Jason Phipps Morgan

This article describes an ocean-bottom seismic data set collected offshore New Zealand’s North Island as part of a multinational amphibious experiment investigating the Auckland volcanic field. From August 2023 to September 2024, 20 broadband ocean-bottom seismographs were deployed at water depths of 1200–2700 m, yielding approximately one year of continuous recordings from 15 successfully recovered instruments. We present comprehensive quality metrics including instrument orientations (determined via Rayleigh-wave polarization with uncertainties < 7.5°), tilt characteristics (predominantly < 2.5°), and clock drift corrections (achieved to 0.1 s precision using ambient noise cross correlation for five units lacking Global Positioning System synchronization). The data set exhibits excellent signal fidelity in the microseismic band (2–20 s), with vertical components showing substantially lower noise than horizontal components. Notably, two instruments (R12 and R42) had a delayed recovery, and recorded 16–18 months of continuous data. All data will be publicly archived at the U.S. National Geophysical Facility Data Repository following a three-year proprietary period, providing critical offshore coverage to complement New Zealand’s land-based seismic network for studies of intraplate volcanism and subduction zone dynamics.

How submerged oceanic domains trapped within growing continents become stable, subaerial continental interiors and what that conversion means for the long-term carbon cycle remain poorly understood. Traditional models of such ocean-to-continent conversion emphasize significant horizontal shortening or plume activity. Here we identify an alternative, surface-process-mediated pathway recorded in Central Asia, where formerly submerged cratonic blocks became intracontinental domains in the late Paleozoic. Geodynamic simulations of subduction, whose predicted uplift is used to drive landscape-evolution simulations of erosion and deposition, show that uplift of the surrounding orogenic belts can generate strong topographic gradients that drive intense erosion, sediment transport, and basin stabilization. This tectonic configuration leads to enclosed intracontinental basins with high-sedimentation rates, promoting efficient organic carbon burial and long-term sequestration. Although oxidative weathering of rock organic carbon and metamorphic degassing partly offset burial, thick Devonian–Carboniferous carbonate sequences indicate substantial net carbon sequestration. This coupling between deep Earth dynamics, surface erosion, and carbon burial amplified the late Paleozoic CO ₂ decline and global cooling during Pangaea’s assembly. A similar process may have operated during the Cenozoic, when Tethys closure and Tibetan Plateau uplift enhanced erosion and carbon storage in the Mediterranean, Black, and Caspian basins, contributing to long-term cooling.

Subducting slabs play a crucial role in deep Earth carbon cycling, yet the carbon-cycling efficiency of different subduction zones remains debated. Here, we systematically compare decarbonation characteristics in the extremely cold Tonga and hot Colombia subduction zones using observation-constrained thermo-petrological modeling, integrated with volcanic degassing and arc geochemistry. Two contrasting slab decarbonation patterns are identified: efficient “focused” decarbonation in hot subduction zones versus sluggish “diffuse” decarbonation in cold subduction zones. Hot Colombia exhibits a high decarbonation efficiency >15%, dominated by metamorphic carbon release within a narrow depth range of ∼75–115 km and supplemented by partial melting-induced decarbonation. In contrast, cold Tonga shows a much lower decarbonation efficiency of ∼1% by dissolution-dominated carbon removal over a broad depth range of ∼75–310 km. These findings suggest a contrasting efficiency of different subducting slabs for deep carbon cycling and volcanic degassing in present-day and ancient subduction systems.

Several types of magmatism are typically associated with continental stretching and rift formation. The South Atlantic Rifted Margin is a particularly well-studied system that exhibits: (1) Thousands-km-long massive dike swarms – likely linked to the Tristan Plume/hotspot; (2) the Parana-Etendeka continental flood basalt (CFB); and (3) the formation of extensive seaward dipping reflector sequences (SDRs) along the southern portion of this rifted margin. Here we review the distribution, timing, and volumes of these different modes of rift-related magmatism in relation to rift evolution.Great dike swarms formed prior to, during, and soon after the Parana-Etendeka flood basalt event at 136.5-135.5 Ma. Although comparable in spatial extent and volume to the well-known Proterozoic Mackenzie dike swarm that similarly extended from a continental flood basalt, summed dike volumes appear to only be ~10% (0.15e6 km^3) of the Parana CFB magmatism (~1.5e6) and ~2% of total magmatism (~6e6) associated with South Atlantic Rifting including SDR provinces.The defining characteristic of the CFB event is that it occurred very rapidly, which appears most consistent with a sudden lithospheric thinning event (e.g. lower lithospheric delamination) in the presence of hot plume material. A plume-head rising under thick continental lithosphere simply could not create this sudden burst of volcanic activity, thus an abrupt lithospheric thinning event appears needed to explain this melting anomaly. Note that there is seismic evidence consistent with such a delamination event both in the thinned lower lithosphere beneath Parana and the presence of a delaminated lithospheric fragment in the transition zone near the site of the modern Tristan Plume.Finally, the largest volcanism associated with South Atlantic rifting is linked to the SDR province including associated underplated magmas offshore the southern margins of South American and Africa. This post-CFB magmatic activity can be quantitatively explained by more extensive melting of southward flowing Tristan Plume material after extensive rifting has thinned the extending lithosphere to

