Gray E. Bebout (Dept. Earth & Environmental Sciences, Lehigh University, Bethlehem, PA 18015 USA; 1-610-758-5831; e-mail: geb0@lehigh.edu). Ken Caldeira (Climate System Modeling Group, Lawrence Livermore National Laboratory, 7000 East Ave, L-103, Livermore, CA 94550 USA; 1-925-423-4191; e-mail: kenc@llnl.gov)
We develop and present results from a carbonate-silicate cycle model of the global carbon cycle that considers subduction zone processes and interaction with the mantle. Global carbon cycle models often have had no adequate reresentation of subduction zone processes and have considered global carbon cycling as a process that is confined to Earth's atmosphere, oceans, land-surface, and crust. Here, we demonstrate the impact that consideration of subduction zone processes can have on predictions from this class of global carbon cycle models.
A few initial observations: (1) If, as seems reasonable, more rapid subduction leads to greater fractional return of subducted C to the mantle, then subduction arc volcanic C fluxes may be less than that estimated by a direct proportionality to subduction rates; (2) If sea-floor basalt-carbonate reservoirs are limited by fracture volume and this carbonate is a major contributor to arc volcanic fluxes, then an increase in subduction rates would produce an increase in CO2 degassing and atmospheric CO2 content, that would then diminish on the time-scale of about 50 Myr; (3) If, as seems reasonable, subduction zones have been returning organic C to the mantle in a ratio of greater than 1:4 organic-C:inorganic-C, then many isotope-based estimates of crustal abundances of organic C would have been overestimates; and (4) in many carbonate-silicate cycle models including Sr-isotopes, the assumption is made that both mid-ocean-ridge exchange and subduction zone degassing occur in direct proportionality; because this direct proportionality need not be maintained, subduction zone processes may have a significant impact on the intepretation of Sr-isotope records. Some insights from studies of subduction-related metamorphic suites (Franciscan-type and Alps) which bear on the model include: (1) a large fraction of the initially subducted reduced C (organic) and oxidized C (the latter in sediments and altered oceanic crust) is retained to great depth beyond forearcs, perhaps to >100 km (for sedimentary carbonate, the more impure calcareous lithologies, e.g., calcareous shales, are more likely to decarbonate in forearcs than the more pure limestones), depending on subduction-zone thermal structure. Imbalance between subducted C flux and that returned from the mantle by magmatism would indicate net modern buildup of subducted surficial/crustal C in the mantle; (2) in subducting sedimentary sequences, both the oxidized and reduced C fractions appear to be driven by prograde devolatilization from seafloor C-isotope values toward mantle C-isotope values near -5 per mil (PDB); and (3) a significant reservoir of vein and cement carbonate with C-isotope compositions of -20 to -3 per mil may be produced in sediments during forearc diagenesis and low-grade metamorphism and then more deeply subducted.
We will motivate the model with a brief review of the important subduction zone processes, present the model, and quantitatively discuss its results.