Valentí Sallarès

In subduction zones, seismic imaging reveals increasing permanent active deformation toward the trench, particularly within accretionary systems. Yet, how these rock bodies deform coseismically and influence megathrust rupture behavior is elusive. Here we combine geophysical observations from seismic imaging with visco-plastic dynamic rupture simulations to investigate how realistic upper-plate rock bodies influence megathrust earthquake dynamics and off-fault deformation. Our models reproduce the elastic structure of three subduction systems that differ primarily in the width of the accretionary prism, a key parameter for comparison. These configurations, referred to as Models I, II, and III, include a narrow compliant prism of approximately 20 km, an intermediate prism of about 60 km, and a wide prism exceeding 100 km, respectively. Elastic rock properties for each upper-plate model are derived from 2D P-wave velocity models obtained from controlled-source seismic data. Upper-plate bulk cohesion and bulk friction define visco-plastic strength and are set to depend on rigidity distribution and empirical observations. Based on laboratory measurements from JFAST drilling samples, we use rate-and-state friction law with strong velocity weakening in the shallow portion of the fault.Results show that coseismic upper-plate plastic deformation in Model I is confined to the ~20-km-wide wedge, whereas in Models II and III it extends 40–60 km landward from the trench. This is consistent with seismic reflection profiles that reveal increasing active internal deformation of the prism at similar distances from the trench in regions such as the Japan Trench, Chile, and Sumatra. Such contrasting upper-plate deformation patterns lead to distinct uplift scenarios, particularly in their high-frequency response. In particular, Model I produces shorter-wavelength uplift near the trench, likely generating a tsunami with higher-frequency content than Model II and III, where uplift exhibits a longer wavelength. Although we do not explicitly simulate independent faults within the prism, the bulk plastic strain can be considered a proxy for the amount of deformation that is accommodated by these structures. Our results suggest that permanent deformation within accretionary prisms is active during trench-breaching megathrust earthquakes, indicating that substantial prism deformation occurs coseismically. Plastic deformation leads to a reduction in slip toward the trench, implying that coseismic energy is absorbed by the overlying rock body. This effect explains the low-radiated-energy of near-trench earthquakes, including tsunami earthquakes. Depending on the plastic strength of upper-plate material and the available energy along the fault, this effect may even prevent the rupture from reaching the trench, while still producing substantial coseismic uplift and horizontal seafloor displacement. Overall, this study indicates that identifying permanently deformed, low-rigidity regions near the trench can serve as a proxy for locating areas where coseismic deformation is strongly accommodated and tsunamigenic uplift is likely amplified.

The structure of the Newfoundland–West Iberian conjugate margins has been extensively studied during the past 50 years in hundreds of papers. The crustal structure has been evaluated through seismic surveys and drilling expeditions, but those are not equally distributed in Iberia and Newfoundland. More work, and in particular recent studies on the West Iberian margin, have identified a complex crustal architecture characterised by continental, oceanic, and exhumed mantle domains that vary along the margin. This structural complexity has only been recently documented with modern data that allow to image the basement domains in detail.In contrast, the Newfoundland basement remains comparatively less well understood due to a relative scarcity of seismic and drilling data. The main wide-angle and streamer data for this area, the SCREECH survey, were acquired in 2000 and modelled under the computational limitations of that time. The resulting models and images have been subject to debate and failed to unequivocally define the nature of the basement domains of the margins. This uncertainty has left open key questions regarding the evolution of deformation during rifting and, thus, also the degree of symmetry of this conjugate pair of margins.The SCREECH acquisition parameters were similar to modern marine acquisition standards. We leveraged their inherent data quality with the current computational facilities and up-to-date methodologies to re-process the data, imaging the structure and modelling seismic phases. Recent advancements in parallel computing and novel geophysical techniques now allow for enhanced-resolution seismic models and a mathematically robust uncertainty analysis—tasks that were previously very computationally demanding.Our study utilises the original SCREECH field data, consisting of three transects with coincident multichannel seismic (MCS) reflection data (6-km streamer) and wide-angle data recorded by short-period OBS and OBH stations at ~15 km spacing. By performing a joint inversion of the streamer and wide-angle data (utilising both reflection and refraction arrivals), we significantly improved the definition of geological units and the spatial resolution of the velocity models. A statistical uncertainty analysis was conducted to validate the reliability of these observed features.Our findings reveal previously unrecognised crustal heterogeneity at the Newfoundland margin, including significant variations in thickness and composition along the margin. Notably, we challenge prior classifications of the crustal domains and the location and dimensions of the Continent-Ocean Transition (COT). Previous models identified an intra-basement deep reflector as the Moho, defining a 4–5 km thick layer interpreted as continental crust. However, our results suggest this reflector may not represent the Moho, as the observed crustal properties are inconsistent with typical continental or oceanic crust, and rather support a COT formed by >250 km of exhumed mantle. By integrating MCS imagery with these new velocity models, we provide a re-interpretation of the margin’s crustal structure and propose a refined evolutionary model for the West Iberian–Newfoundland conjugate system.

