Argo profiling floats recorded temperature increases and salinity decreases at roughly 1.75–2 km depth, below the level of the eruptive vent, during the July 2012 Havre submarine caldera eruption. Mittal and Delbridge (2019) interpreted these anomalies, significant at more than 3σ against background ocean variability, as the signature of large-scale ocean mixing driven by the eruption plume rather than direct heating, using scaling analysis and a modified one-dimensional integral plume–gravity-current model. The open problem is the quantification of that mechanism and its consequences: how full three-dimensional turbulent dynamics, particle–flow coupling, and transient eruption variability modify the intrusion depths, detrainment multiplicity, and mixing efficiency predicted by parameterized 1D integral models; how sensitive these predictions are to poorly constrained source parameters across eruption styles and magnitudes; and whether eruption-induced mixing, once embedded in coupled climate–ocean models, has significant long-term climatic and biogeochemical effects up to the scale of submarine Large Igneous Provinces.

Background

More than 85% of Earth's volcanism occurs underwater, at mid-ocean ridges, intraplate hotspots, and oceanic island arcs, yet the interaction between a submarine eruption and the overlying stratified water column is barely observed. The 2012 eruption of Havre volcano (Kermadec arc, southwest Pacific) — the largest deep-ocean silicic (silica-rich) eruption of the past century (Carey et al., 2018) — produced a giant floating pumice raft (a raft of low-density frothy volcanic glass) and, unnoticed at the time, a measurable perturbation of the deep ocean. Mittal and Delbridge (2019) showed that Argo floats, a global array of autonomous instruments that profile temperature and salinity in the upper 2 km of the ocean roughly every ten days, had fortuitously recorded temperature increases and salinity decreases at depths of about 1.75–2 km — below the depth of the eruptive vent itself — significant at more than 3 standard deviations relative to background ocean variability, persisting out to hundreds of kilometers from the volcano.

To explain these deep anomalies, the authors combined scaling analysis with a modified one-dimensional integral plume–gravity-current model. An integral plume model is a cross-section-averaged description of a buoyant jet: conservation equations for mass, momentum, buoyancy, and transported phases are integrated over the plume cross-section, with entrainment of ambient seawater parameterized by an entrainment coefficient; models of this class (for example the TAMOC framework — the Texas A&M Oilspill Calculator — of Dissanayake et al., 2018, developed for oil-spill and gas-blowout multiphase plumes) treat the dispersed phases — bubbles, droplets, and particles such as volcanic ash — as Lagrangian or Eulerian components coupled to the mean flow. In a density-stratified ocean, a rising multiphase plume entrains heavy ambient water, so the mixed plume fluid becomes progressively denser; where its density matches that of the surroundings (the level of neutral buoyancy), plume fluid detaches, or detrains, and spreads laterally as a subsurface intrusion. Because particles continue to settle out of the detrained fluid, the residual plume can re-ascend and detrain again at a shallower level, producing the multiple stacked intrusions the model predicted for Havre. Mittal and Delbridge argued that this process efficiently advects deep water masses toward the surface while injecting mixed water below the vent level, i.e., that submarine eruption plumes are a mechanism of basin-scale ocean mixing, with implications for ocean chemistry, dynamics, and ecology, including during the emplacement of submarine Large Igneous Provinces (LIPs) — episodic massive flood-basalt eruptions on the ocean floor that coincide in the geological record with oceanic anoxic events and mass extinctions.

