Phases In Seismograms Research Articles

PickBlue: Seismic Phase Picking for Ocean Bottom Seismometers With Deep Learning

AbstractDetecting phase arrivals and pinpointing the arrival times of seismic phases in seismograms is crucial for many seismological analysis workflows. For land station data, machine learning methods have already found widespread adoption. However, deep learning approaches are not yet commonly applied to ocean bottom data due to a lack of appropriate training data and models. Here, we compiled an extensive and labeled ocean bottom seismometer (OBS) data set from 15 deployments in different tectonic settings, comprising ∼90,000 P and ∼63,000 S manual picks from 13,190 events and 355 stations. We propose PickBlue, an adaptation of the two popular deep learning networks EQTransformer and PhaseNet. PickBlue joint processes three seismometer recordings in conjunction with a hydrophone component and is trained with the waveforms in the new database. The performance is enhanced by employing transfer learning, where initial weights are derived from models trained with land earthquake data. PickBlue significantly outperforms neural networks trained with land stations and models trained without hydrophone data. The model achieves a mean absolute deviation of 0.05 s for P‐waves and 0.12 s for S‐waves, and we apply the picker on the Hikurangi Ocean Bottom Tremor and Slow Slip OBS deployment offshore New Zealand. We integrate our data set and trained models into SeisBench to enable an easy and direct application in future deployments.

Characterizing large rockfalls using their seismic signature: A case study of Hongya rockfall

A preliminary, yet timely characterization of rockfall hazards is essential for effective emergency response and mitigation measures. Such characterization is, however, particularly challenging for events occurring on high-mountain slopes in remote locations, for which precursory information remains often undetected. Large rockfalls are known to produce seismic signatures that, if properly interpreted, provide key information on the magnitude and destructiveness of the event. To achieve this, we establish a seismic signal-based assessment scheme and demonstrate its capability by taking a large event – the 5 April 2021 Hongya rockfall (Sichuan, China) – as a case study. First, we show how a rockfall can be distinguished from an earthquake and a rockslide by analyzing its seismic signatures. Then, we demonstrate how the kurtosis-based method can be used to rapidly detect the initiation of a rockfall and determine the seismic wave velocity accordingly, as well as how the arrival-time-based location method can be used to locate the event. The rockfall volume can be estimated from the magnitude of radiated seismic energy. Furthermore, we characterize the different phases in seismograms and link them to the typical stages of rockfall, i.e., precursory small rockfalls, impact and fragmentation, sliding and entrainment, and gravity deposition stages. Finally, a discussion and suggestions are provided for further improving the robustness of the assessment scheme. Our results show that the seismic signal-based scheme presented in this study is suitable for characterizing large rockfalls and has the potential for rapid and effective emergency management.

Nature and origin of an undetected seismic phase in waveforms from Southern Tyrrhenian (Italy) intermediate-depth and deep earthquakes: first evidence for the phase-A in the subducted uppermost lithospheric mantle?

We have found a previously unreported later seismic phase in seismograms of European seismic stations from intermediate-depth and deep earthquakes of the Southern Tyrrhenian subduction zone. We observe this phase at stations from 6 to 9° from the epicentre, towards north. Only seismograms of earthquakes located in a well-defined region of the slab, in the depth range of 215–320 km, show the later x-phase. In this work, we describe the nature and possible origin of this phase, and we provide a simple 2D model to explain the observed arrival times. Our analyses reveal that the x-phase propagates downward in a high velocity layer, possibly located within the deepest part of the slab. We suggest that this layer reveals the presence of the dense hydrous magnesium silicate phase A, introduced from petrological laboratory experiments, inferred to carry water in the upper mantle and predicted to be found in cold subduction zones.

Bayesian transdimensional seismic full-waveform inversion with a dipping layer parameterization

Seismic full-waveform inversion (FWI) has become a popular tool for estimating subsurface models using the amplitude and phase of seismograms. Unlike the conventional gradient-based approach, Bayesian inference using Markov chain Monte Carlo (MCMC) sampling can remove dependence on starting models and can quantify uncertainty. We have developed a Bayesian transdimensional (trans-d) MCMC seismic FWI method for estimating dipping-layer velocity models, in which the number of layers is unknown. A time-domain staggered-grid finite-difference wave equation solver is used for forward modeling. The FWI and MCMC methods are known to be computationally expensive. Two strategies are used to get practical computational performance. A layer-stripping strategy is used to accelerate sampler convergence, and a parsimonious dipping layer parameterization is used so that the MCMC algorithm can search broadly with fewer iterations. The parameters for each layer are velocity, thickness, and lower interface dip angle. We find that this parameterization has sufficient flexibility to invert for narrow 2D velocity models using small offset data. Model stitching is then used to bring several such inversions together to create larger 2D models. In turn, these can be used as starting models for gradient-based adjoint FWI to image complicated geologic settings. Two synthetic 2D numerical examples, including the Marmousi model, are considered, using seismic data dominated by reflections. Creation of good starting models traditionally requires significant human effort. We determine how much of that effort can be substituted with computation.

