The 2014 Science report 'Dehydration melting at the top of the lower mantle' combined diamond-anvil-cell experiments, EarthScope seismic imaging, and mantle-flow models to propose that hydrated transition-zone material descending across the 660-km discontinuity undergoes dehydration melting, producing low-velocity zones at the top of the lower mantle beneath North America and helping trap water in the transition zone. Its discussion left three linked questions open: the water partition coefficients between minerals and melts at 660-km pressure-temperature conditions in the peridotite-saturated system; the density, buoyancy, and fate of hydrous melt near the top of the lower mantle; and whether the observed velocity reductions record partial melt rather than segregated basalt. Together these determine the paper's own open question — whether transition-zone water contents are sufficient to cause dehydration melting where there is downward flow into the lower mantle — and whether, as the paper proposed, dehydration melting acts to trap H2O in the transition zone. The problem is to resolve these linked questions and thereby establish, quantitatively, whether the transition-zone water filter operates.

Background

Earth's mantle is divided by seismic discontinuities at about 410 km and 660 km depth into the upper mantle, the mantle transition zone, and the lower mantle. The transition zone is dominated by the high-pressure forms of olivine — wadsleyite (roughly 410-520 km) and ringwoodite (roughly 520-660 km) — which can incorporate up to about 1-3 weight percent (wt %) H2_2O in their crystal structures, orders of magnitude more than the mantle above and below: the upper mantle, as sampled by mid-ocean-ridge basalts, holds about 0.005-0.02 wt % H2_2O, while the lower mantle consists mainly of bridgmanite (a Mg-Fe silicate perovskite) plus ferropericlase, (Mg,Fe)O, whose water capacity is debated and possibly very small. The 660-km discontinuity (the '660') marks the breakdown of ringwoodite to bridgmanite and ferropericlase at about 23.5 GPa. If material hydrated in the transition zone flows down across the 660, its sudden drop in water storage capacity can drive dehydration melting — partial melting triggered by water exclusion from the crystallizing minerals. Bercovici and Karato (2003) proposed that melts generated at both boundaries of the transition zone, if buoyant, percolate back into it, filtering water and incompatible elements (chemical elements that preferentially concentrate into melts rather than into minerals) and maintaining a hydrated transition zone — the 'transition-zone water filter' hypothesis. A hydrous ringwoodite inclusion containing about 1.5 wt % H2_2O, found in a natural diamond (Pearson et al. 2014), showed that local water saturation in the transition zone is possible. Schmandt et al. (2014) tested the hypothesis with three independent lines of evidence: (1) laser-heated diamond-anvil-cell experiments — a high-pressure apparatus that compresses a microscopic sample between two diamond anvils while lasers heat it to mantle temperatures — in which hydrous ringwoodite (about 1 wt % H2_2O) transformed to perovskite plus (Mg,Fe)O at 30 GPa and 1600 °C produced nanoscale intergranular silicate melt; (2) receiver-function imaging — mapping P-to-S (Ps) converted waves from teleseismic earthquakes into a common-conversion-point image — with the approximately 70-km-spaced EarthScope transportable seismic array across the United States, which revealed abrupt shear-velocity decreases of about 2.6% in the uppermost ~120 km of the lower mantle almost exclusively where mantle-circulation models predicted downwelling across the 660; and (3) melt-fraction estimates of 0.68-1 vol %, obtained by scaling the velocity drop with a published relation of 2.6-3.8% shear-velocity decrease per volume percent of melt. The quantitative interpretation of these observations, however, rested on quantities that were unknown in 2014, and the paper's discussion enumerated them as open questions. Two conventions are needed to read what follows: a mineral-melt partition coefficient is the ratio of the concentration of a component (here H2_2O) in a mineral to its concentration in the coexisting melt, and 'peridotite-saturated' refers to experiments on bulk peridotite (the representative mantle-rock composition, approximated by pyrolite) rather than on simplified chemical subsystems; the relevant conditions at the top of the lower mantle are about 23.5-26 GPa and roughly 1600-2000 K along mantle geotherms.

