Classical mantle plume theory holds that hotspot volcanism is fed by narrow, hot upwellings rising from the deep mantle, and its strongest observational test is seismic imaging of plume conduits beneath hotspots. Synthetic experiments by Maguire et al. (2018) showed that teleseismic traveltime tomography recovers purely thermal lower-mantle plume tails only under wide-aperture, densely sampled networks, and that plume-induced delays (typically under 1 s) are easily concealed by measurement noise or mimicked by inversion artifacts such as vertical smearing (the stretching of a shallow anomaly into an apparently deep conduit by steep ray paths). Those tests, however, used only thermally generated plume models in a pyrolitic mantle, leaving open the scientifically central cases: plumes whose buoyancy, width, and velocity contrast are controlled by composition (thermochemical plumes with eclogitic or other non-peridotitic entrainment), plumes embedded in whole-mantle convection models, and the question of whether full-waveform or multiple-frequency methods can overcome the sparse and uneven data coverage that afflicts real, mostly oceanic, hotspot settings. The open problem is to determine, for compositionally and dynamically realistic plume scenarios, which seismic signatures genuinely indicate a deep-mantle plume rather than an imaging artifact, and whether denser sampling or more complete waveform-based inversion can raise detection capability above the limits established for traveltime tomography.

Background

Mantle plumes are hypothesized narrow columns of hot rock rising from the deep mantle that are held to feed long-lived volcanic hotspots such as Hawaii, Iceland, and the South Pacific hotspots; verifying that such conduits exist and reach the lower mantle is a central unresolved test of the plume hypothesis. The primary seismological tool is teleseismic traveltime tomography, which reconstructs seismic wave-speed anomalies in the mantle from small delays in the arrival times of body waves (P and S waves) generated by distant earthquakes and recorded by networks of seismic stations. Because plume tails in the lower mantle are narrow (hundreds of kilometres wide) and only mildly slow, several physical effects conspire against their detection: wavefront healing, the progressive recovery of a diffracted wavefront so that a delayed arrival loses its delay as it propagates past a small anomaly, was shown by Hwang, Ritsema, van Keken, Goes and Styles (2011) to render deep plumes effectively invisible at long periods; Maguire, Ritsema, van Keken, Fichtner and Goes (2016) quantified the resulting P- and S-wave traveltime delays using full numerical wave simulations. In addition, the inversion of traveltimes is an underdetermined, damped least-squares problem, so the imaging operator itself can create plume-like structures, notably vertical smearing, in which a shallow low-velocity anomaly is stretched downward into an apparently deep-rooted conduit by the steep, poorly crisscrossed ray paths beneath a regional array.

Maguire, Ritsema, Bonnin, van Keken and Goes (2018) performed the most systematic controlled assessment to date: they generated plume structures from axisymmetric compressible thermal convection simulations of a uniform pyrolitic composition, converted temperature to seismic velocities with equilibrium mineral-physics calculations (Perple_X, a widely used thermodynamic software package) and an anelastic correction (accounting for the way seismic attenuation reduces wave speeds, an effect that strengthens at long periods), propagated teleseismic P, S, and SKS waves through these models (SKS is a shear phase that traverses the Earth's outer core as a compressional wave and converts back to S, so it crosses the mantle steeply and nearly vertically) with the spectral-element method (a numerical technique that solves the wave equation on a 3-D grid, capturing diffraction and scattering), and inverted the synthetic traveltimes with both ray-theoretical and finite-frequency formulations (the finite-frequency approach uses sensitivity kernels describing traveltime dependence on wave speed in a volume around the ray, rather than only along the ray). Their conclusions: narrow (<500 km wide) thermal plume tails below roughly 1,600–2,000 km depth require wide-aperture (4,000–6,000 km) networks with dense station spacing (<100–200 km) to be imaged, and then only with large amplitude loss; if traveltime measurement uncertainty (typically 0.1–1 s in practice) exceeds the plume-induced delay (typically <1 s), the tail disappears from the image; and joint S+SKS inversion of a realistic deployment geometry (the PLUME temporary broadband seismic deployment around French Polynesia) smears a purely upper-mantle anomaly into an apparent lower-mantle plume. These results calibrate how to interpret, and how to design, plume-imaging experiments.