It has long been supposed that Earth’s asthenosphere contains small amounts of seismically visible melt; how and why this melt persists has remained a similarly long-supposed mystery. Here we show how this observation is a simple consequence of the preferential diffusion of hydrogen (‘water’) from a harder-to-melt peridotite lithology forming ~80% of the mantle into an easier-to-melt pyroxenite lithology that exists as ~m-10km blobs within a peridotitic ‘matrix’. Pyroxenites, due to their higher Al content, will have higher trace water contents when in diffusive equilibrium with neighboring peridotite. Their higher water contents, in turn, will tend to lower their solidi, and favor their partial melting over nearby peridotite sharing similar p-T conditions. In addition, the latent heat consumed during early pyroxenite melting can locally cool this mantle, favoring the inward diffusion of both heat (~1e-6 m^2/s) and hydrogen from surrounding peridotites.Here we use 2-D numerical models of flow and melting in upwelling mantle that include the possibility of both heat and hydrogen diffusion between nearby peridotite and pyroxenite lithologies, assuming experimentally measured hydrogen diffusivities of ~1e-7 – 1e-8 m^2/s. Several interesting effects are found. ‘Thin’ (~1-100m) pyroxenite layers will rapidly suck both heat and water from nearby peridotite, so locally cooling and drying this peridotite before it starts to pressure-release melt –– while at the same time increasing its viscosity with respect to warmer and damper peridotite. At ~10-100mm/yr ascent rates, larger (~1-10km-scale) blobs of recycled pyroxenitic basalts will instead tend to melt as chemically isolated regions that more slowly suck heat and water from their surrounding peridotites.Finally, laterally moving regions of asthenosphere containing partially melting pyroxenitic blebs and blobs will continue to partially melt for ~10s of Ma due to inward water diffusion even as small-degree melts form and escape from this partially molten bilithologic asthenosphere. This provides a simple geodynamic mechanism for why Earth’s suboceanic asthenosphere appears to persistently contain small amounts of partial melt at depths shallower than ~150km, while also leading to the formation of small degree melts far from plumes, ridges, or subduction zones. We present and discuss numerical experiments that illustrate each of these effects.

Mantle plumes are commonly believed to generate uplift of their overlying lithosphere, while also impacting climate and life. However, sedimentary deposits show a component of subsidence during plume-lithosphere interaction, challenging traditional uplift-only models of plume-lithosphere interaction. We investigate whether plume-induced delamination can induce subsidence as well as uplift through computations of plume-lithosphere interaction. When a denser lithospheric keel delaminates, initial plume ascent will induce ~2 km of uplift, followed by a more rapid ~2 km of subsidence over a few millions of years as the dense lithospheric keel delaminates. Delamination, distinct from back-arc extension and slab-induced flow, may explain observed subsidence patterns during continental flood basalt activity, such as those seen in the Paraná Flood Basalt, Columbia River Flood Basalt, Deccan Traps and Congo Basin regions.

Hotspot swells have typically been associated with mantle plume upwelling, yet the origins of much broader but lower superswells remain unclear. This study first reveals that swell width increases while height decreases at younger seafloor ages; it is possible for a single hot spot swell size to approach that of a superswell. Compressible numerical models replicate the observed trend. Modeling results suggest that both hot spot swells and superswells can form from interactions between mantle plumes and oceanic plates. Beneath young seafloors, the lateral redistribution of restitic, hence buoyant and relatively low-viscosity, lower lithospheric mantle due to plume impingement explains the formation of superswell-like morphology. These models effectively demonstrate lithospheric lid effects on mantle melting and support a reassessment of conventional methods for estimating core heat flux.

Mantle plumes are a key phenomenon in geodynamics, connecting the deep Earth interior with surficial tectonic plates. Numerical simulations are essential for studying plume dynamics and their interactions with overlying lithosphere. While 3D models could better capture the geometric features of plumes, 2D simulations offer superior computational efficiency in large‐scale and high‐resolution scenarios, particularly when rheologically complex geological processes are involved. Thus, both 2D and 3D models have been widely applied in previous numerical studies. However, due to geometric effects, they may yield different results for the same parameters; we need to know how to build the proper scaling relationship between 2D and 3D models. Here, we systematically compare the 2D versus 3D mantle plume in two different regimes. In the first regime with only a plume head, the 2D plume should have a smaller diameter (65%–100% of the 3D value) and lower temperature (reduced by 0–60 K relative to the 3D case) to best match 3D model result. In the second regime with a continuous plume tail, a much smaller diameter (30%–45%) but counter‐intuitively higher temperature (increased by 20–100 K) are needed for the 2D model to best approximate 3D result. Further analytical studies indicate that such discrepancies are mainly controlled by the conservations of area (2D) versus volume (3D) of plume materials. These numerical and analytical results provide quantitative relationships between 2D and 3D plume models, which act as a theoretical reference for interpretation of previous models as well as guidance for future studies.

Volcanic feature identification on topographic maps is an emerging area that can leverage advanced machine learning techniques. However, the field faces challenges due to the scarcity of current labeled volcanic data, class imbalances, and limitations in existing models. Current approaches often fail to effectively capture the spatial continuity and hierarchical patterns inherent in volcanic terrains, in particular for intraplate volcanism such as petit-spot volcanoes.This study began with manually labeled data and employed Support Vector Machines as a preliminary classification tool, which revealed significant limitations in its handling of complex spatial patterns. Subsequently, efforts shifted to Convolutional Neural Networks (CNNs) with transfer learning to enhance feature extraction and classification. High-resolution topographic data from the Japanese Volcanic Islands and the Second Institute of Oceanography Ministry of Natural Resources of China were integrated. These datasets, enriched by grid-based labeling and data augmentation strategies, have provided the foundation for model training and validation. To date we have prioritized the identification of small volcanic features on Pacific-type (fast-spreading) seafloor. Work to date has emphasized the need for high-quality labeled datasets and innovative preprocessing techniques to have reliable machine recognition of volcanic bathymetric features. By incorporating high-precision data from Pacific expeditions, this research will also contribute to the development of future deep learning approaches, laying the groundwork for further advancements in automated volcano identification.