• Mid-Ocean Ridges
• Shear Zones
• Postseismic
• Subduction

Mid-Ocean Ridge

Thermal Structure of Ultraslow Ridges

Bathymetry of the Southwest Indian Ridge (SWIR)



Expected melt body underneath the SWIR

Melt extraction appears inefficient at the slowest ridges. We show that ultraslow spreading corresponds to an effective spreading rate below 6.2 mm/yr. According to our analytical solution for mantle flow beneath oblique spreading centers and our scaling analysis of their thermal structure, vertical melt extraction by porous flow is deeper than ~30km underneath ultraslow ridges. With Mark Behn; Publication 17; Support from The National Science Fundation, grant OCE-0649103.

Patterns of melt extraction at segmented Mid-Ocean Ridges

We contruct numerical models of the relief at the base of the lithosphere underneath segmented mid-ocean ridges. These models allow us to follow the trajectory of magma at it rises towards the ridge axis. We are now working on producing synthetic gravity and geochemical datasets to better compare these models against data. With Mark Behn, Henry Dick, Jeff Standish, and Jennifer Barry (undergraduate student); Support from The National Science Fundation, grant OCE-0649103

Spreading rate dependence of gravity anomalies along oceanic transform faults

We contruct numerical models of the thermal structure underneath segmented mid-ocean ridges at various spreading rates and use them to remove the contribution of the mantle thermal structure from observed gravity fields. This study revealed a change from thinned to thickened crust in the transform domain as spreading rate increases. This probably reflects a change in the style of crustal accretion with spreading rate. With Trish Gregg (graduate student), Jian Lin, and Mark Behn. Publication 15.

Rift propagation at Ridge-Ridge-Ridge triple junctions

The Galapagos Ridge stops a few km away from the East Pacific Rise. We show using numerical models of elastic stress fields in presence of free surfaces that instead of a direct connection between the ridges, the approach of the Galapagos Ridge is expected to form small rifts away from the triple junction, exactly at the position currently occupied by the Incipient Rift. With Hans Schouten, Debbie Smith, Wenlu Zhu, and Garrett Mitchell (undergraduate student); Publication 18.

Grain size reduction underneath a strike-slip fault

Shear Zones

Microstructure evolution

Ductile shear zone are marked by reduced grain size and phase separation. I have proposed constitutive models for grain size evolution and development of layers. I am currently incorporating these formulations into 2D models of lithosphere deeformation underneath a brittle fault to examine their efficiency at forming a ductile shear zone. with Frédéric Gueydan, Greg Hirth; Publications 9, 14.

Localization processes

Sequence of graben in a stretched icy crust

in order to compare the efficiency of various localization processes, I developped a measure of the rate at which a given process proceeds, the effective stress exponent. I am currently examining a new criterion based on a coinmparison between the work required to deform a rock in a distributed and localized state. In this way, I can show which mechanism is most likely to dominat in planetary lithospheres. with Maria Zuber, Frédéric Gueydan; Publications 2.

Fault spacing

Using an analytical model, I showed the relation between spacing of faults and the structure of the lithosphere. This work contrained the structure of the Indian plate in the Diffuse Plate Boundary area and the heat of Mars at the time of wrinkle ridge formation. Numerical models show consistency with this analysis and are helpful to understand the importance of reactivation and lateral thermal gradients. New application to Venus and ganymede are under work, as well as the inclusion of ductile localization processes. New software using Adaptative Mesh Refinement is under development. With Maria Zuber, Geoff Collins; Publications 6, 7, 8.

Postseismic deformation

Time-dependence of postseismic creep

Postseismic creep refers to the long-lived deformation transient that is observed to follow major earthquakes. I have derived analytical solution of the expected time-dependence of postseismic creep and related that to the non-linearity of the rheology of the active region. Inversion of actual geodetic records indicates stronger non-linearity than any published flow law, which could indicate tectonic reloading or effect of depth-dependent rheology. We are developing numerical models with depth-dependent rheology and include them in an inversion framework. With Jeff McGuire; Publication 10

3D visualization of seismicity in the Aleutians

Strength of ductile shear zones

Postseismic creep is often observed to originate below the seismogenic zone. Stress transfer models combined with the aseismic velocity jump trigerred by earthquakes imply a very low strength to the shear zone. Very small grain size may be required to explain the amplitude of the observed signals although I am currently applying rheological models for layered multiphase rocks to explore if these would be compatible with geodetic data. With Greg Hirth; Publication 11.

Subduction zones

Earthquake clusters

Intermediate seismicity does not occur uniformly at subduction zones. We determined the position and robustness of earthquake clusters in the Alaska-Aleutian subduciton zone and related their position to the edges and possible tears of the slab. With Piyapa Dejtrakulwong (undergraduate student).

Slab advance

Thermal structure of a subduciton zone

We explored the thermal evolution of the mantle wedge sujected to advance of a subducting slab to explain the origin of the Denali volcanic gap in Alaska and its apparent replacement by a zone of melt accumulation at the base of the lithosphere. With Stéphane Rondenay and Geoff Abers

Slab Rollback

We are constructing 3D finite element models of mantle flow around a finite-width slab during rollback and work on determining strain ellipses. We find that boundary conditions control whether slab-parallel flow exists or not and are in the process of comparing model results with actual subduction zones.