect of inaccurate Green’s functions and station distributions. We also perform a spectral stability analysis. Our results show that it may be possible to resolve both a supershear rupture front and reactivated secondary slipping using the linear inversion method if those complex features are well separated from the main rupture and produce a fair amount of seismic energy. It may be desirable to assume the full complexity of an earthquake rupture when we first develop finite source models after a major event occurs and then assume a simple rupture model for stability if the estimated models do not show a clear pattern of complex rupture processes.

On 6 February 2016, an M L 6.6 earthquake occurred in the Meinong area of southern Taiwan, causing anomalously large ground shaking. In this study, a joint source inversion was performed to understand the rupture process of this event and to identify how the anomalous strong motion was produced. The results show that the rupture process of this event was complex; at least two asperities were identified on the fault plane. The rupture mainly developed in the down‐dip direction and propagated toward the northwest. Results from a 3D wave propagation simulation further indicate that the strong ground shaking observed in southwest Taiwan was caused by a combination of three effects: (1) rupture directivity, (2) source radiation pattern, and (3) sedimentary amplification. The results demonstrate that constructive source rupture and wave propagation effects due to a moderate earthquake can cause anomalously strong ground motion and serious seismic hazards in nearby densely populated areas.

We reviewed joint inversion studies of the rupture processes of significant earthquakes, using the definition of a joint inversion in earthquake source imaging as a source inversion of multiple kinds of datasets (waveform, geodetic, or tsunami). Yoshida and Koketsu (Geophys J Int 103:355–362, 1990), and Wald and Heaton (Bull Seismol Soc Am 84:668–691, 1994) independently initiated joint inversion methods, finding that joint inversion provides more reliable rupture process models than single-dataset inversion, leading to an increase of joint inversion studies. A list of these studies was made using the finite-source rupture model database (Mai and Thingbaijam in Seismol Res Lett 85:1348–1357, 2014). Outstanding issues regarding joint inversion were also discussed.

In the estimation of spatiotemporal slip models, kinematic source inversions using Akaike's Bayesian Information Criterion (ABIC) and the multiple-time-window method have often been used. However, there are cases in which conventional ABIC-based source inversions do not work well in the determination of hyperparameters when a non-negative slip constraint is used. In order to overcome this problem, a new source inversion method was developed in this study. The new method introduces a fully Bayesian method into the kinematic multiple-time-window source inversion. The multiple-time-window method is one common way of parametrizing a source time function and is highly flexible in terms of the shape of the source time function. The probability distributions of model parameters and hyperparameters can be directly obtained by using the Markov chain Monte Carlo method. These probability distributions are useful for simply evaluating the uniqueness and reliability of the derived model, which is another advantage of a fully Bayesian method. This newly developed source inversion method was applied to the 2011 Ibaraki-oki, Japan, earthquake (Mw 7.9) to demonstrate its usefulness. It was demonstrated that the problem with using the conventional ABIC-based source inversion method for hyperparameter determination appeared in the spatiotemporal source inversion of this event and that the newly developed source inversion could overcome this problem.

Waveform backprojection is a key technique of earthquake-source imaging, which has been widely used for extracting information of earthquake source evolution that cannot be obtained by kinematic source inversion. The technique enjoys considerable popularity, owing to the simplicity of its implementation and the robustness of its processing, but the physical meaning of backprojection images has remained elusive. In this study, we reviewed the mathematical representation of backprojection (BP) and hybrid backprojection (HBP) methods, following the pioneering work of Fukahata et al. (Geophys. J. Int. (2014) 196, 552–559), to clarify the physical implications of BP images. We found that signal intensity in BP and HBP images is scaled with the amplitude of the Greens function that corresponds to a unit-step slip, which results in the signal intensity being depth dependent. We propose variants of BP and HBP, which we call kinematic BP and HBP, respectively, to relate the BP signal intensity to slip motion of an earthquake by modifying the normalizing factors used in the original BP and HBP methods. The original BP and HBP images remain useful for assessing the spatiotemporal strength of the wave radiation, which scales with the amplitude of the Greens function, whereas the kinematic BP and HBP methods are suitable for imaging the slip motion that is responsible for the high-frequency radiation produced during the source-rupture process.

