Yong Yu

The 15 January 2022 submarine volcanic eruption of Hunga Tonga‐Hunga Ha'apai released immense energy throughout the ocean, solid Earth, and atmosphere. We analyze mid‐oceanic column acoustic pressure recordings from 24 freely drifting Mobile Earthquake Recorder in Marine Areas by Independent Divers sensors, and from 11 moored hydrophones in the International Monitoring System. We focus on the pulsed hydroacoustic phase which propagated horizontally through the ocean as a 30‐min T wave with energy around 2.5–10 Hz. The records show high correlation between some receivers, significant variation among others, and varying amplitudes that cannot be explained by distance alone. We investigate the origin of this heterogeneity via the influence of bathymetric features that may block, or occlude, T ‐wave propagation, affecting both shape and amplitude of the records received. We count the number of seafloor obstacles within the horizontal plane of the first (ray‐theoretical) Fresnel zone at a depth of 1,350 m, where the fundamental‐mode T ‐wave eigenfunction is maximal. Adjusted for geometric spreading, the cross‐correlations and sound pressure level differences between receivers systematically relate to differences in occlusion count. Our model of signal loss due to seafloor interactions predicts a 5.6 dB reduction in sound pressure level per logarithm of occlusion count, explaining 88% of the T ‐wave sound pressure variance across the ocean. Source characterization requires adequate path models. Our findings describe how to correct signal amplitudes for seafloor roughness. This is important for constraining volcanic or explosive yield estimates and earthquake magnitudes, and useful to model detectability through various oceanic corridors when designing hydroacoustic monitoring networks of the future.

The 15 January 2022 submarine volcanic eruption of Hunga Tonga‐Hunga Ha'apai released immense energy throughout the ocean, solid Earth, and atmosphere. We analyze mid‐oceanic column acoustic pressure recordings from 24 freely drifting Mobile Earthquake Recorder in Marine Areas by Independent Divers sensors, and from 11 moored hydrophones in the International Monitoring System. We focus on the pulsed hydroacoustic phase which propagated horizontally through the ocean as a 30‐min T wave with energy around 2.5–10 Hz. The records show high correlation between some receivers, significant variation among others, and varying amplitudes that cannot be explained by distance alone. We investigate the origin of this heterogeneity via the influence of bathymetric features that may block, or occlude, T ‐wave propagation, affecting both shape and amplitude of the records received. We count the number of seafloor obstacles within the horizontal plane of the first (ray‐theoretical) Fresnel zone at a depth of 1,350 m, where the fundamental‐mode T ‐wave eigenfunction is maximal. Adjusted for geometric spreading, the cross‐correlations and sound pressure level differences between receivers systematically relate to differences in occlusion count. Our model of signal loss due to seafloor interactions predicts a 5.6 dB reduction in sound pressure level per logarithm of occlusion count, explaining 88% of the T ‐wave sound pressure variance across the ocean. Source characterization requires adequate path models. Our findings describe how to correct signal amplitudes for seafloor roughness. This is important for constraining volcanic or explosive yield estimates and earthquake magnitudes, and useful to model detectability through various oceanic corridors when designing hydroacoustic monitoring networks of the future.

We have inverted a very large combination of delay times for P, pP, PcP, PKPab, PKPbc from the ISC-EHB catalog, augmented with hand-picked OBS and MERMAID onset data and crosscorrelation delays from broadband stations to derive a new global model, UNICA25, of P-wave velocity variations in the mantle. This paper presents the model, summarizes the data processing and inversion, and shows resolution tests at different scales. After rejecting 109 917 outliers (1% of the total), the data set comprises 10 571 152 arrival times. The misfit of the observed delays, with respect to a background model consisting of the crust of LITHO1.0 on top of model AK135, is on average 2.14 standard errors (σ), which is reduced to 0.99σ by inverting for P-wave velocity variations using both ray theory and finite frequency theory, and 5000 iterations of the iterative solver LSQR. Extensive resolution tests show a resolution at 500 km scale or better in most of the lower mantle. An upper mantle resolution of the order of 300 km is obtained under densely instrumented continents, but upper mantle resolution remains deficient under much of the oceans except were covered by MERMAIDs. Linearized tomography is at least three orders of magnitude faster than recent full waveform inversions (FWI), and reaches much higher frequencies for cross-correlated delays. Model UNICA25 has a P-velocity resolution that rivals that of FWI and shows a superior performance in fitting delay times, notably the ISC-EHB P-delays.