Crucially, all of these tests used purely thermal plumes in a mantle of uniform composition, whereas geochemical and geodynamic evidence summarized by Koppers, Becker, Jackson, Konrad, Müller, Romanowicz, Steinberger and Whittaker (2021) indicates that many plumes are thermochemical, meaning their buoyancy and dynamics involve density contrasts from compositional differences, not temperature alone. Composition can weaken, cancel, or even reverse the seismic velocity anomaly of a plume conduit (dense eclogitic material entrained from recycled oceanic crust is the canonical example), so detectability conclusions derived from purely thermal models do not transfer directly: although plume-tail delay times are approximately linear in velocity perturbation, so that the thermal results do offer partial insight into the resolvability of stronger or weaker thermochemical tails of similar width, entrainment of dense material can also reduce plume buoyancy and broaden plumes, and non-peridotitic components can change the amplitude of the seismic anomaly (Maguire et al. 2018). Earlier synthetic-tomography work on thermochemical structures by Bull, McNamara and Ritsema (2009) addressed global-scale plume clusters and thermochemical piles (dense, compositionally distinct domains resting on the core–mantle boundary), not the regional teleseismic resolution of individual plume tails. Whole-mantle waveform tomography, such as the model of French and Romanowicz (2015) that imaged broad low-velocity conduits rooted at the core–mantle boundary beneath major hotspots, motivated the very resolution debate that Maguire et al. (2018) sharpened; whether such waveform-based methods, or multiple-frequency tomography (exploiting the frequency dependence of wavefront healing by inverting traveltimes measured in several frequency bands), can overcome the sparse data coverage that dominates real oceanic hotspot settings remains undetermined. Subsequent work has only circumscribed the question: Stockmann, Cobden, Deschamps, Fichtner and Thomas (2019) tested the visibility of dynamically plausible thermochemical plume models with seismic array (beam-forming/scattering) methods rather than traveltime inversion; Civiero, Armitage, Goes and Hammond (2019) converted upper-mantle plumelet simulations to seismic structure for the East African Rift; Marignier, Ferreira and Kitching (2020) quantified the probability that plume-like features in global tomography models are statistically distinct from noise; Simon, Simons and Irving (2021) began deploying autonomous MERMAID hydrophone floats (Mobile Earthquake Recording in Marine Areas by Independent Divers) in the South Pacific precisely to fill the oceanic data gap; Gülcher, Golabek, Thielmann, Ballmer and Tackley (2022) showed that strain-weakening lower-mantle rheology yields even narrower, cooler plume conduits than classical thermal convection, making seismic detection harder still; and Wamba, Simons and Irving (2025) demonstrated that state-of-the-art global full-waveform inversion models remain comparatively poorly resolved in the South Pacific superswell region because of uneven receiver coverage. Reviews by Ritsema, Maguire, Cobden and Goes (2021), Schuberth and Bigalke (2021), Koppers et al. (2021), and Montagner, Romanowicz, Wamba and Burgos (2025) all treat the seismic detection problem for deep plumes as current and unresolved.

Problem Statement

The problem, as left open by Maguire et al. (2018) and taken up by the subsequent literature, is to determine what seismic observations genuinely diagnose a deep mantle plume once the simplification of purely thermal, uniform-composition plume models is dropped. Concretely, for geodynamically derived plume scenarios that are thermochemically realistic — including compositional buoyancy, dense eclogitic or other non-peridotitic entrainment, and plumes embedded in whole-mantle convection models — establish, by controlled synthetic experiments with realistic wave propagation, measurement noise, and source–receiver geometries: (i) the traveltime and waveform (including frequency-dependent) signatures such plumes impart to teleseismic P, S, and SKS waves, and how they differ from the purely thermal case in which delays are typically under 1 s; (ii) quantitative criteria by which a genuine deep plume conduit can be distinguished from the known imaging artifacts of regional teleseismic tomography (vertical smearing of upper-mantle anomalies, damping-induced side-lobes) and from statistical noise in tomographic models; and (iii) whether more complete waveform-based approaches — multiple-frequency tomography, full-waveform inversion — and denser, ocean-capable sampling can overcome the sparse and uneven data coverage that presently limits detection at real, largely oceanic hotspot settings. These three components jointly serve one scientific objective: a reliable seismic test of whether narrow, deep-rooted mantle plumes exist beneath hotspots. They correspond to the future directions separately identified in the Conclusions and Discussion of Maguire et al. (2018).