International audience; The ill-posed nature of earthquake source estimation derives from several factors including the quality and quantity of available observations and the fidelity of our forward theory. Observational errors are usually accounted for in the inversion process. Epistemic errors, which stem from our simplified description of the forward problem, are rarely dealt with despite their potential to bias the estimate of a source model. In this study, we explore the impact of uncertainties related to the choice of a fault geometry in source inversion problems. The geometry of a fault structure is generally reduced to a set of parameters, such as position, strike and dip, for one or a few planar fault segments. While some of these parameters can be solved for, more often they are fixed to an uncertain value. We propose a practical framework to address this limitation by following a previously implemented method exploring the impact of uncertainties on the elastic properties of our models. We develop a sensitivity analysis to small perturbations of fault dip and position. The uncertainties of our fixed fault geometry are included in the inverse problem under the formulation of the misfit covariance matrix that combines both prediction and observation uncertainties. We validate this approach with the simplified case of a fault that extends infinitely along strike, using both Bayesian and optimization formulations of a static slip inversion. If epistemic errors are ignored, predictions are overconfident in the data and slip parameters are not reliably estimated. In contrast, inclusion of uncertainties in fault geometry allows us to infer a robust posterior slip model. Epistemic uncertainties can be many orders of magnitude larger than observational errors for great earthquakes (Mw > 8). Not accounting for uncertainties in fault geometry may partly explain observed shallow slip deficits for continental earthquakes. Similarly, ignoring the impact of epistemic errors can also bias estimates of near-surface slip and predictions of tsunamis induced by megathrust earthquakes.

International audience; Large earthquakes in stable continental regions remain puzzling as; unlike at plate boundaries, they do not result from the local buildup of strain driven by plate tectonics. The 2017 Mw6.5, Bostwana normal faulting earthquake occurred in a region devoid from recent tectonic activity and where present-day deformation is negligible. The depth of the event (29 ± 4 km), in a felsic lower crust where ductile deformation is expected, likely requires a transient pulse of fluids from a deep source to activate brittle faulting. The mainshock was preceded by two foreshock swarm-like sequences that may be further evidence for fluid movement in a critically loaded fault network. Contrary to plate boundary events, the Mw6.5 Botswana earthquake did not require prior localized stress or strain accumulation. We propose that the crust in stable continental regions, even long after the last tectonic episode, constitutes a reservoir of elastic stress that can be released episodically, for instance, as a result of deep fluid migration.

This paper considers the importance of model parameterization, including dispersion, source kinematics, and source discretization, in tsunami source inversion. We implement single and multiple time window methods for dispersive and nondispersive wave propagation to estimate source models for the tsunami generated by the 2011 Tohoku-Oki earthquake. Our source model is described by sea surface displacement instead of fault slip, since sea surface displacement accounts for various tsunami generation mechanisms in addition to fault slip. The results show that tsunami source models can strongly depend on such model choices, particularly when high-quality, open-ocean tsunami waveform data are available. We carry out several synthetic inversion tests to validate the method and assess the impact of parameterization including dispersion and variable rupture velocity in data predictions on the inversion results. Although each of these effects has been considered separately in previous studies, we show that it is important to consider them together in order to obtain more meaningful inversion results. Our results suggest that the discretization of the source, the use of dispersive waves, and accounting for source kinematics are all important factors in tsunami source inversion of large events such as the Tohoku-Oki earthquake, particularly when an extensive set of high-quality tsunami waveform recordings are available. For the Tohoku event, a dispersive model with variable rupture velocity results in a profound improvement in waveform fits that justify the higher source complexity and provide a more realistic source model.

On 24 November 2015 two events of magnitude Mw 7.5 and Mw 7.6 occurred at 600 km depth under the Peru–Brazil boundary. These two events were separated in time by 300 s. Deep event doublets occur often under South America. The characteristics that control these events and the dynamic interaction between them are an unresolved problem. We used teleseismic and regional data, situated above the doublet, to perform source inversion in order to characterize their ruptures. The overall resemblance between these two events suggests that they share similar rupture process. They are not identical but occur on the same fault surface dipping westward. Using a P-wave stripping and stretching method we determine rupture speed of 2.25 km/s. From regional body wave inversion we find that stress drop is similar for both events, they differ by a factor of two. The similarity in geometry, rupture velocity, stress drop and radiated energy, suggests that these two events looked like simple elliptical ruptures that propagated like classical sub-shear brittle cracks.

