Post-seismic stress evolution for a strike-slip fault in the presence of a viscoelastic asthenosphere

Abstract

SUMMARY A model for post-seismic, near-field effects on transcurrent vertical faults in the presence of a viscoelastic asthenosphere is presented. The lithosphere is modelled as an elastic layer welded to a viscoelastic half-space (the asthenosphere). The earthquake is modelled as a stress-drop event (crack), whereas in previous similar papers it has usually been modelled as a ‘constant-slip event’. Unlike crack models, constant slip dislocations are not suitable for describing the stress field in the proximity of the fault, which is the main concern of the present paper. The sliding section is supposed to have a fixed extension and to reach the free surface. Two cases are considered a fault which becomes welded immediately after an earthquake (‘locked’ problem), and the case in which the slip distribution evolves in the post-seismic phase (‘unlocked’ problem), following the stress relaxation of the asthenosphere. In both cases, the singular integral equation of crack equilibrium is solved in the Laplace transform domain. The locked problem is assumed to reproduce the post-seismic stress recovery following an earthquake on a fault, which preserves the initial strength threshold after the earthquake. Solutions show that major aftershocks in the deep section of the fault can occur if the fault cuts most of the lithosphere owing to relaxation of the asthenosphere. The unlocked problem concerns a fault on which the post-seismic strength is equal to the residual stress left after the earthquake, and post-seismic creep occurs; owing to relaxation of the asthenosphere, the fault keeps sliding after the earthquake, and afterslip can be observed at the ground surface. We compare the results obtained with the advanced phases of afterslip observed in some Californian faults. The stress evolution is a slower process in this case, with respect to the locked model, owing to the two-way relaxation, i.e. through fault creep and asthenosphere flow. The comparison of the unlocked model with afterslip data seems to require a fault cutting most of the lithosphere if the relaxation time for the asthenosphere is greater than ∼10 yr.