Melting and Reactive Flow of Carbonated Peridotite Beneath Mid-Ocean Ridges

Abstract:

The mantle carbon reservoir is four orders of magnitude more massive than that of the atmosphere and ocean combined. The behaviour of carbon in the mantle, especially its transport and extraction, is thus of crucial importance to understanding the coupling between the deep interior and the surface environment of Earth. Laboratory experiments indicate that even small concentrations of carbon dioxide (and other volatiles like H₂O) in the upper mantle significantly affect silicate melting [HK96,DH06] by stabilising carbon-rich melt at high pressure. The presence of carbon in the mantle substantially extends the region where partial melt is stable and has important consequences for the dynamics of magma transport and chemical differentiation [H10,DH10]. We have developed theory and numerical implementation to simulate thermo-chemically coupled magma/mantle dynamics in terms of a two-phase (rock+melt), three component (dunite+MORB+carbonated MORB) physical model. The fluid dynamics is based on McKenzie’s equations [McK84]. The thermo-chemical formulation of the system is represented by a novel, disequilibrium, multi-component melting model based on thermodynamic theory [RBS11]. This physical model is implemented as a parallel, two-dimensional, finite-volume code that leverages tools from the PETSc toolkit.

First results show that carbon and other volatiles cause a qualitative difference to the style of melt transport, potentially enhancing its extraction efficiency – measured in the carbon mass flux arriving at the mid-ocean ridge axis – by at least an order of magnitude. The process that controls magma transport in our models is a volatile flux-induced reactive infiltration instability, causing carbonated melt to rise from depth in localized channels. These results add to our understanding of melt formation and transport at mid-ocean ridges (the most important magmatic system in the mantle) and may have important implications for subduction zones.

REFERENCES
HK96 Hirth & Kohlstedt (1996), EPSL
DH06 Dasgupta & Hirschmann (2006), Nature
H10 Hirschmann (2010), PEPI 
DH10 Dasgupta & Hirschmann (2010), EPSL
McK84 McKenzie (1984), J Pet
KW12 Katz & Weatherley (2012), EPSL
RBS11 Rudge, Bercovici & Spiegelman (2011), GJI