The interpretation rests on strongly idealized physics: a 1D integral model parameterizes rather than resolves turbulence, treats particle–flow coupling and microphysical processes (bubble and particle size evolution, dissolution, breakup) simplistically, and assumes a steady, sustained source. Related work has advanced pieces of the underlying fluid mechanics — laboratory and integral studies of sedimentation from particle-bearing plumes in stratification (Sutherland and Hong, 2016; Mingotti and Woods, 2022), of fountain collapse and sedimentation-driven secondary plumes in deep-submarine explosive eruptions (Newland et al., 2022), and three-dimensional large-eddy simulation of bubble plumes in stratified crossflow as a methodology (Xiao et al., 2021; large-eddy simulation computes the large, energy-containing turbulent eddies directly and parameterizes only the smallest scales) — but no study quantifies what full three-dimensional turbulent dynamics, two-way particle–flow coupling, and transient eruption variability do to the intrusion depths, detrainment multiplicity, and mixing efficiency of a submarine eruption plume, nor whether the resulting mixing matters for climate and biogeochemistry when carried to LIP scale in coupled models. That gap is the open problem here.

Problem Statement

The overarching objective is to determine, quantitatively, how much and in what structure submarine eruption plumes mix the ocean, and whether that mixing matters for ocean dynamics, climate, and biogeochemistry. Concretely, for submarine explosive eruptions of the Havre 2012 type (and spanning other magnitudes, styles, and settings such as mid-ocean ridges versus arcs):

  1. Quantify how full three-dimensional turbulent dynamics, two-way coupling between the flow and the dispersed phases (bubbles, droplets, volcanic particles of distributed sizes), and transient eruption variability modify — relative to the predictions of parameterized one-dimensional integral plume–gravity-current models — (a) the depths of subsurface intrusions formed by the plume, (b) the number and vertical spacing of detrainment levels (the multiplicity of stacked intrusions), and (c) the bulk mixing efficiency, i.e., how much ambient water the plume transports vertically; together with the sensitivity of these quantities to the poorly constrained source parameters (dispersed-phase mass fractions, particle-size distributions, initial vent flow speed, vent geometry), so that expected temperature/salinity anomaly magnitudes and depth structure can be bounded for different eruption styles and magnitudes, with retrospective consistency against the published Havre Argo anomalies as the observational constraint.

  2. Establish whether eruption-induced enhanced mixing, represented through plume-derived mixing parameterizations in coupled ocean–climate–biogeochemical models, produces significant long-term climatic and biogeochemical effects — up to and including the scales of submarine Large Igneous Province emplacement — or is negligible against background ocean mixing.

These two components are linked as the plume-scale and global-scale halves of one question: component 1 supplies the physically validated mixing parameterization that component 2 requires, and component 2 decides its scientific significance. The sparse, non-synoptic nature of the Havre Argo sampling is a limitation, but the published anomalies remain the observational validation constraint on the plume-scale quantification. A statistical characterization of Argo (and its deep-profiling and biogeochemical-sensor extensions, Deep-Argo and BGC-Argo) eruption-detection rates is a related but distinct observational-monitoring question and is not part of this problem.