Guided wave observations and evidence for the low-velocity subducting crust beneath Hokkaido, northern Japan

At the western side of the Hidaka Mountain range in Hokkaido, we identify a clear later phase in seismograms for earthquakes occurring at the uppermost part of the Pacific slab beneath the eastern Hokkaido. The later phase is observed after P-wave arrivals and has a larger amplitude than the P wave. In this study, we investigate the origin of the later phase from seismic wave observations and two-dimensional numerical modeling of wave fields and interpret it as a guided P wave propagating in the low-velocity subducting crust of the Pacific plate. In addition, the results of our numerical modeling suggest that the low-velocity subducting crust is in contact with a low-velocity material beneath the Hidaka Mountain range. Based on our interpretation for the later phase, we estimate P-wave velocity in the subducting crust beneath the eastern part of Hokkaido by using the differences in the later phase travel times and obtain velocities of 6.8 to 7.5 km/s at depths of 50 to 80 km. The obtained P-wave velocity is lower than the expected value based on fully hydrated mid-ocean ridge basalt (MORB) materials, suggesting that hydrous minerals are hosted in the subducting crust and aqueous fluids may co-exist down to depths of at least 80 km.

Array-conditioned deconvolution of multiple-component teleseismic recordings

We investigate the applicability of an array-conditioned deconvolution technique, developed for analysing borehole seismic exploration data, to teleseismic receiver functions and data pre-processing steps for scattered wavefield imaging. This multichannel deconvolution technique constructs an approximate inverse filter to the estimated source signature by solving an overdetermined set of deconvolution equations, using an array of receivers detecting a common source. We find that this technique improves the efficiency and automation of receiver function calculation and data pre-processing workflow. We apply this technique to synthetic experiments and to teleseismic data recorded in a dense array in northern Canada. Our results show that this optimal deconvolution automatically determines and subsequently attenuates the noise from data, enhancing P-to-S converted phases in seismograms with various noise levels. In this context, the array-conditioned deconvolution presents a new, effective and automatic means for processing large amounts of array data, as it does not require any ad-hoc regularization; the regularization is achieved naturally by using the noise present in the array itself.

Waveform modelling of teleseismicS, Sp, SsPmP, and shear-coupled PL waves for crust- and upper-mantle velocity structure beneath Africa

SUMMARY We describe a waveform modelling technique and demonstrate its application to determine the crust- and upper-mantle velocity structure beneath Africa. Our technique uses a parallelized reflectivity method to compute synthetic seismograms and fits the observed waveforms by a global optimization technique based on a Very Fast Simulated Annealing (VFSA). We match the S, Sp, SsPmP and shear-coupled PL phases in seismograms of deep (200–800 km), moderate-to-large magnitude (5.5–7.0) earthquakes recorded teleseismically at permanent broad-band seismic stations in Africa. Using our technique we produce P- and S-wave velocity models of crust and upper mantle beneath Africa. Additionally, our use of the shear-coupled PL phase, wherever observed, improves the constraints for lower crust- and upper-mantle velocity structure beneath the corresponding seismic stations. Our technique retains the advantages of receiver function methods, uses a different part of the seismogram, is sensitive to both P- and S-wave velocities directly, and obtains helpful constraints in model parameters in the vicinity of the Moho. The resulting range of crustal thicknesses beneath Africa (21–46 km) indicates that the crust is thicker in south Africa, thinner in east Africa and intermediate in north and west Africa. Crustal P- (4.7–8 km s−1) and S-wave velocities (2.5–4.7 km s−1) obtained in this study show that in some parts of the models, these are slower in east Africa and faster in north, west and south Africa. Anomalous crustal low-velocity zones are also observed in the models for seismic stations in the cratonic regions of north, west and south Africa. Overall, the results of our study are consistent with earlier models and regional tectonics of Africa.

