Dominance of transient rheology in far-field postseismic deformation following the 2012 Mw 7.8 Haida Gwaii and 2013 Mw 7.5 Craig earthquakes

Abstract

The Mw 7.8 2012 Haida Gwaii earthquake and the Mw 7.5 2013 Craig earthquake occurred along the Queen Charlotte transpressive margin, two months and 340 km apart. On and near Vancouver Island, hundreds of kilometres southeast of the two events, many continuous Global Navigation Satellite Systems sites underwent no discernable coseismic motion but up to 1 cm southeast postseismic motion (away from epicenters) which lasted for only 2 – 3 years. To resolve the postseismic signal of the two earthquakes, we used a minimum of four years of measurements before the Haida Gwaii earthquake to remove the longer-term motion of these sites that reflect mainly the locking of the Cascadia megathrust. To study the geodynamic process that governed the observed far-field postseismic motion, in particular its short duration, we employ a model of an elastic lithosphere with thickness 20 km and rigidity (μE) 32 GPa overlying Burgers-viscoelastic mantle layers in a self-gravitating spherical Earth which captures the regionally integrated effect of Earth rheology despite the complex and controversial tectonics of the study area. We find that the short-lived far-field postseismic motion was controlled entirely by the transient rheology. The observations can be explained by a steady-state (Maxwell) viscosity ηM of the upper mantle no lower than 1×1019 Pa s with a transient (Kelvin) viscosity ηK of about 6×1017 Pa s (Kelvin relaxation time τK ∼1.2 years) extending to large depths such as > 400 km. The far-field postseismic deformation is the result of the elastic plate serving as a stress guide but resisted by the asthenospheric mantle undergoing transient creep. The lithosphere-asthenosphere coupling is quantified by a “plate Kelvin response time” τKE=ηK/μE. Rheological parameters can be scaled to other values, but as long as τK and τKE stay unchanged, the far-field deformation results remain the same. Our observations presented a unique opportunity to quantify the effect of the transient rheology separately from that of the steady-state rheology.