3D melt transport and crustal accretion at mid-ocean ridges

Posted on 08.25.2015
Model output showing the depth to the top of a melt surface for a model of a segmented oceanic transform fault. The plate boundary is indicated by the white line. Black arrows show melt migration pathways and illustrate how melt formed in the mantle will flow either to the ridge or into the transform fault, where it will be erupted.

The process of generating melt beneath mid-ocean ridges becomes inherently more complex in the vicinity of transform fault offsets. Melt generation within the mantle beneath mid-ocean ridges is generally thought as a relatively two dimensional process that is a function of spreading rate. As the mantle rises beneath mid-ocean ridges, the pressure decreases and this triggers the onset of melting (i.e., decompression melting). Once the melt is generated and there is enough of it present, it will begin to coalesce and form pathways so it can continue to move up towards the spreading center. How much melt is generated and where it is generated within the mantle is a function of the thermal structure (and/or spreading rate) and starting composition.

In the presence of a transform fault offset, flow pathways will be interrupted due to the colder transform fault boundary. This presents a number of complexities when try to understand and model what happens to melt. During the final stage of my thesis, I developed a method which combines petrological models that calculate fractional mantle melting with geodynamic models that calculate mantle flow and thermal structure. Since the problem of transform fault offsets is inherently three-dimensional, 3D models are used and applied them to a variety of spreading rates and transform fault geometries. 

This work allowed us to better understand how transform fault offsets affect the overall process of crustal accretion at mid-ocean ridges. 

COLLABORATORS: Dan Fornari (WHOI), Mike Perfit (U Florida), Jian Lin (WHOI), Mark Behn (WHOI), Tim Grove (MIT)

FUNDING: This work was funded by a National Science Foundation Graduate Research Fellowship and a WHOI Graduate Research Fellowship (Gregg)