← Back to Sessions

Spatially Variable Creep on the Bartlett Springs Fault Inferred From Bayesian, Stress-Driven Models

Session: Effects and Uses of Aseismic Deformation and Fault Creep in Seismic Hazard and Warning [Poster]

Type: Poster

Date: 4/20/2021

Presentation Time: 04:15 PM Pacific

Description: 

Near-surface creep, if it extends to seismogenic depths, can reduce a fault’s potential for generating damaging earthquakes by influencing dynamic rupture propagation. Murray et al. (2014) inferred spatially-variable creep rates on the Bartlett Springs (BSF) and Maacama faults, and slip rates below assumed locking depths on these faults and the San Andreas Fault, by developing kinematic models using Bayesian inversion applied to GPS-derived velocities. The resulting spatially averaged creep rate estimates suggest that the BSF has substantial creep (~4.5 – 7.5 mm/yr) at all depths, and the data permitted a shallow locking depth (~5 km).

Lozos et al. (2015) used the creep rate distribution from Murray et al. (2014), and one estimated by Lienkaemper et al. (2014) using different data and methods, as initial conditions for dynamic rupture models of BSF earthquakes. The resulting ruptures varied substantially in the amplitude and along-strike extent of slip, with magnitudes ranging from M_w 6.32 – M_w 7.24. However, the kinematic creep rate models have low spatial resolution and lack information about where stress is accumulating, thus limiting how much detail the dynamic rupture models could provide.

The depth-extent of creep, the along-strike extent of non-creeping zones, and the creep rate relative to the deep slip rate are the primary factors generating the different dynamic rupture behavior seen using the Lienkaemper et al. (2014) versus Murray et al. (2014) models. Johnson (2013) outlined a Bayesian stress-driven modeling approach in which locations on a fault are either locked (and accumulate stress) or creep at a constant stress . This physically -constrained approach resolves the size of creeping patches better than purely kinematic inversions. We apply this method to an expanded version of the Murray et al. (2014) GPS velocity field to explore the range of BSF creep models that result under the different assumptions and constraints inherent in this physics-based approach with the goal of narrowing the range of initial conditions for future dynamic rupture modeling.

Presenting Author: Jessica R. Murray

Student Presenter: No

Authors