Melt Migration at a Ridge-Ridge-Ridge Triple Junction


DOEI Project Funded: 2005

What are the primary questions you are trying to address with this research?
What is the trajectory of batches of magma at a Ridge-Ridge-Ridge (RRR) triple junction? RRR triple junctions are defined at the junction of three mid-ocean ridges. Often, one ridge spreads much more slowly than the other two and it may be thought that it samples almost passively the melting system of the more dominant ridges. The models I propose to build will address this hypothesis by solving for flow trajectory and related geochemical evolution of the magma.

What is the significance of this research for others working in this field of inquiry and for the broader scientific community?
If indeed the slower ridge has a little effect on magma chemistry and trajectory, it is justified to use intersecting ridges as probes into mid-ocean ridge melting systems. If they have a strong influence, we’d better know about it before interpreting data! The three-dimensional code I will build can be applied to other tectonic settings

What is the significance of this research for society?
The software I want to build can address magma migration problems in volcanically active areas such as Hawaii and the Cascades, but it application to the relatively simple setting of RRR triple junction will facilitate model validation by comparing its results to geochemical observations

When and where will this investigation be conducted?
This project is all about theory and software development. I will work on it in early 2006 at WHOI. A recent cruise to the Incipient Rift area in the Pacific Ocean returned data that can be compared to model predictions

What are the key tools or instruments needed to conduct this research?
The numerical code that I will build uses both Finite Element and Finite Difference techniques to solve mantle flow, mantle temperatures, and melt trajectories. Much of it will be take advantage of the Linear Algebra library PETsC ( and will be linked to other software for geodynamics developed by the Computational Infrastructure for Geodynamics (CIG

What are the greatest challenges - physical or intellectual - to conducting this investigation?
There will be many different pieces to the envisioned software that have to be able to interact tightly with one another. This is not commonly done. The software will also need to be flexible and expandable so that it can be reused in many other research projects.

Is this research part of a larger project or program?
Hopefully, this software will be the starting point of several other projects aimed at understanding melt migration in three dimensions in various plate boundary settings.

Biographical information:
I was born in the Lorraine region of Northeast France, and studied for my undergraduate studies in Nancy and Paris. I majored in Physics but realized I was more interested in the Earth than tiny particles that I would never see (an emphasis of my program). I then specialized in geodynamics, squeezing rocks at SUNY Stony Brook and mapping a Martian volcano back in Paris. I moved permanently to the States to enroll for a PhD at MIT initially in Planetary Sciences and then in Geophysics. In spite of the excitement of witnessing images arriving daily from Mars in our lab, I grew more interested in exploring the workings of Earth, focusing on how melt is distributed, and on using patterns of faulting and geodetic signal to better understand the large scale behavior of the crust and how it influences earthquake generation.