HR: 14:10h
AN: V33D-03    [Abstracts]
TI: Integrating Geodynamic and Petrological Numerical Models; Mid-Ocean Ridge Flow Dynamics Revisited
AU: * Smith, P M
EM: psmith@gps.caltech.edu
AF: GPS Caltech, 1200 E. California Blvd., Pasadena, CA 91125 United States
AU: Baker, L J
EM: labaker@gps.caltech.edu
AF: GPS Caltech, 1200 E. California Blvd., Pasadena, CA 91125 United States
AU: Hall, C E
EM: chall@gps.caltech.edu
AF: GPS Caltech, 1200 E. California Blvd., Pasadena, CA 91125 United States
AU: Asimow, P D
EM: asimow@gps.caltech.edu
AF: GPS Caltech, 1200 E. California Blvd., Pasadena, CA 91125 United States
AU: Gurnis, M C
EM: gurnis@gps.caltech.edu
AF: GPS Caltech, 1200 E. California Blvd., Pasadena, CA 91125 United States
AB: We have adapted the adiabat_1ph front-end to the pHMELTS petrological model and coupled it with a 2-D variable viscosity flow model, such that energy and mass transfer are treated self-consistently. The combined model, GyPSM, captures thermal and chemical controls on mantle dynamics and feedback between them. We have developed mid-ocean ridge and subduction zone versions (see Baker et al, this volume). MORs have been the most extensively studied igneous system but the lateral and vertical extent of melting is still uncertain and is generally represented as uniform upwelling in a triangular region extending to the base of the crust. How do processes such as cooling due to latent heat and the effect of volatiles on melting and rheology affect this geometry and the sensitivity to spreading rate, temperature and source heterogeneity? Most empirical parameterizations of melting invoked by geophysical models make simplifying assumptions e.g. anhydrous source, batch melting. The approach used in pMELTS (Ghiorso et al, 2002), and the pHMELTS extension for water-undersaturated conditions, ensures that predicted melt productivity, chemistry and key thermodynamic quantities are internally consistent. By adding the flow model, conservation of mass, momentum and energy can be simultaneously satisfied. At each incremental isentropic step, the equilibrium assemblage is found and then adjusted for melt extraction, diffusion and advection. Each of the thousands of pHMELTS calculations performed per iteration is associated with a particle and its path tracked. Scripts control program execution and exchange energetic and chemical information. The parallelized scheme is surprisingly robust and effective so long as ascending melt is not expected to react with the residue. Preliminary results suggest more rapid cooling and thickening of the lithosphere when melting is accounted for. The low-pressure viscosity is dominated by the cooling effect whereas the effect of water in olivine is noticeable at depth. Water promotes deep small degree melts that enhance the viscosity variation. The melting region has quite concave upper boundaries and seemingly does meet the base of the crust, at least for the chosen conditions, but the parameter space needs fuller investigation.
DE: 1011 Thermodynamics (0766, 3611, 8411)
DE: 1032 Mid-oceanic ridge processes (3614, 8416)
DE: 1065 Major and trace element geochemistry
DE: 8120 Dynamics of lithosphere and mantle: general (1213)
DE: 8159 Rheology: crust and lithosphere (8031)
SC: Volcanology, Geochemistry, Petrology [V]
MN: Fall Meeting 2005