Relative importance of poroelastic effects and viscoelastic relaxation for postseismic velocity fields after normal and thrust earthquakes: Insights from 2D finite-element modelling

Abstract

Earthquakes on faults in the brittle upper crust evoke sudden changes in pore fluid pressure as well as postseismic viscoelastic flow in the lower crust and lithospheric mantle but the relative importance of these processes during the postseismic phase has not been systematically studied. Here, we use two-dimensional finite-element models to investigate how pore fluid pressure changes and postseismic viscoelastic relaxation interact during the earthquake cycle of an intracontinental dip-slip fault. To isolate the effects from pore fluid flow and viscoelastic relaxation from each other, we performed experiments with and without pore fluid flow and viscoelastic relaxation, respectively. In different experiments, we further varied the permeability of the crust and the viscosity of lower crust or lithospheric mantle. Our model results show poroelastic effects dominate the velocity field in the first months after the earthquake. In models considering poroelastic effects, the surfaces of both hanging wall and footwall of the normal fault subside at different velocities, while they move upwards in the thrust fault model. Depending on the permeability and viscosity values, viscoelastic relaxation dominates the velocity field from about the second postseismic year onward although poroelastic effects may still occur if the permeability of the upper crust is sufficiently low. With respect to the spatial scales of poroelastic effects and viscoelastic relaxation, our results show that pore fluid pressure changes affect the velocity field mostly within 10–20 km around the fault, whereas the signal from viscoelastic relaxation is recognizable up to several tens of kilometres away from the fault. Our findings reveal that both poroelastic effects and viscoelastic relaxation may overlap earlier and over longer time periods than previously thought, which should be considered when interpreting aftershock distributions, postseismic Coulomb stress changes and surface displacements.