ABSTRACT
A model for intermediate-depth earthquakes of subduction zones is evaluated based on shear localization, shear heating, and runaway creep within thin carbonate layers in an altered downgoing oceanic plate and the overlying mantle wedge. Thermal shear instabilities in carbonate lenses add to potential mechanisms for intermediate-depth seismicity, which are based on serpentine dehydration and embrittlement of altered slabs or viscous shear instabilities in narrow fine-grained olivine shear zones. Peridotites in subducting plates and the overlying mantle wedge may be altered by reactions with CO2-bearing fluids sourced from seawater or the deep mantle, to form carbonate minerals, in addition to hydrous silicates. Effective viscosities of magnesian carbonates are higher than those for antigorite serpentine and they are markedly lower than those for H2O-saturated olivine. However, magnesian carbonates may extend to greater mantle depths than hydrous silicates at temperatures and pressures of subduction zones. Strain rates within altered downgoing mantle peridotites may be localized within carbonated layers following slab dehydration. A simple model of shear heating and temperature-sensitive creep of carbonate horizons, based on experimentally determined creep laws, predicts conditions of stable and unstable shear with strain rates up to 10/s, comparable to seismic velocities of frictional fault surfaces. Applied to intermediate-depth earthquakes of the Tonga subduction zone and the double Wadati-Benioff zone of NE Japan, this mechanism provides an alternative to the generation of earthquakes by dehydration embrittlement, beyond the stability of antigorite serpentine in subduction zones.
ABSTRACT
Fracture energy is a form of latent heat required to create an earthquake rupture surface and is related to parameters governing rupture propagation and processes of slip weakening. Fracture energy has been estimated from seismological and experimental rock deformation data, yet its magnitude, mechanisms of rupture surface formation and processes leading to slip weakening are not well defined. Here we quantify structural observations of the Punchbowl fault, a large-displacement exhumed fault in the San Andreas fault system, and show that the energy required to create the fracture surface area in the fault is about 300 times greater than seismological estimates would predict for a single large earthquake. If fracture energy is attributed entirely to the production of fracture surfaces, then all of the fracture surface area in the Punchbowl fault could have been produced by earthquake displacements totalling <1 km. But this would only account for a small fraction of the total energy budget, and therefore additional processes probably contributed to slip weakening during earthquake rupture.
