Search

This study presents an approach to better characterize the P‐wave and S‐wave velocity structure of the seafloor sediment layer using ocean bottom seismometers. The presence of low‐velocity seafloor sediment layers influences the observed seismic record at the seafloor over a broad frequency range, such that detailed knowledge of this sediment structure is essential to predict its effect on teleseismic records. We use the radial component of teleseismic P waves and autocorrelation functions of the radial, vertical, and pressure components of teleseismic P and S waves to obtain sediment layer models using the Markov chain Monte Carlo approach with parallel tempering. Synthetic tests show that the body waves constrain the P‐ and S‐wave impedances and travel times and the P‐ to S‐wave velocity ratio of the sediment layers. The proposed method resolves thin layers at a high resolution, including the uppermost thin (∼50 m to a few hundred meters) low S‐wave velocity layer. Real data applications at sites across the Pacific Ocean that are coincident with previous in situ studies demonstrate the effectiveness of this method in characterizing the seafloor sediment unit. The sediment models characterized by this new approach will allow us to more accurately predict and correct the effects of sediment layers in generating P‐ and S‐wave reverberations. Plain Language Summary: Ocean bottom seismometers are powerful tools for investigating the detailed structure of the oceanic mantle. However, most ocean bottom seismometers are placed above the low‐velocity water‐saturated sediment layer, such that their seismic records are inevitably affected by this seafloor sediment layer. Detailed constraints on the structure of the seafloor sediment layer allow us to accurately predict its effects on the observed seismic records. The seafloor sediment layer may consist of multiple discontinuities and vary regionally. This study proposes a new approach for characterizing the seismic properties of the seafloor sediment layer using teleseismic body waves recorded at ocean bottom seismometers. Synthetic tests show that the method can resolve thin sublayers within the sediment layer, such as a very thin low shear‐wave velocity layer and sharp seismic velocity jumps. Real data applications show that the sediment models obtained using this approach are in good agreement with those obtained from direct measurements and previous body‐wave analyses. Our proposed method will enable more detailed prediction and correction of the influence of the sediment layer on ocean bottom seismometer records, thereby providing more accurate constraints on deeper Earth structures. Key Points: The seismic properties of the seafloor sediment layer are characterized using ocean bottom seismometer teleseismic body‐wave recordsThe P‐ and S‐wave impedances and travel times and the VP/VS values of each sublayer within the sediment layer are resolvedIn situ seafloor sediment structure constraints will enable accurate correction and prediction of shallow structure effects [ABSTRACT FROM AUTHOR]

This study presents reflectivity images of the northeast (NE) Japan subduction zone continuous across the ocean and land. As nearly half of its forearc region is under the ocean, data from ocean bottom seismometers (OBSs) must be utilized to fully image the region by passive seismic analysis. The use of OBS data has been a challenge due to inherent characters of the ocean bottom observations: high noise level and effects of seafloor sediment. Now, decent imaging is possible in NE Japan overcoming the high level noise due to the accumulated data set of the OBSs. The low‐velocity of seafloor sediment significantly delays and amplifies S waves passing through them, and thus complicates teleseismic waveforms. We identify and correct these effects to produce coherent receiver function images throughout the amphibious subduction zone. Our images provide a potential for discussing new structural features and will help better understanding of the dynamics of the NE Japan subduction zone. Plain Language Summary: In northeast Japan, the Pacific plate is subducting beneath the Japanese arc, and delineating how the actual subduction is taking place at depths has been the target of many seismological studies. This study presents our effort of continuous imaging of subsurface interfaces across the ocean and land employing an amphibious data set. On the land side, there are dense seismic networks that have been used to reveal the seismic interfaces using converted phases of distant earthquakes. On the ocean side, the accumulation of data from temporary ocean bottom seismometers in the last decade and the recently installed ocean bottom cable network has made it possible to use a similar approach as on the land. However, the low‐velocity sediment beneath the seafloor causes systematic slower arrivals and larger amplitudes of the seismic phases of interest that result in highly disturbed images if we analyze them in the same way as the land. Here, we identify and correct the effects of the seafloor sediment to analyze the seismic interfaces across the ocean and land. As a result, we succeed in imaging the interfaces of the subducting oceanic crust throughout the region, as well as previously unidentified features under the ocean that may deserve future investigation. Key Points: The northeast Japan subduction zone is imaged using receiver functions (RFs) of an amphibious data setWe identify significant amplitude and time delay effects on RFs from seafloor sedimentCorrections of the sediment effects significantly improve images across an amphibious region [ABSTRACT FROM AUTHOR]