A tsunamigenic earthquake with a moment magnitude of 7.3 occurred near the Japan Trench, off Tohoku, northeast Japan, on 7 December 2012. Operational seismic monitoring inferred that the earthquake was composed of doublet sources of comparable magnitudes: the first event was reverse faulting and the second event, which occurred 10–20 s later, was normal faulting. An associated tsunami was observed at offshore stations, having an amplitude of 10−1 to 101 cm. Inverse modeling using the observed tsunami records reveals possible vertical seafloor deformations. We can then constrain, in terms of the source location, that the tsunamigenic earthquake involves two subfaults across the Japan Trench: one is the outer trench reverse faulting as the first event and the second is the inner trench normal faulting. The present study shows that tsunami analysis with seismograph analyses effectively constrains the epicenter locations, even in the case of complex earthquakes like doublets.

International audience; The kinematic rupture process of the 2011 Mw 9.0 Tohoku earthquake is inverted with an elliptical-patches method, using a genetic algorithm, for the purpose of rapid and robust estimation of the source parameters of a mega-earthquake. We use the ground-displacement field provided by a continuous GPS network and the ground-velocity field recorded by acceleration networks. In addition to the typical inversion procedure in which a data duration long enough to cover the whole rupture process is used, inversions based on shorter signals, giving an incomplete view over the ground shaking sequence, are also carried out. How fast can a robust estimation of the source parameters be obtained? Using the elliptical approximation, we find that robust solutions of Mw ~ 9.0-earthquakes are rather quickly obtained regardless of the frequency band and the elliptical patch description. It is also confirmed that, because of the absence of off-shore recording stations on the east side of the fault, some uncertainties in the rupture process cannot be completely removed. In fact, at the very low frequencies considered, the western part of the fault (deeper portion) contributes more significantly to the recorded signals than does the other end close to the trench. This problem also prevents refinement of the description of the rupture process, in particular when using more than one ellipse.

On 24 May 2013, the largest ever-recorded deep earthquake occurred beneath Sea of Okhotsk. A multiple point source inversion procedure is applied to constrain source process of this earthquake, based on waveform modeling of both direct P and SH waves and near-surface reflected pP and sSH waves. Our results indicate that the earthquake consists of six major subevents separated in space and time, encompassing a horizontal dimension of 64 ± 4 km along ~ N160°E and a downward depth extension of 35 ± 4 km. The geographic distribution and focal mechanisms of the inferred subevents and foreshock/aftershock locations do not fit into plane rupture. We suggest that the earthquake can be best explained by a cascading failure of shear instability within preexisting weak zones in the region, with the perturbation of stress generated by a shear instability triggering another.

Earthquake source characterization has been significantly speeded up in the last decade with the development of rapid inversion techniques in seismology. Among these techniques, the W-phase source inversion method quickly provides point source parameters of large earthquakes using very long period seismic waves recorded at teleseismic distances. Although the W-phase method was initially developed to work at global scale (within 20 to 30 min after the origin time), faster results can be obtained when seismological data are available at regional distances (i.e., Δ ≤ 12°). In this study, we assess the use and reliability of regional W-phase source estimates in China and neighboring areas. Our implementation uses broadband records from the Chinese network supplemented by global seismological stations installed in the region. Using this data set and minor modifications to the W-phase algorithm, we show that reliable solutions can be retrieved automatically within 4 to 7 min after the earthquake origin time. Moreover, the method yields stable results down to Mw = 5.0 events, which is well below the size of earthquakes that are rapidly characterized using W-phase inversions at teleseismic distances.

Finite-fault earthquake source inversions infer the (time-dependent) displacement on the rupture surface from geophysical data. The resulting earthquake source models document the complexity of the rupture process. However, multiple source models for the same earthquake, obtained by different research teams, often exhibit remarkable dissimilarities. To address the uncertainties in earthquake-source inversion methods and to understand strengths and weaknesses of the various approaches used, the Source Inversion Validation (SIV) project conducts a set of forward-modeling exercises and inversion benchmarks. In this article, we describe the SIV strategy, the initial benchmarks, and current SIV results. Furthermore, we apply statistical tools for quantitative waveform comparison and for investigating source-model (dis)similarities that enable us to rank the solutions, and to identify particularly promising source inversion approaches. All SIV exercises (with related data and descriptions) and statistical comparison tools are available via an online collaboration platform, and we encourage source modelers to use the SIV benchmarks for developing and testing new methods. We envision that the SIV efforts will lead to new developments for tackling the earthquake-source imaging problem.

