Physically realistic forward geochemical modeling of MORB: a comparison of INVMEL and pHMELTS
Abstract
Traditional methods of geochemical modeling mid-ocean ridge basalt (MORB) petrogenesis via melting, mixing and fractional crystallization have predominantly focused on reproducing observed chemical signatures with minimal consideration of whether the physical scenarios are dynamically possible. We use a 2D kinematic model of melting beneath a mid-ocean ridge to provide physically realistic constraints for two existing geochemical programs: pHMELTS (Asimow et al., 2004; Ghiorso et al., 2002) and INVMEL (McKenzie & O'Nions, 1991). Both programs have been updated with recent mantle source compositions and partition coefficients for trace element calculations. The geophysical model is based on passive upwelling, described by simple 2D corner flow, and adiabatic decompression melting beneath a spreading ridge. The model setup allows for time-dependent thermal and melt profile calculations which can be extracted at appropriate time steps and desired distances from the ridge axis. 1D melt-fraction depth profiles are used to directly constrain the forward geochemical calculation using INVMEL and determine an appropriate P-T-path for the evolution of pHMELTS. Both geochemical modeling techniques have been applied to major and trace element data from a variety of mid-ocean ridge settings including: the Juan de Fuca Ridge and Vance Seamounts; the 9-10°N segment at the East Pacific Rise; and plume-ridge interaction at Iceland. One important observation is the great disparity of predicted melt productivity variation with depth between the kinematic model and pHMELTS calculations. The depth at which melting starts has a significant impact on the trace element composition of melts due to variable amounts of garnet-field melting. pHMELTS significantly underestimates the amount of garnet field melting required to reproduce the degree of depletion of the heavy Rare Earth Elements (REEs), relative to the middle REEs, in MORB. Studies such as this one are critical for trying to conduct dynamically accurate geochemical simulations and making software users aware of the advantages and disadvantages of each program for determining geochemical behavior.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2012
- Bibcode:
- 2012AGUFM.T13G2723W
- Keywords:
-
- 1009 GEOCHEMISTRY / Geochemical modeling;
- 1032 GEOCHEMISTRY / Mid-oceanic ridge processes;
- 3035 MARINE GEOLOGY AND GEOPHYSICS / Midocean ridge processes