We performed waveform inversion of the P waveforms recorded by our BroadBand Ocean Bottom Seismometers (BBOBSs) deployed in the Northwestern Pacific. Consequently, the depth profile of the P velocity of the oceanic upper mantle, which has not been well resolved by previous surface wave or receiver function analyses, was revealed. We considered the azimuthal anisotropy in the lithosphere, which significantly improved the variance reduction from 34% to 44%. The resulting P model exhibited higher and lower velocities in the lithosphere and asthenosphere, respectively. The velocity contrast was found to be second/third of that observed in the previous S models; however the obtained model appeared to have some trade‐off with the VS structures in the vicinity of the source. We compared our P model with the previous S model obtained using our BBOBSs and obtained the VP/VS model. The resulting VP/VS model has two notable features. First, the lithosphere is characterized by a rapid increase in the VP/VS values with depth, which implies chemical stratification. Second, the VP/VS values in the vicinity of the lithosphere‐asthenosphere boundary (LAB) are larger than the synthetic values of any major upper mantle mineral predicted by considering the anharmonic effects, which suggests the effects of anelasticity or melt. Key Points: We inverted P waveforms recorded by our BBOBSs to obtain the depth profile of the P velocity beneath the older Pacific plateCombining our P model with a previous S model revealed two notable features in Vp/VS structuresThe obtained Vp/Vs model suggests chemical stratification in the lithosphere and the effects of anelasticity or melt at the LAB [ABSTRACT FROM AUTHOR]

We investigate slab and mantle structure of the NE Japan subduction zone from P wave azimuthal and radial anisotropy using travel time tomography. Trench normal E-W-trending azimuthal anisotropy (AA) and radial anisotropy (RA) with VPV > VPH are found in the mantle wedge, which supports the existence of small-scale convection in the mantle wedge with flow-induced LPO of mantle minerals. In the subducting Pacific slab, trench parallel N-S-trending AA and RA with VPH > VPV are obtained. Considering the effect of dip of the subducting slab on apparent anisotropy, we suggest that both characteristics can be explained by the presence of laminar structure, in addition to AA frozen-in in the subducting plate prior to subduction. [ABSTRACT FROM AUTHOR]

Abstract: We report anomalous core phase PKPbc‐PKPdf differential travel times relative to 1‐D spherically symmetric model with a uniformly anisotropic inner core recorded by stations in Alaska for South Sandwich Islands (SSI) earthquakes. The data sample the inner core for the polar paths, as well as the lowermost mantle beneath Alaska. Our major observations are the following: (1) fractional travel time residuals of PKPbc‐PKPdf increase rapidly within 2 °, (2) a clear shift of the residual pattern could be seen for earthquakes with different locations, and (3) the residuals show systematic lateral variation: at the northern part, no rapid increase of residual can be seen. A structural boundary with a P wave velocity contrast of about 3% at the lowermost mantle beneath East Alaska is invoked to explain the observation, and the required strength of anisotropy in the quasi‐western hemisphere of the inner core might be reduced if those anomalous data are excluded from analysis. [ABSTRACT FROM AUTHOR]

We constrain olivine fabrics in the oceanic lithosphere using active and passive seismic observations of Pn azimuthal anisotropy. We first analyze active‐source data derived from a broadband ocean bottom seismometer array that was deployed in the Northwestern Pacific. We infer the azimuthal dependence of the Pn velocity, including the 2ϕ and 4ϕ terms of the sinusoidal functions, where ϕ is the back azimuth. We observe a skewed azimuthal dependence of the Pn velocity, with a large peak‐to‐peak amplitude of about 12%. Similar features are observed using an independent data set of teleseismic Pn waves. We constrain the direction of the crystallographic axes of olivine to explain the observed azimuthal dependence and identify A‐type olivine with a slightly dipping a‐axis and slightly tilting b‐axis being compatible with our observations. In contrast, we find that D‐type olivine with any direction of crystallographic axes cannot explain our observations. Secondary deformation and recrystallization in the older Pacific may be responsible for this strong and complex seismic anisotropy in the study region. Plain Language Summary: Olivine is the most abundant mineral in the oceanic upper mantle and exhibits preferred crystallographic orientations when it experiences plastic strain‐induced deformation (Nicolas & Christensen, 1987, https://doi.org/10.1029/GD016p0111). Olivine in the oceanic lithosphere is considered to retain the orientations from its formation at the mid‐ocean ridge unless it experiences tectonic perturbations such as reheating by hotspot volcanisms. Seismic anisotropy observations in the oceanic lithosphere therefore provide clues to plate tectonics processes. In this study, we analyze two independent data sets (active‐ and passive‐source data) and reveal the detailed azimuthal dependence of the P velocity traversing the oceanic lithosphere in the northwestern Pacific. We find that the two data sets consistently show a strong and skewed anisotropic signal. We infer that secondary deformation and recrystallization in the Pacific Plate may be responsible for these observations. Key Points: Strong and skewed azimuthal dependence of Pn velocity is robustly observedCompatible direction of the crystallographic axes of olivine is constrainedSecondary deformation and recrystallization in the older Pacific is suggested [ABSTRACT FROM AUTHOR]