The force applied to the Earth by the calving of two icebergs at Jakobshavn Isbrae, Greenland, has been quantified. The source force history was recovered by inversion of regional broadband seismograms without any a priori constraint on the source time function, in contrast with previous studies. For periods 10-100 s, the three-component force can be obtained from distant stations alone and is proportional to the closest station seismograms. This inversion makes it possible to quantify changes of the source force direction and amplitude as a function of time and frequency. A detailed comparison with a video of the event was used to identify four forces associated with collision, then bottom-out and top-out rotation of the first and second icebergs, and ice mélange motion. Only the two iceberg rotations were identified in previous studies. All four processes are found here to contribute to the force amplitude and variability. Such a complete time-frequency force history provides unique dynamical constraints for mechanical calving models. ERC. Grant Number: ERC-CG-2013-PE10-617472 ANR. Grant Number: ANR-11-BS01-0016 LANDQUAKES, CNCS‐UEFISCDI. Grant Number: PN-II-ID-PCE-2011-3-0045

We present a quantitative comparison between seismic moment tensor (MT) inversion solutions using 1D and 3D synthetic seismograms at two frequency bands for events in the Middle East, to assess the effects of 3D models on source studies. Complex geology associated with the active continental convergent margin leads to a scarcity of reliable, available data, necessitating a thorough examination of solution stability and robustness to assure an accurate description of sources with well‐characterized source parameters. Solutions were calculated for 195 events ( M w>5.5) using a full‐waveform MT inversion matching both phase and amplitude. Seismic data processed at two frequency bands compares short‐ and long‐period performance for 1D and 3D synthetic seismograms. An improvement in fit between data and synthetics is seen using 3D over 1D synthetic seismograms, especially for complex body‐wave propagation and surface‐wave dispersion. At short periods, 3D synthetics provide a more robust solution compared with 1D, showing a reduction in error of the source mechanism. The percentage of double‐couple components increases with the addition of 3D structure and suggests the percentage of non‐double‐couple components is a result of poorly constrained Earth structure. Event solutions contained in the catalog (![Graphic][1] Table S1 in the electronic supplement to this paper) have an average cross‐correlation value of 0.87, with good amplitude ratios, and are improved (i.e., increased variance reduction) yet consistent with longer period solutions from the Global Centroid Moment Tensor (CMT) catalog. Online Material: Table of 184 event solutions from 3D moment tensor inversion. [1]: /embed/inline-graphic-1.gif

The 2016 Kaikōura (New Zealand) earthquake generated large ground motions and resulted in multiple onshore and offshore fault ruptures, a profusion of triggered landslides, and a regional tsunami. Here we examine the rupture evolution using two kinematic modelling techniques based on analysis of local strong-motion and high-rate GPS data. Our kinematic models capture a complex pattern of slowly (Vr < 2km/s) propagating rupture from the south to north, with over half of the moment release occurring in the northern source region, mostly on the Kekerengu fault, 60 seconds after the origin time. Both models indicate rupture re-activation on the Kekerengu fault with the time separation of ~11 seconds between the start of the original failure and start of the subsequent one. We further conclude that most near-source waveforms can be explained by slip on the crustal faults, with little (

Dynamic rupture simulations have been performed in order to generate a physically consistent slip distribution. In these simulations, an attempt has been made to investigate the governing parameters, called dynamic parameters, of historical earthquakes. However, it has been reported that the parameters are not estimated accurately when only the ground motion records are employed; this is because of the trade-offs existing between the parameters, for example, between the strength excess and the slip-weakening distance. In this article, we apply a dynamic source inversion to 2D synthetic tests in order to discuss parameter trade-offs. Sensitivity tests, which consist of numerous sets of inversions including a fixed stage and a released stage, are conducted such that a particular set of estimation parameters is not estimated in the fixed stage; then all of the parameters are estimated in the released stage. The results imply trade-offs between the rupture time and other parameters. From detailed comparisons between slip rates and waveforms, it is observed that the shapes of the slip-rate profiles between the rupture time and the peak time do not contribute well to the generated waveforms. From a simple 1D assumption, a gradient of the slip-weakening friction law, namely, a weakening gradient, is analytically related to the peak-to-rupture time with a -1 slope in the log–log plot. The values estimated from the synthetic tests almost mirror the values of the -1 slope. The results imply the trade-offs between the weakening gradient and the peak-to-rupture time.

