Kinematic Models of Southern California Deformation calibrated to GPS Velocities and a Strain Energy Minimization Criterion: How do they Differ?

Abstract:

Fault slip rates inferred from GPS-calibrated kinematic models may be influenced by seismic-cycle and other transient effects, whereas models that minimize strain energy (“TSEM models”) represent average deformation rates over geological timescales. To explore differences in southern California fault slip rates inferred from these two approaches, I have developed kinematic, finite-element models incorporating the UCERF3 block model-bounding fault geometry and slip rates from the UCERF3 report (Field et al., 2014). A fault segment (the “Ventura-Oak Ridge segment”) was added to represent shortening accommodated collectively by the San Cayetano, Ventura, Oak Ridge, Red Mountain and other faults in the Transverse Ranges. Fault slip rates are randomly sampled from ranges given in the UCERF3 report, assuming a “boxcar” distribution, and models are scored by their misfit to GPS site velocities or to their total strain energy, for cases with locked and unlocked faults. Both Monte Carlo and Independence Sampler MCMC methods are used to identify the best models of each category.

All four suites of models prefer low slip rates (i.e. less than about 5 mm/yr) on the Ventura-Oak Ridge fault system. For TSEM models, low rates (< 12 mm/yr) are strongly preferred for the San Gorgonio segment of the SAF. The GPS-constrained, locked model prefers a high slip rate for the Imperial Fault (over 30 mm/yr), though the TSEM models prefer slip rates lower than 30 mm/yr. When slip rates for the Ventura-Oak Ridge fault system are restricted to less than 5 mm/yr, GPS-constrained models show a preference for high slip rates on the southern San Jacinto and Palos Verde Faults ( > 13 and > 3 mm/yr, respectively), and a somewhat low rate for the Mojave segment of the SAF (25-34 mm/yr).

Because blind thrust faults of the Los Angeles Basin are not represented in the model, the inferred Ventura-Oak Ridge slip rate should be high, but the opposite is observed. GPS-calibrated models decisively prefer a lower slip rate along the Mojave segment of the SAF than TSEM models, consistent with viscoelastic velocity perturbations late in the SSAF seismic cycle. The low San Gorgonio SAF slip rates inferred from TSEM models are consistent with strain transfer from the SAF Coachella segment to the ECSZ, bypassing the San Gorgonio SAF segment (e.g., Wallace, 1984).