Reconciling extensive mantle hydration at subduction trenches and limited deep H2O fluxes

The long-term global sea level depends on the balance of H₂O exchanges between the Earth's mantle and the surface through both volcanism (mantle degassing) and subduction of hydrous minerals (mantle regassing). The estimates of H₂O fluxes by the current thermopetrological subduction models predict that regassing exceeds degassing by 60%, which may lead to a sea-level drop of at least a hundred meters in the last 540 Ma [Parai & Mukhopadhyay, 2012, Earth Planet. Sc. Lett., 317, 396-406. https://doi.org/10.1016/j.epsl.2011.11.024. These models further imply a moderate ( Tg/Myr) global input of H₂O at the subduction trenches. In contrast, geological constraints suggest a near-steady state of long-term sea level while geophysical observations advocate for a larger global H₂O input, especially given the large amounts of hydrated lithospheric mantle that are inferred at present-day subduction trenches. To address this paradox, we revise the subduction-H₂O flux calculations using recently published experimental data on natural hydrated peridotites at high-pressure conditions, which suggest that all hydrated phases destabilize below 800˚C for pressures higher than 8 GPa [Maurice et al., 2018, Contrib. Mineral. Petrol, 173(10), 86. https://doi.org/10.1007/s00410-018-1507-9 ]. Our reassessed thermopetrological models show that a prominent global H₂O input ( Tg/Myr), mainly conveyed by the layer of subducted serpentinized mantle, is compatible with a limited global H₂O retention in subducted slabs at mid-upper mantle depths ( Tg/Myr), including in models that consider some worldwide variability of the input serpentine. We also show that the global H₂O retention at mid-upper mantle depths is only driven by the hydrated mantle of coldest subducting plates. Overall, our models show that the present-day global water retention in subducting plates beyond mid-upper mantle depths barely exceeds the estimations of mantle degassing, and thus quantitatively support the stable-sea level scenario over geological times.