Several source parameters (depth, scalar moment, rupture velocity, focal mechanism) are determined for the moderate-size October 27, 2004 (MW = 5.8) Vrancea (Romania) intermediate-depth earthquake by inversion of teleseismic waveforms.

Geophysical Journal International, 200 (1), ISSN:0956-540X, ISSN:1365-246X

SUMMARY A dynamic rupture model is used to simulate source rupture processes based on the physical relationship between the stress and the strain field near a fault plane. When, however, a dynamic rupture model is used for inversion analysis, much computation time is required. We propose a new dynamic source inversion method, formulated by introducing a suitable set of dynamic parameters, ‘consolidated dynamic parameters’, which consists of rupture time, slip-weakening distance and dynamic stress drop, as estimating parameters. The rupture time is adopted for estimating parameters to derive stable formulation and to decrease CPU time. Compared with the conventional method, calculation time is much lessened. A multiscale approach to solve non-linear inversion problem is that the number of estimated values and the observation frequency ranges increase as the scale changes. Numerical results show the good reconstructing of target dynamic parameter distributions by the proposed inversion method.

In this work, we validated the automatic inversion of the U.S. Geological Survey (USGS) intensity data set of the Whittier Narrows M L 5.9, 1987 earthquake. This validation of our new technique was obtained by comparing the source inversion results with the principal source parameters coming from instrumental measurements independent of our study. To perform the inversion, first, we used a genetic algorithm (GA) with a population of 20,000 individuals (i.e., sources). Second, because the problem is bimodal, we also used a niching genetic algorithm, (NGA) with four demes of 2,000 individuals. This gave us almost the same results. The site intensities were calculated by our kinematic KF function. Twelve source parameters were involved in the inversions, the most sensitive of which are the epicentral coordinates and the fault-plane solution. Two minimum variance models were determined by both the GA and the NGA inversions: (1) one east-west trending dip-slip source, which is in agreement with that already known from instrumental measurements, and (2) one almost coinciding with its auxiliary plane in the fault-plane solution. These findings almost coincide with those produced by the grid-search inversion method, but the GA-NGA inversion is much faster and does not need strong constraints. This confirms that it is possible to get an approximate idea of the source of the studied earthquake also by automatically inverting the regional pattern of the USGS intensities. This result encourages us to validate our inversion technique with more well documented earthquakes and to treat intensities of preinstrumental earthquakes, which are the principal target of our work. Manuscript received 30 July 2003.

We compare two geophysical survey measurements of the same type made at different times in order to characterize the change in the geological medium during the elapsed time. The aim of this study is to develop a strategy using a full non-linear inversion algorithm as the interpretation tool. In this way, not only the location and the form of the changes are recovered, but also the changes in the material parameters of the geological medium can be estimated. In order to solve this fully non-linear problem, the so-called 'multiplicative regularized contrast source inversion' (MR-CSI) method is employed. The unique property of this iterative method is that it does not solve the forward problem at each iterative step. This makes it possible to use the non-linear inversion algorithm for large-scale computation problems. The numerical results show that by taking into account the non-linear nature of the problem, interpretation of the time-lapse data can be significantly improved, compared with that obtained using linear inversion.

Summary In order to obtain knowledge concerning the physics of earthquake rupturing more precisely and to obtain reliable source process for near-source ground motion simulation, we introduced a convolution method that incorporates the effect of a moving dislocation over a rectangular area to a synthetic point source into computing the element source waves for the waveform inversion. Therefore, the rupture directivity effect over the entire fault area is considered fully in the waveform inversion. The rupture process inverted using this technique has a continuous slip distribution inside the subfaults. The moment release history obtained for each subfault directly reflects a slip time function on the subfault because the rupture propagation effect inside each subfault is separated. The validity of this technique is based on the following approximation: the Green's functions from two point sources are approximately the same after a phase correction when the distances between two point sources is much smaller than the distance between the sources and an observation station. The condition for valid usage of this technique has been found through numerical tests. We applied this technique in the waveform inversion of strong motion records for the source process of the 1995 Hyogo-ken Nanbu earthquake. The global characteristics of moment release distributions are similar between two inversion results; one inverted with point-source element waveforms and the other with element source waves considering the finite moving dislocation effects inside each subfault. However, the details are different, e.g. the locations of local maxima of the total moment release are different between these inversion results. A positive correlation between final slip and peak slip velocity is observed.