Converted seismic wave investigation in the Gulf of Corinth from local earthquakes

Abstract We propose a procedure for the detection of converted phases in seismograms between the P- and S-arrivals. This procedure has been successfully applied to waveforms of different stations from the 1991 passive experiment around the Aigion area (Greece). Three major converted phases have been detected at two stations on the northern Gulf of Corinth. The first phase has been interpreted as a P-to-S transmitted wave occurring at an interface located at a depth of 3 km in the northern part of the Psaromita Peninsula. The two later phases that have a more complex polarisation might come from deeper interfaces. To cite this article: D. Latorre et al., C. R. Geoscience 336 (2004).

Mapping the crustal thickness in southwest Japan using Moho-reflected waves

In this work we have tried to detect and collect later phases associated with the Moho discontinuity and use them to study the lateral variations of the crustal thickness in southwest (SW) Japan. We first compute synthetic seismograms for local earthquakes taking into account the focal mechanism solutions estimated from first motion polarity data. Then we compare the synthetics with the observed waveforms to detect the major later phases in seismograms of shallow crustal earthquakes in SW Japan. Taking the advantage of the high quality and great quantity of Hi-net waveform data now available throughout the entire Japan Islands, we have detected 1659 Moho-reflected phases (PmP and SmS) from shallow events in SW Japan. We estimated the crustal thickness in the study area using travel time differences between these later phases and the first P and S arrivals. Our results show that the Moho is deep in Chubu district and becomes shallower along the coastlines of the Japan Sea and the Pacific Ocean. A thinner crust appears in Osaka Bay and Awajii Island where the 1995 Kobe earthquake (M7.2) occurred.

Effect of a plume on long period surface waves computed with normal modes coupling

The physical characterization and depth origin of mantle plumes are not well constrained. In order to address these issues, we look for observable effects of plumes on long period surface wave seismograms. The effect of a weak but sharp heterogeneity on long period surface waves is computed by a first order normal mode theory using a generalization of the spherical harmonic summation theorem to reduce the number of operations. It turns out that it is necessary to couple very far along a given dispersion branch ( l±40 up to ±80) to remove spurious phases in seismograms. The coupling between different overtone branches (different overtone numbers) of the same kind (spheroidal or toroidal) and of different kinds are computed as well. By taking into account a large number of overtones, we are able to compute the effect of the heterogeneity on all seismogram phases from surface waves to body waves, including P–SH and SV–SH coupling. This technique is applied to different plausible models of mantle plume: a small vertical conduit down to 660 km depth, down to the core–mantle-boundary (CMB) and with or without head. We show that for a finite size plume the radiation pattern is essentially forward, which is not the case when the heterogeneity is considered as punctual. The scattered amplitude displays large variations for the different cases according to the temperature contrast, but an effect up to 10% of the incident amplitude can be expected, and should be observable on good quality seismic data.

Detection of anomalous seismic phases by the wavelet transform

SUMMARY Strong motion data at La Union, one of three near-fault stations of the Michoacan, Mexico, earthquake in 1985, are analysed by a wavelet transform in order to identify a clear later energy arrival whose frequency content and particle motion are quite different from the main part of the seismograms. We used an orthonormal set of analysing wavelets, a discrete wavelet transform, proposed by Meyer and Yamada, with which an efficient computational procedure can be achieved utilizing the fast Fourier transform. Results of an application of the above wavelet transform to three-component velocity data of La Union are summarized as follows. A vertical seismogram does not contain any distinguishable later energy arrivals for the entire frequency range recorded. The major energy arrives in the time interval between 10 and 30 s. In contrast, an east-west component seismogram shows a peculiar later arrival more than 10 s later than the main energy arrival, and our wavelet analysis shows that this later phase is characterized by an arrival time of 38 s and a frequency range from 0.1 to 1 Hz where the amplitude of the later phase is even larger than the main part. This example clearly demonstrates the potential of wavelet transforms to identify objectively any phase in seismograms localized in both time and frequency. From the information on its particle motion, the above phase corresponds to a heterogeneous and delayed break of one stronger portion on a fault on which the rupture was very smooth or ‘crack-like’ elsewhere.