The oceanic lithospheric structure is key to understanding the processes related to the interplay between tectonics and magmatism from the end of rifting to the onset and cessation of seafloor spreading. In the oceanic basin of the South China Sea (SCS), the Southwest Sub‐basin (SWSB) is the last‐opening segment before the abandonment of seafloor spreading. The SWSB shows a V‐shaped geometry associated with the SW‐ward propagation of the spreading center during the opening. Here, we present a 540‐km‐long wide‐angle seismic profile along a seafloor‐spreading flow‐line in the central sector of the SWSB. A joint refraction and wide‐angle reflection travel‐time tomography provides a P‐wave velocity (Vp) model of the crust and uppermost mantle. Continental break‐up produced asymmetric conjugate margins, abutted by a typical oceanic crust with a two‐layer Vp structure. The distinct vertical velocity gradients in the upper and lower crust, and a clear Moho, indicate a magmatically‐dominated oceanic system. In contrast, oceanic crust <19.2 Ma displays no Moho interface and a Vp structure indicating reduced magmatism and increased tectonic extension. Anomalously slow Vp structure under the median valley indicates serpentinized mantle resulting from enhanced faulting during the final opening phase. The spatial variation in Vp structure reflects two main temporal stages of the SWSB seafloor spreading before cessation. Previous results of plate kinematics reconstructions appear to support that the transition from magmatically robust to increased tectonics and final cessation of seafloor spreading in the SWSB might relate to the subduction of the proto‐SCS.

SUMMARY: We investigate the lithospheric structure of the Southwest Iberian Margin along an active seismic profile southwest of São Vicente Cape, ranging from the southern Tagus Abyssal Plain to the westernmost part of the Gulf of Cadiz. This profile, approximately 320 km long, intersects almost perpendicularly three major thrust faults: the Tagus Abyssal Plain, Marquês de Pombal and Horseshoe faults. The crustal structure, derived from spatially coincident wide-angle seismic (WAS) and multichannel seismic (MCS) data, was validated and constrained using gravimetric data. Joint traveltime inversion of refracted phases identified in WAS and reflected seismic phases from both WAS and MCS records were used to build a detailed 2-D P-wave velocity (Vp) structure. The resulting model reveals a Vp distribution with abrupt lateral velocity and structural variations, characterized by a rugged basement top and sharp changes in crustal thickness. Three main lithospheric domains consisting of continental, oceanic and exhumed mantle affinity were identified from south to north. The traveltime inversion of the deepest reflected seismic phases reveals four major southeast-dipping reflectors, likely corresponding to major regional thrust faults with significant seismic and tsunamigenic potential. Integrating the modelled and interpreted seismic results with the locations of recent well-constrained earthquakes suggests that the Marquês de Pombal and Tagus Abyssal Plain extend deeper than previously thought, with fairly high seismic activity in the deep levels. This has significant implications for their seismogenic potential and should be taken into account for accurate assessment of seismic hazards in the region.