We present a new reference model for the evolution of oceanic lithosphere, which incorporates the effects of incomplete viscous relaxation, radiogenic heating, and secular cooling. The new reference model is based solely on thermal conduction, that is, without involving the occurrence of small‐scale convection, and unlike the plate model, it does not contain unphysical boundary conditions. Yet, our model can explain both bathymetry and the heat flow data on the normal seafloor. The success of the new model owes to the use of realistic material properties in conduction modeling as well as the consideration of all of major processes that take place ubiquitously beneath seafloor. The effect of secular cooling on the bathymetry of old seafloor is particularly notable. Whereas secular cooling brings only weak temperature variations with an amplitude of ∼20 K, it can nonetheless affect global bathymetry substantially owing to the deep sensitivity of long‐wavelength topography kernels. We suggest that the well‐known fact that Earth has been cooling, which was not considered in any of previous reference models, may be the key to the long‐standing puzzle of seafloor flattening. The new reference model is expected to be useful to better quantify the impact of the emplacement of hotspot islands and oceanic plateaus, the effect of small‐scale convection, and the regional history of secular cooling in the convecting mantle. Plain Language Summary: Seafloor deepens as it moves away from mid‐ocean ridges, and this subsidence reflects how the suboceanic mantle loses its heat by conduction. When seafloor becomes older than 70 million years old, however, it starts to deviate upward from what is predicted by the simple law of thermal conduction. A common approach to model such a deviation is to adopt the so‐called plate model, which can suppress seafloor subsidence with a constant temperature boundary condition at a shallow depth (120–140 km), although the real Earth does not contain such a boundary. Here we show that, by taking into account all of major processes intrinsic to the suboceanic mantle, from the cold shallow part to the hotter deep part, it is possible to explain the evolution of seafloor topography as well as heat loss, without invoking an unphysical boundary condition. In particular, this study illustrates that the fact that Earth is cooling, which is long known in Earth sciences, can have considerable effects on the large‐scale behavior of ocean basins. Key Points: A new reference model for the evolution of normal oceanic lithosphere is proposedThe model is based solely on thermal conduction and yet free of unphysical boundary conditionsThe model incorporates the effects of incomplete viscous relaxation, radiogenic heating, and secular cooling [ABSTRACT FROM AUTHOR]