Problem Statement

Determine, quantitatively, whether dehydration melting operates where hydrated transition-zone material flows down across the 660-km discontinuity, and whether it makes the mantle transition zone a long-lived water filter. This requires resolving the three linked questions left open by Schmandt et al. (2014): (1) the water partition coefficients between minerals and melts at the pressure-temperature conditions of the 660-km discontinuity (about 23.5-26 GPa, 1600-2000 K) in the peridotite-saturated compositional system, which in 2014 had never been measured experimentally and which are needed to predict partial-melt percentages at 660 km depth for any given H2O content of the transition zone; (2) the density of hydrous melt near the top of the lower mantle — whether it is buoyant relative to the surrounding lower mantle, at what rate it percolates, and its fate, namely whether it ponds, rises into the transition zone, or freezes; and (3) whether the observed sub-660 seismic velocity reductions are caused by partial melt or by an alternative such as segregated, neutrally buoyant basalt that also reduces seismic velocities. The hypothesis under test is the paper's closing proposal that the combination of dehydration melting driven by downwelling across the 660 and upwelling across the 410 'could create a long-term H2O trap in the transition zone'; testing it does not additionally require mapping the global lateral extent of transition-zone hydration or closing the mantle water budget over Earth's history, which are outside the scope of this problem. The problem is complete when the mechanism is quantitatively established or refuted end-to-end: melt fractions predictable from the H2O content of the transition zone, melt buoyancy and fate near the top of the lower mantle determined, and the sub-660 seismic signal attributed against the segregated-basalt alternative. These three results decide the headline question in full, including whether the resulting water filter is 'long-lived': the filter's longevity is settled by the melt-fate item above — whether melt generated at the top of the lower mantle is buoyant and returns H2O to the transition zone rather than sinking — and requires no separate analysis of the filter's persistence over geological time or of the mantle water budget.