The crustal structure of the Newfoundland-West Iberian conjugate margins has been extensively studied with seismic data and drilling legs. Recent surveys in the West Iberian margin have revealed a complex crustal architecture with continental, oceanic and exhumed mantle domains that change along the margin. In contrast, the Newfoundland margin, with lower seismic and drilling information available, remains comparatively more poorly understood. The main wide-angle and streamer SCREECH survey was acquired in 2000 and was modelled with comparative computational limitations at the time. The resulting images and model have been debated and did not unequivocally characterize the nature of the basement domains along the margin. Thus, the evolution of the deformation during rifting and the symmetry or asymmetry of the conjugate pair of margins are still discussed. The SCREECH data acquisition parameters are similar to modern data, and we took advantage of their quality to re-process, image the structure and model the seismic phases with methodologies that have been refined during the last decade. Recent developments in parallel computing and novel geophysical approaches provide now the means to obtain a new look at the structure with enhanced resolution seismic models and a mathematically-robust analysis of the data uncertainty, that was formerly difficult, if not unfeasible, to achieve. We use the SCREECH original field data, formed by three transects with coincident multichannel seismic (MCS) reflection data acquired with a 6-km streamer and wide-angle data recorded by short-period OBS and OBH spaced at ~15 km. We reprocessed the streamer data and also performed the joint inversion of streamer and wide-angle OBS/OBH seismic data, using reflections and refraction arrivals, which improved the definition of the geological units and the spatial resolution of the velocity model for each unit. We performed a statistical uncertainty analysis of the resulting model, supporting the improved reliability of the observed features. Our results reveal previously unrecognized crustal heterogeneity, including variations in crustal thickness and composition along the margin. In particular, the crustal domain classification and the COT location were done considering the existence of a deep reflector, interpreted as the Moho and defining a 4-5 km crust that was interpreted as oceanic. Our results suggest that this reflector may not represent the Moho, as the observed crustal properties are not consistent with typical oceanic crust. The integration of the MCS images with the velocity models allowed us to re-interpret the crustal structure of this margin and integrate all the observations in a refined evolution model for the West Iberian – Newfoundland conjugate margins.

Major sites of currently active serpentinization production are slow and ultra-slow mid-ocean ridges (MORs). At magma-poor margins, serpentinization takes place during mantle exhumation at the Continent Ocean Transition (COT), a phase occurring after continental breakup and before the establishment of a steady-state oceanic ridge. During this process, mantle rocks become intensely serpentinized as they interact with seawater that percolates through cracks and faults. However, the precise sequence of serpentinization formation in these regions remains still a topic of debate.In this study, we examine two well-known examples of mantle exhumation: the West Iberia margin, using the FRAME-P3 profile, and the Tyrrhenian back-arc basin, analyzed through the MEDOC-6 profile. Both profiles were recently acquired with high-resolution multi-channel seismic (MCS) data and are complemented by previously modelled wide-angle seismic (WAS) data. This integration significantly enhances the P-wave velocity model, providing a more detailed and refined view of the basement structure and challenging previous interpretations. The improved model offers better velocity resolution, particularly in regions with complex subsurface structures, and delivers a comprehensive characterization of the transition zone between the oceanic crust and exhumed mantle.This analysis revealed unprecedented details such as the extent of serpentinization and the detection of a continuous reflector indicative of a hydration front. They also highlight significant fluid-rock interactions and serpentinization processes in the region. In addition, the time-migrated MCS seismic section provides a detailed view of the tectonostratigraphic framework of the COT transition.

Driven by discovery of contrasting structures of Continent to Ocean Transition (COT) discovered at rifted continental margins during the 90’s, several high-quality seismic datasets were acquired in these margins during the early 2000 to unravel the structure of unexplored regions. Despite the fact that some of these datasets are basically comparable to modern data in quality, the processing, imaging and modelling methodologies at the time of acquisition can be now refined and improved. Recent developments in parallel computing and novel geophysical approaches provide now the means to obtain a new look at the structure with enhanced resolution seismic models and a mathematically-robust analysis of the data uncertainty, that was formerly difficult, if not unfeasible, to achieve. We focused on the Newfoundland margin and applied up-to-date methodologies to the high-quality SCREECH dataset (2000). These data include three primary transects with coincident multichannel seismic (MCS) reflection data acquired with a 6-km streamer and wide-angle data recorded by short-period OBS and OBH spaced at ~15 km. We reprocessed the streamer data and also performed the join inversion of streamer and wide-angle OBS/OBH seismic data, using reflections and refraction arrivals, which significantly improved the resolution of the velocity model. We performed a statistical uncertainty analysis of the resulting model, supporting the reliability of the observed features. In particular the new velocity model provides a detailed definition of the top of the basement where the largest abrupt velocity change occurs. The comparatively high-resolution velocity model obtained from the joint tomography allowed to properly perform a Pre-Stack Depth Migration of the MCS. The improved velocity model and seismic images permit to characterize the different crustal domains of the margin with less uncertainty that previous attempts, and relate them to the tectonic structure.The different domains reveal previously undetected crustal characteristics that change over short distances. The reprocessing of the MCS data allowed to a better understanding of the crustal structure, as the Moho is imaged for the first time under the slope domain.Comparison of these new results on the Newfoundland margin with the most modern data on the West Iberian margin, acquired during FRAME (2018) and ATLANTIS (2022) cruises provides a new view of the evolution of the North Atlantic opening.

