Earthquake Rupturing in Fluid-Overpressured Crust: How Common?

Abstract

Whether or not ruptures nucleate in fluid-overpressured crust (λ v = P f/σ v > 0.4) is important because pore-fluids overpressured above hydrostatic lower fault frictional strength and may also vary through the earthquake cycle, acting as an independent variable affecting fault failure. Containment of fluid overpressure is precarious because pressure-dependent activation of faults and fractures allows drainage from overpressured portions of the crust. Discharge of fluids through activated fault-fracture permeability (fault-valve action) decreases overpressure so that subsequent failure depends on the cycling of both overpressure and frictional strength as well as tectonic stress. Geometric and mechanical considerations suggest that fluid overpressures are more likely to develop and be sustained in compressional/transpressional regimes as opposed to extensional/transtensional tectonic settings. On the basis of geophysical observations and force-balance analyses, subduction interface shear zones appear to be strongly but variably overpressured to near-lithostatic levels (λ v > 0.9) over the full depth range of seismogenic megathrusts. Strong overpressuring at seismogenic depths is also documented in active fold-thrust belts and in areas of ongoing compressional inversion (e.g., northern Honshu) where inherited normal faults are reactivated as steep reverse faults, requiring near-lithostatic overpressures (λ v → 1.0) at depths of rupture initiation. Evidence for overpressuring around strike-slip faults is less clear but tends to be strongest in areas of transpression. In areas of extensional tectonics coincident with particularly high fluid discharge, there is some evidence of overpressuring concentrated towards the base of the seismogenic zone. In general, because of the limited resolution of geophysical techniques, it is easier to make the case for rupture propagation through overpressured crust than to make a definitive case for the direct involvement of overpressured fluids in rupture nucleation, though in some instances the circumstantial evidence is compelling. An unresolved related issue is the heterogeneity of overpressuring. Do the active fault zones themselves serve as fluid conduits that are locally overpressured with respect to the surrounding crust?