Search for: All records

SUMMARY We present a new 3-D radially anisotropic seismic velocity model EARA2024 of the crust and mantle beneath East Asia and the northwestern Pacific using adjoint full-waveform inversion tomography. We construct the EARA2024 model by iteratively minimizing the waveform similarity misfit between the synthetic and observed waveforms from 142 earthquakes recorded by about 2000 broad-band stations in East Asia. Compared to previous studies, this new model renders significantly improved images of the subducted oceanic plate in the upper mantle, mantle transition zone, and uppermost lower mantle along the Kuril, Japan, Izu-Bonin and Ryukyu Trenches. Complex slab deformation and break-offs are observed at different depths. Moreover, our model provides new insights into the origins of intraplate volcanoes in East Asia, including the Changbaishan, Datong-Fengzhen, Tengchong and Hainan volcanic fields. 

SUMMARY Applications of machine learning in seismology have greatly improved our capability of detecting earthquakes in large seismic data archives. Most of these efforts have been focused on continental shallow earthquakes, but here we introduce an integrated deep-learning-based workflow to detect deep earthquakes recorded by a temporary array of ocean-bottom seismographs (OBSs) and land-based stations in the Tonga subduction zone. We develop a new phase picker, PhaseNet-TF, to detect and pick P- and S-wave arrivals in the time–frequency domain. The frequency-domain information is critical for analysing OBS data, particularly the horizontal components, because they are contaminated by signals of ocean-bottom currents and other noise sources in certain frequency bands. PhaseNet-TF shows a much better performance in picking S waves at OBSs and land stations compared to its predecessor PhaseNet. The predicted phases are associated using an improved Gaussian Mixture Model Associator GaMMA-1D and then relocated with a double-difference package teletomoDD. We further enhance the model performance with a semi-supervised learning approach by iteratively refining labelled data and retraining PhaseNet-TF. This approach effectively suppresses false picks and significantly improves the detection of small earthquakes. The new catalogue of Tonga deep earthquakes contains more than 10 times more events compared to the reference catalogue that was analysed manually. This deep-learning-enhanced catalogue reveals Tonga seismicity in unprecedented detail, and better defines the lateral extent of the double-seismic zone at intermediate depths and the location of four large deep-focus earthquakes relative to background seismicity. It also offers new potential for deciphering deep earthquake mechanisms, refining tomographic models, and understanding of subduction processes. 

Abstract Surface wave tomography is widely used to improve our understanding of continental magma reservoirs that may be capable of fueling explosive volcanic eruptions. However, traditional surface wave tomography based on inversions for phase velocity maps and locally 1D shear velocity may have difficulty resolving strong 3D low‐velocity anomalies associated with crustal magma reservoirs. Here, we perform synthetic tomography experiments based on 3D seismic waveform simulations to understand how the limitations of surface wave tomography could affect interpretations of tomography in volcanic settings. We focus our modeling on the Yellowstone volcanic system, one of the largest and most thoroughly studied continental magmatic systems, and explore scenarios in which the maximum shear velocity anomaly associated with the crustal magma reservoir ranges between −10% and −66%. We find that even with the well‐instrumented setting near Yellowstone, the recovered shear velocity anomalies in the mid‐to‐upper crust are severely diminished due to the small spatial scale of the reservoir with respect to the seismic wavelengths that sample it. In particular, recoveredV_Sanomalies could be reduced by a factor of two or more, implying that the inferred melt fraction of large‐scale continental magma reservoirs may be considerably underestimated. 

