Reactive Transport of Carbonated Magma Beneath a Mid-Oceanic Ridge: Theory and Numerical Models

Laboratory experiments indicate that even small concentrations of CO2 in the upper mantle significantly affect the silicate melting behavior [DH06]. CO2 stabilizes carbon-rich melt at high pressure, thus vastly increasing the volume of the upper mantle expected to be partially molten [H10,DH10]. These small-degree melts have important consequences for chemical differentiation and could affect the dynamics of mantle flow. 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+basalt+volatiles) physical model. The fluid dynamics is based on McKenzie's equations [McK84], while the thermo-chemical formulation of the system is represented by a novel, disequilibrium approach to volatile-bearing mantle melting. An experimentally constrained ternary phase diagram, based on an effective solid solution, is used to characterize the equilibrium state in the system. This physical model is implemented as a parallel, two-dimensional finite-volume code that leverages tools from the PETSc library. Early application of this simulation code to a mid-ocean ridge system suggests that the methodology captures some of the first order features of carbonated mantle melting, including deep, low-degree, volatile-rich melt formation. Melt segregation leads to continuous dynamic thermo-chemical dis-equilibration, while phenomenologically derived reaction rates continually move the system towards re-equilibration. The simulations will be used to first characterize carbon extraction from the MOR system assuming chemically homogeneous mantle, and will subsequently be extended to investigate the consequences of heterogeneity in lithology [KW12] and volatile content. Such studies will advance our understanding of the role that the mid-ocean ridge system plays in the deep carbon cycle. REFERENCESDH06 Dasgupta & Hirschmann (2006), doi:10.1038/nature04612.H10 Hirschmann (2010), doi:10.1016/j.pepi.2009.12.003. DH10 Dasgupta & Hirschmann (2010), doi:10.1016/j.epsl.2010.06.039.McK84 McKenzie (1984), J.Pet.KW12 Katz & Weatherley (2012), doi: 10.1016/j.epsl.2012.04.042.