UTIG Seminar Series: Demian Saffer, Penn State
||February 19, 2016 at 10:30 am
||February 19, 2016 at 11:30 am|
| ||Location:||PRC, 10100 Burnet Road, Bldg 196, Rm 1.603, Austin, TX 78758|
| ||Contact:||Peter Flemings/Laura Wallace, firstname.lastname@example.orgemail@example.com, 512-475-8738/512-471|
| ||URL:||Event Link|
"Phyllosilicates, Friction, and Fluids: The Role of Clays in Plate Boundary Fault Processes"
Clay minerals, and hydrous smectite family clays in particular, have been hypothesized to control both the low absolute shear strength and the nature of slip along plate boundary faults, including subduction megathrusts, low-angle normal faults, and continental transforms. In addition to their effects on frictional behavior, hydrous clays are the largest and shallowest source of bound water in systems undergoing prograde diagenesis and metamorphism. For example, the transformation from smectite to illite at temperatures of ~80-150 °C releases ~20 wt% H2O. As a result, the reaction has the potential to generate fluid overpressures through expulsion of bound water; to change the elastic, hydraulic, and frictional properties of the rock or sediment via mobilization of quartz and associated cementation or vein-filling; and to generate geochemical signatures that yield information about fluid flow pathways from source regions at depth to observation points at the surface. Here, I summarize a set of integrated laboratory, numerical modeling, and field studies that address these linked processes, focusing on examples from the Costa Rican and Nankai (SW Japan) subduction zones.
Laboratory shearing experiments on a suite of synthetic and natural fault gouges demonstrate that clay- and mudstones sampled from major fault zones, including the San Andreas Fault and the Costa Rican and Nankai subduction megathrusts, are frictionally weak, with friction coefficients in the range µ = ~0.08-0.35. These fault rocks exhibit rate-strengthening behavior, which should lead to only aseismic creep. However, they become increasingly rate-weakening (and therefore potentially able to host unstable slip) with increasing quartz abundance, raising the possibility that clay transformation, primarily through the by-product of silica mobilization, may play a role in controlling the upper aseismic-seismic transition and its spatial variability or “patchiness” at some subduction margins.
Numerical models describing the kinetics of clay dehydration and associated fluid expulsion show that at both the Costa Rican and Nankai margins, earthquakes nucleate only in areas down-dip of the reaction, after it has peaked and is near completion. This behavior is robust, and it parallels variations in thermal structure along the strike of the Costa Rican margin. Although the number of observations is limited, the correlation further supports the idea that the onset of unstable slip is linked to clay transformation, likely through the combination of changes in rock frictional properties and the dissipation of the bound water source that allows effective normal stress to increase. Interestingly, the rupture areas of large (M>6-7) events extend into the zone of dehydration, suggesting the possibility that nucleation may be restricted to depths where the clay reaction has progressed, but that slip may propagate up-dip into weak and potentially overpressured regions where the dehydration reaction is peaking. Finally, these models can also be used to predict the evolution of pore fluid composition within subducting and accreting sediments due to the release of bound (fresh) water and the desorption of fluid-mobile elements like B and Li. Comparison of simulated pore fluid compositions in these source regions with observations at seafloor seeps and in shallow drillholes provides evidence for long-distance (10’s of km) up-dip fluid migration along fault zones.