References

  1. Tushar Mittal and Brent Delbridge (2019), 'Detection of the 2012 Havre submarine eruption plume using Argo floats and its implications for ocean dynamics', Earth and Planetary Science Letters 511, 105–116. DOI: 10.1016/j.epsl.2019.01.035, https://doi.org/10.1016/j.epsl.2019.01.035
  2. Rebecca Carey, S. Adam Soule, Michael Manga, James D. L. White, Jocelyn McPhie, Richard Wysoczanski, Martin Jutzeler, Kenichiro Tani, Dana Yoerger, Daniel Fornari, Fabio Caratori-Tontini, Bruce Houghton, Samuel Mitchell, Fumihiko Ikegami, Chris Conway, Arran Murch, Kristen Fauria, Max P. Jones, Ryan Cahalan, and Warren McKenzie (2018), 'The largest deep-ocean silicic volcanic eruption of the past century', Science Advances 4, e1701121. DOI: 10.1126/sciadv.1701121, https://doi.org/10.1126/sciadv.1701121
  3. Anusha L. Dissanayake, Jonas Gros, and Scott A. Socolofsky (2018), 'Integral models for bubble, droplet, and multiphase plume dynamics in stratification and crossflow', Environmental Fluid Mechanics 18, 1167–1202. DOI: 10.1007/s10652-018-9591-y, https://doi.org/10.1007/s10652-018-9591-y
  4. Eric L. Newland, Nicola Mingotti, and Andrew W. Woods (2022), 'Dynamics of deep-submarine volcanic eruptions', Scientific Reports 12, 3276. DOI: 10.1038/s41598-022-07351-9, https://doi.org/10.1038/s41598-022-07351-9
  5. Nicola Mingotti and Andrew W. Woods (2022), 'Dynamics of sediment-laden plumes in the ocean', Flow 2, E26. DOI: 10.1017/flo.2022.12, https://doi.org/10.1017/flo.2022.12
  6. Bruce R. Sutherland and Youn Sub (Dominic) Hong (2016), 'Sedimentation from particle-bearing plumes in a stratified ambient', Physical Review Fluids 1, 074302. DOI: 10.1103/physrevfluids.1.074302, https://doi.org/10.1103/physrevfluids.1.074302
  7. Shuolin Xiao, Chen Peng, and Di Yang (2021), 'Large-eddy simulation of bubble plume in stratified crossflow', Physical Review Fluids 6, 044613. DOI: 10.1103/physrevfluids.6.044613, https://doi.org/10.1103/physrevfluids.6.044613
  8. Johan T. Gilchrist and A. Mark Jellinek (2021), 'Sediment waves and the gravitational stability of volcanic jets', Bulletin of Volcanology 83, 64. DOI: 10.1007/s00445-021-01472-1, https://doi.org/10.1007/s00445-021-01472-1
  9. Arran P. Murch, James D. L. White, Thibaut Barreyre, Rebecca J. Carey, Rhiannon Mundana, and Fumihiko Ikegami (2020), 'Volcaniclastic Dispersal During Submarine Lava Effusion: The 2012 Eruption of Havre Volcano, Kermadec Arc, New Zealand', Frontiers in Earth Science 8. DOI: 10.3389/feart.2020.00237, https://doi.org/10.3389/feart.2020.00237
  10. Maggie Zheng, Tushar Mittal, Kristen E. Fauria, Ajit Subramaniam, and Martin Jutzeler (2022), 'Pumice Raft Detection Using Machine-Learning on Multispectral Satellite Imagery', Frontiers in Earth Science 10. DOI: 10.3389/feart.2022.838532, https://doi.org/10.3389/feart.2022.838532
  11. K. M. Bisson, S. Gassó, N. Mahowald, S. Wagner, B. Koffman, S. A. Carn, S. Deutsch, E. Gazel, S. Kramer, N. Krotkov, C. Mitchell, M. E. Pritchard, K. Stamieszkin, and C. Wilson (2023), 'Observing ocean ecosystem responses to volcanic ash', Remote Sensing of Environment 296, 113749. DOI: 10.1016/j.rse.2023.113749, https://doi.org/10.1016/j.rse.2023.113749
  12. Rebecca J. Carey, David A. Butterfield, and Malcolm R. Clark (2025), 'Submarine Volcanic Eruptions and Their Impacts on Hydrothermal Systems and Biological Communities', Annual Review of Earth and Planetary Sciences 53, 397–423. DOI: 10.1146/annurev-earth-040522-095654, https://doi.org/10.1146/annurev-earth-040522-095654
  13. Casey M. S. Schine, Jens-Erik Lund Snee, Alex Lyford, Gert van Dijken, and Kevin R. Arrigo (2025), 'Southern Ocean net primary production influenced by seismically modulated hydrothermal iron', Nature Geoscience 19, 106–112. DOI: 10.1038/s41561-025-01862-6, https://doi.org/10.1038/s41561-025-01862-6
  14. Biswajit Basu and Manish Kanojia (2023), 'Modelling the effect of submarine volcanic eruption on equatorial oceanic flows', Nonlinear Analysis: Real World Applications 74, 103956. DOI: 10.1016/j.nonrwa.2023.103956, https://doi.org/10.1016/j.nonrwa.2023.103956