We analyze seismic records collected at the oldest (170–180 Ma) Pacific seafloor using broadband dispersion array analysis. Using ambient noise and teleseismic waveforms, we measure Rayleigh‐wave phase velocities in a period range of 5–200 s that are inverted for array‐average one‐dimensional isotropic and azimuthally anisotropic shear‐wave velocity depth profiles from the crust to a depth of 300 km. The high‐velocity Lid and the low‐velocity zone are well‐resolved with velocity difference of ∼4%, whose transition occurs at depths between 80 and 100 km. The profile is compared with that obtained at the 130‐ and 140‐Ma seafloor. Accounting for the cooling effect due to the plate age difference, the low‐velocity zone of the oldest Pacific seafloor is ∼1.3% slower (∼110°C warmer) than that beneath the 140‐Ma seafloor, suggesting the occurrence of some reheating process beneath the oldest lithosphere. The azimuthal anisotropy at shallow depths (<50 km) is significantly different between the western and eastern areas of the array where the peak‐to‐peak amplitudes are estimated to be ∼2.8% and ∼1.6%, respectively. The fast direction is nearly parallel to the past seafloor spreading direction (perpendicular to the magnetic lineation) in the west, while it largely deviates in the east. The observed difference in azimuthal anisotropy may represent complicated evolution dynamics of the infant Pacific plate that involved a ridge‐ridge‐ridge triple junction. Plain Language Summary: One of the classical enigmas of plate tectonics is the cause of seafloor flattening where the seafloor depth becomes nearly constant for old ages (>∼70 Ma) while a simple cooling model of plate evolution predicts the age dependence. To solve this enigma, it is essential to resolve the subsurface structure of the oceanic plate and the asthenosphere below. As a part of an international effort, Pacific Array, that aims to cover the Pacific seafloor by arrays of ocean bottom seismometers, a team of Japanese and Korean scientists deployed an array in the oldest (170–180 Ma) Pacific seafloor. We determine isotropic and anisotropic shear‐wave velocity structures as a first step toward resolving the enigma. The results indicate that the asthenosphere beneath the oldest Pacific lithosphere is significantly slower than the prediction of the simple cooling and that the occurrence of some reheating process may be needed. Furthermore, azimuthal anisotropy (the dependence of the wave velocity on the horizontal propagation direction) in the shallow lithosphere shows that the anisotropy patterns differ between the eastern and western areas of the array, and simulation results indicate that the evolution process of the infant Pacific plate is not merely a simple triangular plate spreading. Key Points: Broadband Rayleigh‐wave dispersion array analysis has been conducted in the oldest (∼170 Ma) Pacific seafloorThe low‐velocity zone is significantly slower to indicate the occurrence of some reheating process beneath the oldest Pacific lithosphereThe spatial variation of the shallow azimuthal anisotropy indicates a complicated evolution of the infant Pacific plate [ABSTRACT FROM AUTHOR]

Investigation of fluid distribution along megathrust faults is an important issue, since the fluid affects frictional properties and thus slip behavior on faults. Scattered teleseismic phases, or receiver functions (RFs), have made significant contributions to understanding the fluid content of subducting plates beneath the onshore regions but have been rarely applied in offshore settings. In this study, we conducted receiver function inversion analysis to investigate detailed seismic properties near the megathrust fault using ocean bottom seismometers deployed off the Kii Peninsula, southwest Japan. RFs were calculated at high frequencies (up to 4 Hz), removing the effect of water reverberations from vertical component records. Our inversion was performed in two steps: first, we modeled sediment layer by a simple stacking method and then solved for deeper structure by a waveform inversion. The results indicate the presence of a thin low-velocity zone (LVZ) of a thickness of 0.2-1.2 km with a S wave velocity of 0.7-2.4 km/s along the plate interface. We interpret this LVZ as thin fluid-rich sediment layer between the overriding and subducting plates that acts as a pathway of fluid migration. [ABSTRACT FROM AUTHOR]

Strong long‐period seismic signals at periods around 25 and 18 s appear in the ambient noise cross‐correlation functions recorded by an array of ocean bottom seismometers (OBSs) in the western‐central Pacific. The signal amplitude varies from time to time, and the apparent travel times of the signals are typically smaller than those expected for the Rayleigh waves propagating along the great circle connecting the station pairs. From the cross‐correlation functions, the signal sources are located in the Vanuatu Arc. Local data analysis suggests the signals originate from two different sources possibly located at depths of ~0–1 km below the sea level beneath the active cones of Ambrym volcano. Plain Language Summary: Spatially localized sources that persistently generate long‐period seismic waves at periods longer than 10 s are reported worldwide. For some sources, the excitation mechanisms have been well‐investigated. A volcano in Japan generates such waves that travel more than thousands of kilometers. For others, however, the excitation origins are still poorly understood. These include the one observed in the Vanuatu Arc. A temporal observation network of seismometers installed on the seafloor in the western‐central Pacific observed persistent waves peaking at around two different periods. From the data analysis, we find the signals originate from an active volcano, Ambrym volcano, in the Vanuatu Arc. Besides, we find that those seismic waves are most likely originated from two different sources possibly located at shallow depths beneath the volcano. Key Points: Persistent long‐period (25 and 18 s) seismic signals appear in ambient noise cross‐correlations of OBS data in the western‐central PacificThe signals are originated from Ambrym volcano in the Vanuatu ArcSource locations of the two signals may be different, but both are possibly located at shallow depths right beneath the active cones [ABSTRACT FROM AUTHOR]