Numerical modeling of volatile cycling beneath subduction zones

The subduction zone water cycle, i.e. the hydration and dehydration of subducting oceanic lithosphere, is a key process in understanding arc magmatism and volatile recycling processes. However, budgets of the subduction zone water cycle continue to have large error bars attached to them. In this study we will show how numerical modeling techniques can help to integrate various geological, geophysical, and geochemical datasets and to put bounds on the likely amounts of water being subducted, released into the arc melting region, and recycled into the deeper mantle. To achieve this task we use a suite of numerical models of different complexity. Bending related faulting and hydration of the incoming lithosphere offshore Chile and Nicaragua is resolved using a novel reaction-transport model that couples water circulation to serpentinization reactions. We find that the temperature dependent kinetics of the serpentinization reaction are likely to control the hydration patterns at the trench outer rise leading to the formation of a band of highly serpentinized mantle around the 270°C isotherm. The models further predict a reduction in surface heat flow values in the outer rise region that is qualitatively consistent with observations. A detailed analysis reveals, however, that the observed seafloor temperature gradient in the bend-fault region is too low to be caused by ‘one-pass’ downward water flow into the serpentinizing lithosphere, but rather implies that bend-faults are areas of active hydrothermal flow, with the implied prediction that serpentine-sourced vents and chemosynthetic vent communities should be found in this deep-sea environment, too. Dehydration occurs deeper within the subduction zone by fluid releasing metamorphic reactions. These rising fluids flux the mantle wedge where they are commonly believed to trigger arc melting. Using 2D and 3D numerical models we have resolved the likely flow field in the mantle wedge. The kinetics of serpentinization results in maximum serpentinization around the 270°C isotherm of the incoming slab and also has the maximum potential for water release during dehydration at depth. The depth of maximum serpentinization increases with increasing plate age and is consistent with the spacing of double Benioff zones (DBZs) now observed in several subducting slabs. Finally, we have resolved the mantle flow field within the mantle wedge as a function of subduction rate and slab fluid release in 3D. We find that the classical 2D corner-flow solution is only a small subset of all possible mantle wedge flow fields. In fact, a more “natural” flow field involves 3D diapirs fuelled by low-density slab fluids rising from the slab surface. These diapirs provide a potential mechanism for decompression melting in the mantle wedge, break the classic corner flow solution, and illustrate the need for high-resolution three-dimensional subduction zones models.