White Paper Title: 
Modeling mantle melting at mid-ocean ridges

A self-consistent “bottom-up” geodynamical model that combines thermal and petrologic models has been developed to investigate ridge processes. Specifically, mantle flow and thermal models are used in conjunction with the mantle melting model of Kinzler and Grove (JGR, 1992a, b; JGR, 1993) and the fractional crystallization model of Yang et al. (Contrib. Mineral. Petrol., 1996).  Combining fractional melting and crystallization models with mantle thermal models quantifies both the termination of melting and the onset of crystallization.  Outputs from these models include physical parameters such as crustal thickness and the major element composition of the extruded lavas.  We constrain our model using geophysical data (e.g., seismic and gravity) and major element geochemistry data.

This modeling technique has extensive applications for studying ridge processes and takes advantage of both geophysical and geochemical datasets. Some topics we are currently investigating include:

  1. Constraining local/global variations in mantle potential temperature
  2. Constraining local/global variations in mantle source composition
  3. Predicting the shape of the mantle melt region
  4. Tracking mantle melt compositions from depth and exploring melt migration through the mantle (e.g., What part of the melting region is tapped?  Which melts make it to the surface?  Where are the melts likely to stall?)
  5. Exploring the processes that affect the melt as it fractionally crystallizes in the crust (e.g., hydrothermal cooling)

To date, we have used this approach to investigate the shape of the melt region and resultant major element and crustal thickness variations associated with transform fault offsets and examined the specific example of the segmented Siqueiros transform fault (Gregg et al., JGR, 2009).  We are also utilizing the model for a global ridge study in which we investigate variations in mantle potential temperature and source composition at both the local (segment-scale) and global scale.