References

  1. Schmandt, B., Jacobsen, S. D., Becker, T. W., Liu, Z., & Dueker, K. G. (2014). Dehydration melting at the top of the lower mantle. Science, 344(6189), 1265-1268. https://doi.org/10.1126/science.1253358
  2. Bercovici, D., & Karato, S. (2003). Whole-mantle convection and the transition-zone water filter. Nature, 425, 39-44. https://doi.org/10.1038/nature01918
  3. Pearson, D. G., Brenker, F. E., Nestola, F., McNeill, J., Nasdala, L., Hutchison, M. T., Matveev, S., Mather, K., Silversmit, G., Schmitz, S., Vekemans, B., & Vincze, L. (2014). Hydrous mantle transition zone indicated by ringwoodite included within diamond. Nature, 507, 221-224. https://doi.org/10.1038/nature13080
  4. Schulze, K., Marquardt, H., Kawazoe, T., Boffa Ballaran, T., McCammon, C. A., Koch-Müller, M., Kurnosov, A., & Marquardt, K. (2018). Seismically invisible water in Earth's transition zone? Earth and Planetary Science Letters, 498, 9-16. https://doi.org/10.1016/J.EPSL.2018.06.021
  5. Nakajima, A., Sakamaki, T., Kawazoe, T., & Suzuki, A. (2019). Hydrous magnesium-rich magma genesis at the top of the lower mantle. Scientific Reports, 9, 7420. https://doi.org/10.1038/s41598-019-43949-2
  6. Fu, S., Yang, J., Karato, S., Vasiliev, A., Presniakov, M. Yu., Gavriliuk, A. G., Ivanova, A. G., Hauri, E. H., Okuchi, T., Purevjav, N., & Lin, J.-F. (2019). Water concentration in single-crystal (Al,Fe)-bearing bridgmanite grown from the hydrous melt: Implications for dehydration melting at the topmost lower mantle. Geophysical Research Letters, 46(17-18), 10346-10357. https://doi.org/10.1029/2019GL084630
  7. Gréaux, S., Irifune, T., Higo, Y., Tange, Y., Arimoto, T., Liu, Z., & Yamada, A. (2019). Sound velocity of CaSiO3 perovskite suggests the presence of basaltic crust in the Earth's lower mantle. Nature, 565, 218-221. https://doi.org/10.1038/s41586-018-0816-5
  8. Panero, W. R., Thomas, C., Myhill, R., Pigott, J. S., Raepsaet, C., & Bureau, H. (2020). Dehydration melting below the undersaturated transition zone. Geochemistry, Geophysics, Geosystems, 21(2), e2019GC008712. https://doi.org/10.1029/2019GC008712
  9. Yuan, L., Steinle-Neumann, G., & Suzuki, A. (2020). Structure and density of H2O-rich Mg2SiO4 melts at high pressure from ab initio simulations. Journal of Geophysical Research: Solid Earth, 125(10), e2020JB020365. https://doi.org/10.1029/2020JB020365
  10. Karato, S., Karki, B., & Park, J. (2020). Deep mantle melting, global water circulation and its implications for the stability of the ocean mass. Progress in Earth and Planetary Science, 7, 76. https://doi.org/10.1186/s40645-020-00379-3
  11. Fei, H. (2021). Water content of the dehydration melting layer in the topmost lower mantle. Geophysical Research Letters, 48(1), e2020GL090973. https://doi.org/10.1029/2020GL090973
  12. Liu, Z., Fei, H., Chen, L., McCammon, C., Wang, L., Liu, R., Wang, F., Liu, B., & Katsura, T. (2021). Bridgmanite is nearly dry at the top of the lower mantle. Earth and Planetary Science Letters, 570, 117088. https://doi.org/10.1016/j.epsl.2021.117088
  13. Amulele, G., Karato, S., & Girard, J. (2021). Melting of bridgmanite under hydrous shallow lower mantle conditions. Journal of Geophysical Research: Solid Earth, 126(9), e2021JB022222. https://doi.org/10.1029/2021JB022222
  14. Dong, J., Fischer, R. A., Stixrude, L. P., & Lithgow-Bertelloni, C. R. (2021). Constraining the volume of Earth's early oceans with a temperature-dependent mantle water storage capacity model. AGU Advances, 2(1), e2020AV000323. https://doi.org/10.1029/2020AV000323
  15. Andrault, D., & Bolfan-Casanova, N. (2022). Mantle rain toward the Earth's surface: A model for the internal cycle of water. Physics of the Earth and Planetary Interiors, 322, 106815. https://doi.org/10.1016/j.pepi.2021.106815
  16. Drewitt, J. W. E., Walter, M. J., Brodholt, J. P., Muir, J. M. R., & Lord, O. T. (2022). Hydrous silicate melts and the deep mantle H2O cycle. Earth and Planetary Science Letters, 581, 117408. https://doi.org/10.1016/j.epsl.2022.117408
  17. Ishii, T., Ohtani, E., & Shatskiy, A. (2022). Aluminum and hydrogen partitioning between bridgmanite and high-pressure hydrous phases: Implications for water storage in the lower mantle. Earth and Planetary Science Letters, 583, 117441. https://doi.org/10.1016/j.epsl.2022.117441
  18. Gu, T., Pamato, M. G., Novella, D., Alvaro, M., Fournelle, J., Brenker, F. E., Wang, W., & Nestola, F. (2022). Hydrous peridotitic fragments of Earth's mantle 660 km discontinuity sampled by a diamond. Nature Geoscience, 15, 950-954. https://doi.org/10.1038/s41561-022-01024-y
  19. Huang, R., Boffa Ballaran, T., McCammon, C. A., & Frost, D. J. (2023). The composition and redox state of hydrous partial melts generated at the top of the lower mantle. Earth and Planetary Science Letters, 624, 118447. https://doi.org/10.1016/j.epsl.2023.118447
  20. Hao, S., Wei, S. S., & Shearer, P. M. (2024). Substantial global radial variations of basalt content near the 660-km discontinuity. AGU Advances, 5(6), e2024AV001409. https://doi.org/10.1029/2024AV001409
  21. Ni, P., Shirey, S. B., Walter, M. J., Czas, J., Novella, D., Nestola, F., Kueter, N., Smith, E. M., Stachel, T., Pearson, D. G., Steele, A., Gardner, L. L., Jacobsen, S. D., Harte, B., Harris, J. W., & Shahar, A. (2025). Onset of slab mantle melting in Earth's lower mantle: Evidence from ferropericlase in superdeep diamonds. Science Advances, 11(42), eadt9106. https://doi.org/10.1126/sciadv.adt9106
  22. Khan, S., Andrault, D., Condamine, P., Schiavi, F., Devidal, J.-L., & Bolfan-Casanova, N. (2025). Genesis and fate of incipient melt in shallow lower mantle. Contributions to Mineralogy and Petrology, 180(10), 75. https://doi.org/10.1007/s00410-025-02260-2
  23. Chen, L., Liu, Z., Zhao, X., Xu, C., Ju, D., Feng, B., He, J., Liu, R., Li, X., Du, W., Liu, B., & Xu, W. (2025). Water solubility in nominally anhydrous minerals in a silica-rich system at the top lower mantle. Earth and Planetary Science Letters, 660, 119360. https://doi.org/10.1016/j.epsl.2025.119360
  24. Zhang, L., & Xia, Q.-K. (2025). Water in the lower mantle: constraints from mineral physics and phase equilibrium data. Physics of the Earth and Planetary Interiors, 369, 107475. https://doi.org/10.1016/j.pepi.2025.107475
  25. Carr, S. A. B., Olugboji, T., Park, J., & Karato, S. (2025). High-resolution mapping of North America suggests numerous low-velocity zones above and below the mantle transition zone. Tectonophysics, 908, 230775. https://doi.org/10.1016/j.tecto.2025.230775
  26. Wang, H., Magni, V., Conrad, C. P., & Domeier, M. (2025). Hydrous regions of the mantle transition zone lie beneath areas of continental intraplate volcanism. Geochemistry, Geophysics, Geosystems, 26(3), e2024GC011901. https://doi.org/10.1029/2024GC011901
  27. Li, S., Guo, J., Zhang, Y., Li, Y., Zhou, Z., Xu, W., Manthilake, G., & Weng, A. (2025). Water reservoirs in the lower mantle beneath northeastern Asia. Geophysical Research Letters, 52, e2024GL114180. https://doi.org/10.1029/2024GL114180