DAMAGE AND DRAINAGE: CAN RUPTURING FAULTS HOLD THEIR WATER?

The walls of a fault experience high stress fluctuations associated with the passage of an earthquake rupture front. The effect is particularly severe for ruptures moving close to the normal limit speeds for propagation, i.e., the Rayleigh speed for mode II and shear speed for mode III. Off-fault stresses, as calculated assuming elastic response outside a slip-weakening zone at the advancing rupture tip, are found to exceed the Mohr-Coulomb strength limit over distances into the fault walls with dimensions typically measured in meters (Poliakov et al., J. Geophys. Res., 2002). Sometimes local tensile stressing is predicted over limited regions (Rice et al., manuscript, 2003).

It is proposed that such a damage cycle near the rupture tip transiently changes fault zone hydrology during major earthquakes. The normally low-permeability fault walls are rendered much more permeable. That occurs typically to one side of the fault only, but possibly to both (the orientation of the pre-stress state controls). A consequence is that the walls of the shearing fault core are ineffective in helping to contain shear-heated pore fluids. The chief impediment to drainage and relaxation of their thermally elevated pore pressure becomes the core itself. Recent work suggests that active core zones, although extremely fine-grained and of small permeability (e.g., 10^{-19} m^2), may nevertheless be very thin (5 mm or less). Also, their permeability should be increased by granular dilatancy associated with rapid shear rates.

These considerations suggest that fault cores will sometimes have trouble bottling up their hot pore fluids, so that the effective stress reduction mechanism (Sibson, 1973; Lachenbruch, 1980; Mase and Smith, 1987) can operate. In such cases continued shear heating and partial melting of the gouge zone are expected outcomes in large slip events.