The present-day Eurasia-Africa plate convergence offshore SW Iberia gives rise to a diffuse plate boundary marked by deep lithospheric thrust and strike-slip faults. The Horseshoe Abyssal Plain Thrust (HAT) stands out as a key structure accommodating plate convergence, and it has been the site of deep (> 30 km depth) and large magnitude (Mw > 6) earthquakes. Additionally, the HAT has been proposed to be the source of the 1755 Lisbon earthquake (estimated Mw≥8.5), one of the most destructive earthquakes and tsunami in the history of Europe. The geometry of the fault and the physical properties of rocks surrounding it have been determined through tomographic models derived from controlled-source seismic data. Although large earthquakes along the HAT primarily occur at considerable depths within the peridotitic mantle (~40 km depth), the fault intersects a region of serpentinized mantle at shallower depths (10-20 km depth). In contrast to peridotite that undergoes seismic deformation, the frictional behaviour of serpentinized peridotite depends on factors such as pressure, water content, temperature, and slip velocity. Laboratory measurements indicate that serpentinite transitions from rate-strengthening behaviour at plate tectonic rates to rate-weakening at seismic slip rates. This dual nature suggests that large deep earthquakes, nucleated in pristine peridotite, could rupture seismically through the weaker serpentinized peridotite. While this mechanism has been proposed to explain the HAT's potential to generate large tsunamigenic earthquakes, it remains untested. In this study, we use dynamic rupture numerical simulations to investigate the role of serpentinized peridotite in the rupture process and the tsunamigenic potential of the HAT. In particular, we explore various frictional scenarios to determine the slip pattern necessary to account for the previously estimated tsunamigenic uplift associated with the 1755 Lisbon earthquake.

Seismic methods are one of the most powerful existing geophysical tools to extract information on the structure of the Earth’s subsurface. These techniques continue to be widely used to obtain images of the sediments and crust and to map the variations in physical properties. Particularly the P-wave velocity (Vp) distribution is the most commonly modelled property. The most widely applied seismic method to retrieve velocity models is seismic tomography, using either travel-time information as in travel-time tomography (TTT) or a more complete set of waveform attributes in full-waveform inversion (FWI). Whereas TTT is a robust, moderately non-linear technique providing coarse models, FWI is strongly non-linear and computationally demanding, but with the potential to provide higher resolution models. TTT and FWI are considered to be complementary, so that they are often combined and applied together. TTT is applied first to get a moderate-resolution model, which is then used as an initial model for FWI. In fact, the key to successfully apply FWI to seismic data, is the usage of a kinematically correct initial model, in which simulated and recorded waveforms are not cycle-skipped at the lowest frequency available. On this basis, it is crucial to extract the travel-time information of the refracted waves as accurate as possible. Particularly for marine multichannel reflection seismic (MCS) acquisition systems, where most refractions are masked by reflections and noise, data processing techniques like Downward Continuation (DC) allows to better retrieve refractions, and this is achieved by virtually redatuming streamer field data to the seafloor. We aim at showing the different user-friendly open source HPC software designed and built in the Barcelona Center for Subsurface Image (BCSI) to process seismic data from marine experiments, recorded by ocean bottom seismometers (OBS) and/or multichannel seismic (MCS) data recorded by towed streamers. The three tools, DC, TTT, and FWI, which can be used independently and/or combined together for a better performance, have the potential to produce high resolution models of the physical properties of the subsurface. While DC software is designed for 2D seismic data, the TTT tool also read 3D data, and allows the joint inversion of Vp, Vs and anisotropic properties for active and passive data, including earthquake relocation. At present, DC and TTT software tools are freely available and the last version updated at GitHub repositories and the FWI tool for 2D data is under development.