Deeply-sourced fluids upwelling along a fault can control seismogenesis

Dmitry I. Garagash

Published August 1, 2019, SCEC Contribution #9300, 2019 SCEC Annual Meeting Poster #168

Near lithostatic pore fluid pressure and absolute weakness of major continental transform faults at depth have been suggested by observations of low (tidal) stress triggering of slow slip at the base of the San Andreas Fault (SAF) at Parkfield [Thomas et al 2009]. Mantle volatiles upwelling along deeply rooted faults at rates of ~mm/year, as inferred from geochemical analysis of fault fluids [Kennedy et al 1997], can lead to the pore pressure in excess of hydrostatic [Rice 1992]. We show that the fluid pressure can reach lithostatic values at depth when the hysteresis of the fault rocks’ permeability upon loading/unloading of the effective stress is accounted for in the modeling.

Specifically, we predict the asperity-like distribution of the effective-stress with depth: increasing along the lithostatic-hydrostatic gradient in the shallower part and then rapidly decreasing to near-zero values along the deeper part of the fault. The depth to the peak stress and its magnitude are decreasing functions of the upwelling fluid rate. For parameters representative of SAF fault rocks, varying fluid flow rate from low ~ 0.1 mm/yr to high ~ 3mm/yr results in upward migration of the “stress asperity” from 7.5 km to 3 km depth.

Fault seismogenesis is linked to unstable, e.g. rate-weakening (a-b<0), fault friction rheology over a certain, “seismogenic” depth range, while fault is presumed frictionally stable (a-b>0) at tectonic rates along the shallower/deeper parts of the fault. Earthquakes can nucleate within this “unstable frictional asperity” if the size of the nucleation process region (where dynamic fault instability develops) is not larger than the asperity size. Since nucleation size is inversely dependent on the effective stress, the seimogenesis is then directly linked to the depth “overlap” between the frictional and effective stress asperities, the latter’s position dependent on the deep fluid supply rate.

We therefore propose and demonstrate in 1D fault slip cycle simulations that seimogenic and aseismic/creeping segments of a fault (e.g. SAF) may stem not necessarily from the differences in the rheology, but rather from the along-strike variability of the mantle volatiles supply into the fault: the higher flow rates favoring weaker, creeping segments and slower flow rates favoring stronger, seismogenic ones.

Key Words
fault fluids, lithostatic pore pressure, fault strength, seismogenesis, slip cycle

Citation
Garagash, D. I. (2019, 08). Deeply-sourced fluids upwelling along a fault can control seismogenesis. Poster Presentation at 2019 SCEC Annual Meeting.

Related Projects & Working Groups
Fault and Rupture Mechanics (FARM)