Focusing Fluids towards the Arc: the Role of Rehydration Reactions and Rheology

Abstract:

Aqueous fluids released from the down-going slab in subduction zones are generally thought to be the cause of arc volcanism. However there is a significant discrepancy between the consistent location of the volcanic front with respect to intermediate depth earthquakes (e.g. 100+/-40 km; England et al., GJI, 2004, Syracuse & Abers, G-cubed, 2006) and the large depth range over which dehydration reactions are predicted to occur in the slab (e.g. 80-250 km; van Keken et al., JGR, 2011).

By coupling the fluid flow to the solid rheology through compaction pressure, recent numerical models (Wilson et al., EPSL, 2014) demonstrated a number of focusing mechanisms that can be invoked to explain this apparent discrepancy. Most notable among these were permeability channels within the slab. These were shown to be highly effective in transporting fluid from deeper fluid sources along the slab towards the shallowest source. In the presence of these channels the majority of the fluid is released into the mantle wedge far shallower and closer to the arc than it was originally generated.

While observations consistent with free fluids in the slab have been reported (e.g. Shiina et al., GRL, 2013), it is possible that changing the reactivity and rheology of the slab can change the efficiency of in-slab transport (e.g. Wada et al., EPSL, 2012, Faccenda et al., G3, 2012). We present a series of simplified model problems of fluid flow within the slab and mantle wedge demonstrating the potential effect of these processes on fluid flux. In particular, pseudo-1D models show that if fluids can efficiently rehydrate slab minerals, then these reactions can shut down fluid pathways within the slab, resulting in deeper release of fluid into the mantle wedge. We will expand these results to consider the effects of rehydration in 2-D calculations. In addition, our previous models have considered only the simplest rheologies and geometries for the slab. We will also discuss new results that investigate simple visco-plastic models for the slab that limit the stresses and maximum viscosities in the slab for more realistic slab geometries. Despite these additional complexities, the robust observation of the location of the volcanic front with respect to intermediate depth earthquakes provides a clear test for evaluating subduction zone models.