References

  1. Maguire, Ross; Ritsema, Jeroen; Bonnin, Mickaël; van Keken, Peter E.; Goes, Saskia (2018). Evaluating the Resolution of Deep Mantle Plumes in Teleseismic Traveltime Tomography. Journal of Geophysical Research: Solid Earth 123(1), 384–400. doi:10.1002/2017jb014730. https://doi.org/10.1002/2017jb014730
  2. Hwang, Yong Keun; Ritsema, Jeroen; van Keken, Peter E.; Goes, Saskia; Styles, Elinor (2011). Wavefront healing renders deep plumes seismically invisible. Geophysical Journal International 187(1), 273–277. doi:10.1111/j.1365-246x.2011.05173.x. https://doi.org/10.1111/j.1365-246x.2011.05173.x
  3. Maguire, Ross; Ritsema, Jeroen; van Keken, Peter E.; Fichtner, Andreas; Goes, Saskia (2016). P- and S-wave delays caused by thermal plumes. Geophysical Journal International 206(2), 1169–1178. doi:10.1093/gji/ggw187. https://doi.org/10.1093/gji/ggw187
  4. Bull, Alison L.; McNamara, Allen K.; Ritsema, Jeroen (2009). Synthetic tomography of plume clusters and thermochemical piles. Earth and Planetary Science Letters 278(3–4), 152–162. doi:10.1016/j.epsl.2008.11.018. https://doi.org/10.1016/j.epsl.2008.11.018
  5. French, Scott W.; Romanowicz, Barbara (2015). Broad plumes rooted at the base of the Earth's mantle beneath major hotspots. Nature 525(7567), 95–99. doi:10.1038/nature14876. https://doi.org/10.1038/nature14876
  6. Stockmann, Fabienne; Cobden, Laura; Deschamps, Frédéric; Fichtner, Andreas; Thomas, Christine (2019). Investigating the seismic structure and visibility of dynamic plume models with seismic array methods. Geophysical Journal International 219(Supplement_1), S167–S194. doi:10.1093/gji/ggz334. https://doi.org/10.1093/gji/ggz334
  7. Civiero, Chiara; Armitage, John J.; Goes, Saskia; Hammond, James O. S. (2019). The Seismic Signature of Upper-Mantle Plumes: Application to the Northern East African Rift. Geochemistry, Geophysics, Geosystems 20(12), 6106–6122. doi:10.1029/2019GC008636. https://doi.org/10.1029/2019GC008636
  8. Marignier, Augustin; Ferreira, Ana M. G.; Kitching, Thomas (2020). The Probability of Mantle Plumes in Global Tomographic Models. Geochemistry, Geophysics, Geosystems 21(9), e2020GC009276. doi:10.1029/2020GC009276. https://doi.org/10.1029/2020GC009276
  9. Ritsema, Jeroen; Maguire, Ross; Cobden, Laura; Goes, Saskia (2021). Seismic Imaging of Deep Mantle Plumes. In: Mantle Convection and Surface Expressions, Geophysical Monograph Series, 353–369. doi:10.1002/9781119528609.ch14. https://doi.org/10.1002/9781119528609.ch14
  10. Schuberth, Bernhard S. A.; Bigalke, Tobias (2021). From Mantle Convection to Seismic Observations. In: Mantle Convection and Surface Expressions, Geophysical Monograph Series, 97–119. doi:10.1002/9781119528609.ch4. https://doi.org/10.1002/9781119528609.ch4
  11. Koppers, Anthony A. P.; Becker, Thorsten W.; Jackson, Matthew G.; Konrad, Kevin; Müller, R. Dietmar; Romanowicz, Barbara; Steinberger, Bernhard; Whittaker, Joanne M. (2021). Mantle plumes and their role in Earth processes. Nature Reviews Earth & Environment 2(6), 382–401. doi:10.1038/s43017-021-00168-6. https://doi.org/10.1038/s43017-021-00168-6
  12. Simon, Joel D.; Simons, Frederik J.; Irving, Jessica C. E. (2021). Recording earthquakes for tomographic imaging of the mantle beneath the South Pacific by autonomous MERMAID floats. Geophysical Journal International 228(1), 147–170. doi:10.1093/gji/ggab271. https://doi.org/10.1093/gji/ggab271
  13. Gülcher, A. J. P.; Golabek, G. J.; Thielmann, M.; Ballmer, M. D.; Tackley, P. J. (2022). Narrow, Fast, and "Cool" Mantle Plumes Caused by Strain-Weakening Rheology in Earth's Lower Mantle. Geochemistry, Geophysics, Geosystems 23(10), e2021GC010314. doi:10.1029/2021GC010314. https://doi.org/10.1029/2021GC010314
  14. Faccenda, Manuele; VanderBeek, Brandon P. (2023). On constraining 3D seismic anisotropy in subduction, mid-ocean-ridge, and plume environments with teleseismic body wave data. Journal of Geodynamics 158, 102003. doi:10.1016/j.jog.2023.102003. https://doi.org/10.1016/j.jog.2023.102003
  15. Wamba, Mathurin D.; Simons, Frederik J.; Irving, Jessica C. E. (2025). Data-space cross-validation of mantle structure in global tomographic models underneath the Pacific Ocean. Geophysical Journal International 241(1), 241–259. doi:10.1093/gji/ggaf044. https://doi.org/10.1093/gji/ggaf044
  16. Montagner, Jean-Paul; Romanowicz, Barbara; Wamba, Mathurin; Burgos, Gael (2025). Upwellings and Mantle Ponding Zones in the Lower Mantle Transition Zone (660–1000 km). Geosciences 15(11), 413. doi:10.3390/geosciences15110413. https://doi.org/10.3390/geosciences15110413