AGU2016: Subduction and Dehydration of Slow-Spread Oceanic Lithosphere | Scientific Talk

This a seminar I gave at the American Geophysical Union Fall Meeting in San Francisco on the 16th of December 2016 as part of the session "Subduction Top to Bottom".

I present the latest results from my research project supported by the AXA Research Fund and the OBSIVA project, funded by a prize of the Foundation Simone and Cino del Duca, from the French Academy of Science.

See a 3D rotating animation of the subduction zone here:

For more information see:

Title: "Subduction and dehydration of slow-spread oceanic lithosphere"

Abstract:
Water transported by subducting slabs affects the dynamics of subduction zones and is a major gateway in the global geochemical water cycle. During subduction much of the water stored in the slab is released via pore fluid escape and through metamorphic reactions that depend on the thermal regime. The most notable are eclogitization of hydrated basalt and gabbro and breakdown of serpentinite. Most constraints to date have been obtained at Pacific subduction zones, and have contributed to a model of slab dehydration applicable to normal fast-spread oceanic lithosphere with a mafic crust. Slow-spread crust however, is heterogeneous in thickness and composition and has a different water distribution than fast-spread crust.

We use P-wave traveltimes from several active source seismic experiments and P- and S-wave traveltimes from shallow and intermediate depth (less than 160 km) local earthquakes recorded on a vast amphibious array of OBSs and land seismometers to recover the 3D Vp and Vp/Vs structure of the central Lesser Antilles subduction zone from the surface to 160 km depth. This slab was formed by slow accretion at the Mid-Atlantic ridge and represents the global slow accretion rate end-member. We image the dipping low-Vp layer at the top of the slab corresponding to the hydrated slab crust penetrating to about 100 km depth. High Vp/Vs ratio on the slab top and in the forearc crust is interpreted as evidence of elevated fluid content either as free fluids or as bound water in hydrated minerals. A local minimum in Vp is observed on the slab top at ~50 km depth, and forms an elongated trench-parallel anomaly. This anomaly is interrupted at the projection of the Marathon fracture zone. We suggest that this is the result of lateral variations in slab crust composition from normal mafic oceanic crust to tectonized oceanic crust consisting to a large extent of serpentinized peridotite near the fracture zone. Slab regions with normal mafic oceanic crust likely undergo eclogitization, resulting in voluminous water release over a narrow depth range. Serpentinized ultramafic crust, in contrast, may release water at a more constant rate. We infer that subduction of slow-spread lithosphere may result in heterogeneous water transport and release at subduction zones with implications for seismicity, magma generation and the geochemical budget.