Estimating the amount of hydration of the subducting oceanic mantle using receiver function data

Water subducted into the mantle plays an important role in a wide range of Earth processes; it fuels volcanism, acts as a lubricant of the subduction megathrust fault, lowers the viscosity of the asthenosphere and is the prime catalyst for global element cycles. While the amount of water subducted in sediments and the oceanic crust has been validated through off-shore drilling experiments, little is known about the amount that enters the system through serpentinized outer-rise bend-faults. As measurements are scarce the questions arise: How much water is stored in the subducting mantle? How does the amount vary across subduction zones? What are the factors that control mantle hydration? As a proof of concept, we analyze P-to-S wave conversions in teleseismic receiver function data from a weak stratigraphic horizon in the Cascadia forearc that is located at about the depth to which the slab mantle is thought to be hydrated. The overlying subducted material likely hosts serpentinized bend-faults and is elastically anisotropic. The anisotropy direction matches that of dry olivine fabric inherited during sea-floor spreading, but the magnitude of anisotropy suggests an additional contribution from a bend-faulting fabric. Preliminary petrophysical modeling shows that an addition on the order of >20 vol% of serpentinized bend-faults explains the observed anisotropy pattern better than olivine anisotropy alone. A similar P-to-S conversion signal has previously been detected in the forearcs of Mexico, Costa Rica and Nankai; and there are indications that oceanic mantle hydration occurs globally. We plan to generalize our single seismic station approach to estimate oceanic plate hydration and apply it in subduction zone forearcs around the world.