SCEC Award Number 16239 View PDF
Proposal Category Individual Proposal (Integration and Theory)
Proposal Title Earthquake Cycles with Viscoelasticity and Rate-and-State Friction
Name Organization
Eric Dunham Stanford University
Other Participants Kali Allison (PhD student)
SCEC Priorities 1b, 1e, 3f SCEC Groups USR, Simulators, FARM
Report Due Date 03/15/2017 Date Report Submitted 04/03/2017
Project Abstract
The depth extent of earthquakes, postseismic deformation, and loading of the seismogenic layer are all sensitive to the viscoelastic response of the lower crust and upper mantle. Here we introduce a finite difference method for simulating earthquake sequences in heterogeneous viscoelastic solids with faults governed by rate-and-state friction. This is presently done in the context of the classic 2D antiplane shear model of a vertical strike-slip fault. The model employs power-law flow laws together with an assumed geotherm; temperature is held fixed in time. We vary the geotherm, and hence the viscosity structure, and explore the resulting effects on the partitioning of tectonic displacement into fault slip and viscous flow, as a function of depth, the relative contributions of afterslip and viscous flow to postseismic deformation, and time-dependent crustal strength profiles. We also find that in most cases, the down-slip limit of slip is controlled by the frictional transition from velocity-weakening to velocity-strengthening with increasing depth and temperature, rather than the onset of viscous flow. Another important conclusion is that loading of the seismogenic upper crust, earthquake nucleation, and recurrence intervals are relatively independent of whether tectonic displacement at depth is accommodated by aseismic fault creep or distributed viscous flow. However, there are potentially observable differences in postseismic deformation that could be used to determine the nature of deformation at depth.
Intellectual Merit Earthquake simulators and cycle simulations are now routinely used in many SCEC projects. Yet, with a few exceptions, these simulations assume linear elasticity with faults loaded by backslip. Our work advances simulation capabilities by accounting for viscous flow, using laboratory-derived flow laws, in the lower crust and upper mantle, together with rate-and-state fault friction. This development is the first step toward fully coupled thermomechanical earthquake cycle simulations, which will permit integration of a wide range of data sets to help constrain long-standing issues in earthquake physics, such as fault weakening mechanisms and crustal stress profiles.
Broader Impacts This project supported a female graduate student in the late stages of her PhD. The simulation capabilities developed in this project can be used to help determine the processes which limit the down-dip extent of seismic ruptures. This has important implications for scaling relations between fault dimension and moment, which are key in quantifying seismic hazard in projects like UCERF.
Exemplary Figure Figure 3. Slip histories in four models: (a) elastic medium, (b)-(d) layered viscoelastic medium (Figure 1b) with different geotherms (value given above each panel). Red curves are plotted every 1 s during coseismic rupture (slip velocity > 1 mm/s); blue curves are plotted every 10 yr. As the geothermal gradient increases and effective viscosity in the lower crust and upper mantle decreases, tectonic deformation at depth transitions from being accommodated by aseismic fault creep to bulk viscous flow (see Figure 2). Nucleation and propagation are similar across all models, but the postseismic response (i.e., afterslip) is quite different.