In this study, we obtain a three‐dimensional (3‐D) high resolution seismic anisotropy model of NE Tibet using ambient noise double beamforming (DBF) tomography with ChinArray‐Himalaya II data. DBF tomography technique allows for the simultaneous derivation of local phase velocity and azimuthal anisotropy from ambient noise cross‐correlations without performing tomographic inversion. Based on the DBF tomography results, we have constructed a high resolution 3‐D azimuthal anisotropic shear wave velocity model of NE Tibet. Our model indicates the presence of prominent low velocities and strong azimuthal anisotropy in the mid‐to lower crust beneath the Songpan‐Ganzi belt, while the mid‐to lower crust of the Qilian belt exhibits relatively weak low velocity and azimuthal anisotropy. We propose that crustal channel flow may be the primary mechanism driving crustal deformation in the Songpan‐Ganzi belt, which is in the mature part of the plateau. As the Qilian belt is at an early stage of plateau formation, it is characterized by crustal shortening with the development of a sub‐vertical foliation. Additionally, the misalignment between predicted and observed shear wave splitting reveals decoupled crust and upper mantle deformation west of the Ordos block where eastward mantle flow is deflected by the lithospheric keel of Ordos.

We obtained a 3‐D crustal radial anisotropy model beneath northeastern (NE) Tibet by joint inversion of Rayleigh and Love dispersion curves from ambient noise tomography. Positive crustal radial anisotropy and significant low velocity are dominant beneath the Songpan‐Ganzi Terrane (SGT), indicating the presence of crustal channel flow. The Qilian orogen is characterized by negative radial anisotropy, which could be caused by folding and thrusting due to crustal shortening. This difference in radial anisotropy suggests that crustal shortening deformation may have occurred at a relatively early stage of the plateau evolution (the Qilian orogen) and crustal channel flow at the later stage (the SGT). The crustal radial anisotropy clearly defines the NE crustal boundary of the expanding Tibetan Plateau, which is roughly situated along the north Qilian frontal thrust from 97° to 103.5°E in the north and turns north‐south, passing the Zhuanglanghe Fault and terminating at the Kunlun Fault in the south.

During the last 50 Ma, the eastern Eurasian continent has experienced widespread crustal and lithospheric deformation. However, mechanisms in creating such widespread intra‐continental deformation are still not well understood. Here, we present a 3‐D S‐wave velocity model of the northeastern Tibetan Plateau and adjacent regions by jointly inverting teleseismic body and surface waves. The resulting model clearly suggests that lateral extrusion of asthenosphere from northern Tibet is being blocked by the thick cratonic keels of Ordos and Sichuan, and therefore the asthenosphere is channeled into a strong eastward flow beneath the northern Qinglin between these two cratonic keels. Our results also demonstrate that continental deformation in eastern Eurasia is likely influenced by the small‐scale convection when the lateral asthenospheric flow from NE Tibet encounters preexisting continental lithospheric steps to the east of Ordos.

The collision between India and Eurasia forming the Tibetan Plateau since the early Cenozoic had led to the crustal and upper‐mantle (asthenospheric) extrusion from the plateau to the east. However, the Ordos and Yangtze cratonic keels next to the northeastern and eastern Tibetan Plateau obstruct this eastward asthenospheric extrusion. With a dense seismic array, for the first time, we imaged a narrow low P‐wave and S‐wave velocity anomaly characterized with higher Vp/Vs between these two Archean cratons at depths ranging from 150 to 300 km using finite‐frequency tomography. We interpret this low‐velocity anomaly as a channel of the eastward asthenospheric flow from the Tibetan Plateau. We provide solid seismic observations for a pathway of the Tibetan asthenospheric extrusion due to the ongoing collision of the Indian and Eurasian continents. This asthenospheric extrusion from the Tibetan Plateau could influence the lithospheric reworking in the central and eastern North China Craton.