Abstract The lithospheric structure of the contiguous US and surrounding regions offers clues into the tectonic history, including interactions between subducting slabs and cratons. In this paper, we present a new radially anisotropic shear wave speed model of the upper mantle (70–410 km) of the contiguous US and surrounding regions, constrained by seismic full‐waveform inversion. The new model (named CUSRA2021) utilizes frequency‐dependent travel time measurements, from 160 earthquake events recorded by 5,280 stations. The data coverage in eastern US is improved by incorporating more intraplate earthquakes. The final model exhibits clear and detailed shear wave speed anomalies correlating well with tectonic units such as North America Craton (high‐Vs), Cascadia subduction zones (high‐Vs), Columbia Plateau (low‐Vs), Basin and Range (low‐Vs), etc. In particular, the detailed structure of the North America Craton beneath Illinois basin is revealed. The depth of high‐Vs anomaly beneath the North America Craton correlates well with S‐to‐P receiver function and SH reflection results. Besides, the radial anisotropy in the Craton lithosphere shows a layering structure, which may relate to the process of lithospheric accretion and the origin of mid‐lithosphere discontinuities. 

Abstract Accurate seismic images of the crust are essential for assessing seismic hazards and elucidating tectonic processes that shape surface landforms. Although California and Nevada have been studied extensively using various seismic datasets and tomographic methods, the region lacks a seismic model that can accurately define both the shallow (<8 km) and deeper crust. We take the advantage of recent increases in seismic data coverage to build a new 3D shear wave speed model by jointly inverting Rayleigh wave ellipticity, phase velocity, and teleseismic P waveforms. In the Great Valley, the new model reveals an asymmetric basement, steeply dipping in the west and gently dipping in the east. Beneath its western margin, in the Coast Ranges, we resolve a wedge‐shaped, low‐velocity zone in the upper‐middle crust, interpreted as Franciscan Complex. Our images confirm that uplift of the western Great Valley and an eastward shift of its depositional center are caused by wedging and underthrusting of the complex during subduction. Across the Basin and Range, the resolved crust has an average thickness of 38 km in the southern half of the northern Basin and Range, about 5 km thicker than neighboring regions. The thickened crust overlaps with major volcanic centers of the mid‐Cenozoic ignimbrite flare‐up. This spatial correlation may suggest magmatic intrusions and underplating contributed to crustal growth and thickening prior to Miocene Basin and Range extension. Overall, the new model is consistent with active source studies in the region but provides a more comprehensive view of shallow and deep structures across this large and tectonically complex region. 

Seismic imaging shows a melt fraction of up to 20% in the depth range that supplied prior Yellowstone eruptions. 

Abstract The SPECFEM3D_Cartesian code package is widely used in simulating seismic wave propagation on local and regional scales due to its computational efficiency compared with the one-chunk version of the SPECFEM3D_Globe code. In SPECFEM3D_Cartesian, the built-in meshing tool maps a spherically curved cube to a rectangular cube using the Universal Transverse Mercator projection (UTM). Meanwhile, the geodetic east, north, and up directions are assigned as the local x–y–z directions. This causes coordinate orientation issues in simulating waveform propagation in regions larger than 6° × 6° or near the Earth’s polar regions. In this study, we introduce a new code package, named Cartesian Meshing Spherical Earth (CMSE), that can accurately mesh the 3D geometry of the Earth’s surface under the Cartesian coordinate frame, while retaining the geodetic directions. To benchmark our new package, we calculate the residual amplitude of the CMSE synthetics with respect to the reference synthetics calculated by SPECFEM3D_Globe. In the regional scale simulations with an area of 1300 km × 1300 km, we find a maximum of 5% amplitude residual for the SPECFEM3D_Cartesian synthetics using the mesh generated by the CMSE, much smaller than the maximum amplitude residual of 100% for the synthetics based on its built-in meshing tool. Therefore, our new meshing tool CMSE overcomes the limitations of the internal mesher used by SPECFEM3D_Cartesian and can be used for more accurate waveform simulations in larger regions beyond one UTM zone. Furthermore, CMSE can deal with regions at the south and north poles that cannot be handled by the UTM projection. Although other external code packages can be used to mesh the curvature of the Earth, the advantage of the CMSE code is that it is open-source, easy to use, and fully integrated with SPECFEM3D_Cartesian.