Study of seismogram envelopes based on scattering by random inhomogeneities in the lithosphere: a review

Abstract Instead of the precise examination of individual phases in seismograms, we will focus on the relation between random inhomogeneities in the lithosphere and the systematic change in seismogram envelopes. Recent developments in the study of high-frequency seismogram envelopes are briefly reviewed. The smooth amplitude decay of coda envelopes of local earthquakes first attracted seismologists' interest in scattering and was investigated based on the single-scattering assumption with a uniform distribution of point-like scatterers. Subsequently, the importance of multiple scattering was recognized and theoretical models were introduced. As a bridge between conventional seismology and the scattering approach, a hybrid model in which a large impedance contrast plane was embedded in a scattering medium was proposed for the interpretation of a reflected S phase in the coda for shallow events around a volcanic front in Japan. Next, three-component seismogram envelopes of a small local earthquake were synthesized using the Born approximation in inhomogeneous media. The S coda is excited in all directions from the hypocenter. The P coda appears even in the nodal direction of the P-wave radiation mostly owing to SP scattering near the hypocenter. Both pseudo P and S phases appear even on the null axis. These synthetic seismogram envelopes are in harmony with characteristics of observed seismogram envelopes of local earthquakes. Scattering due to random inhomogeneities quantitatively explains the frequency dependence of S-wave attenuation. From the analysis of seismogram envelopes around direct S waves of earthquakes with intermediate hypocentral distances, it was recently found that the time lag of the maximum amplitude arrival from the S-wave onset increased and the envelope broadened with travel distance. Diffraction due to slowly varying velocity inhomogeneities is modeled using the parabolic approximation, and is used to explain the envelope around the S-wave arrival observed at intermediate distances. A nonlinear inversion of the full seismogram envelope is used for the evaluation of randomness. Thus, the analysis of seismogram envelopes offers us useful information on the randomly inhomogeneous structure of the lithosphere complementary to the array observation analysis of individual phases.

Subducting plate boundary beneath the northeastern Japan arc estimated from SP converted waves

In the Tohoku District, the northeastern part of Honshu, Japan, a conspicuous phase is frequently observed between P and S phases in seismograms for events in the lower seismic plane of the double-planed deep seismic zone. This phase is dominant in the vertical component and has an apparent velocity almost equal to or slightly smaller than the P phase. We have never found this phase in seismograms for events in the upper seismic plane. From these characteristics, the phase is interpreted as SP, converted from S to P at the upper boundary of the descending oceanic plate (UBP). Using the travel times of the SP waves, we estimated the location of the UBP beneath the Tohoku District to be just above the upper seismic plane of the double-planed deep seismic zone. This position is consistent with those determined from the analyses of the travel times of the ScSp and PS waves, which are respectively the seismic wave converted from ScS to P and that from P to S at the UBP.

Attenuation of Love-waves in the low-period range near volcanic island margin

Generally Love-wave is absent in the corresponding spectrum of the phase in the low period range when the path tranversed is oceanic. ButOliver, Ewing & Press were able to observe this phase in seismograms of volcanic islands like Honolulu, Bermuda and others. Here, it has been attempted to explain why these waves could be recorded at volcanic stations and not elsewhere. It has been supposed that the lava settles down to form a transversely isotropic layer. Green's Function technique has been adopted to solve the problem of propagation of Love-waves through a crust whose thickness gradually decreases towards the ocean.

Wave Generations from Line Sources within the Ground

Wave generations from line sources within the ground are studied. Numerical details are worked out for the line source of step function type in time. The results to be noted are as follows;(1) At the epicenter, the displacement is a simple pulse followed by a gradual decrease to a permanent displacement.(2) RAYLEIGH waves appear at a certain distance from the epicenter.(3) Permanent displacements are rather large even at points far away from the origin. They are comparable with the amplitudes of RAYLEIGH waves at the respective paints.(4) There is no S phase in seismograms to be obtained for the wave origin of dilatational type.

Wave Generations from Line Sources within the Ground.

Wave generations from line sources within the ground are studied. Numerical details are worked out for the line source of step function type in time. The results to be noted are as follows;(1) At the "epicenter", the displacement is a simple pulse followed by a gradual decrease to a permanent displacement.(2) RAYLEIGH waves appear at a certain distance from the epicenter.(3) Permanent displacements are rather large even at points far away from the origin. They are comparable with the amplitudes of RAYLEIGH waves at the respective paints.(4) There is no S phase in seismograms to be obtained for the wave origin of dilatational type.

Unexplained phases in seismograms*

Summary Seismograms of more than 100 deep-focus earthquakes originating at epicentral distances from 1° to about 30° from Huancayo, Peru, and recorded there, are used in an investigation of unexplained phases. Three groups of waves, following P after about 5, 10, and 20 seconds, are possibly waves which have changed from longitudinal to transverse at depths of roughly 40, 80, and 150 km., respectively. Other phases may be waves reflected at a depth of between 900 and 1,000 km. The possibility of a sudden small decrease in wave velocity there is discussed. Other observed unexplained phases are mentioned. Reasons are discussed for usually minor disagreements between observed and calculated times, as well as amplitudes of phases in seismograms.