Estimating variations in locking depth for the Mojave segment of the San Andreas Fault over the past 1500 years from paleoseismic stress drop

The most fundamental model for characterizing earthquake occurrence [ Reid, 1910] assumes a constant stress accumulation rate on a locked fault segment, which eventually fails at a threshold stress level. While this model provides a conceptual framework for investigating the conditions that prepare a fault for failure, there is little evidence that faults rupture periodically or at a uniform threshold stress. Moreover, this model fails to explain the large differences in earthquake recurrence intervals and paleoseismic slip for segments of major fault systems: why do some faults rupture on the order of 10\u27s of years, while others require 100\u27s of years to accumulate substantial stress before failing in a large earthquake? Previous work has shown that the rate of stress accumulation is highly sensitive to fault locking depth, where shallowly locked faults accumulate stress at higher rates than deeply locked faults. Here we investigate the role of variable locking depth on multiple earthquake cycle stress accumulation spanning the last 1500 years. We focus on the Wrightwood paleoseismic site, located along the Mojave segment of the San Andreas Fault System. We use published slip estimates (0.5-7.0 m) [Weldon et al., 2004] to simulate 14 earthquake cycles of interseismic stress accumulation, coseismic stress drop, and postseismic stress relaxation over the last 1500 years using a 4D viscoelastic deformation model with realistic fault geometry. For each earthquake cycle, we assume a constant slip rate and systematically adjust the locking depth of the Mojave segment (5-25 km) to match paleoseismic stress drop. Finally, we investigate the implications of paleoseismic stress accumulation and variable locking depth in context with present-day stress estimated along the Mojave segment of the San Andreas Fault System