Seismic Moment Accumulation Response to Lateral Crustal Variations of the San Andreas Fault System

Abstract

Rheologic variations in the Earth's crust (like elastic plate thickness [EPT] or crustal rigidity) modulate the rate at which seismic moment accumulates for potentially hazardous faults of the San Andreas Fault System (SAFS). To quantify rates of seismic moment accumulation, Global Navigation Satellite Systems, and Interferometric Synthetic Aperture Radar data were used to constrain surface deformation rates of a four‐dimensional viscoelastic deformation model that incorporates rheological variations spanning a 900 km section of the SAFS. Lateral variations in EPT, estimated from surface heat flow and seismic depth to the lithosphere‐asthenosphere boundary, were converted to lateral variations in rigidity and then used to solve for seismic moment accumulation rates on 32 fault segments. We find a cluster of elevated seismic moment rates (11–20 × 1015 Nm year−1 km−1) along the main SAFS trace spanning the historical Mw 7.9 1857 Fort Tejon earthquake rupture length; present‐day seismic moment magnitude on these segments ranges from Mw 7.2–7.6. We also find that the average plate thickness in the Salton Trough is reduced to only 60% of the regional average, which results in a ∼60% decrease in moment accumulation rate along the Imperial fault. Likewise, a 30% increase of average plate thickness results in at least a ∼30% increase in moment rate and even larger increases are identified in regions of complex plate heterogeneity. These results emphasize the importance of considering rheological variations when estimating seismic hazard, suggesting that meaningful changes in seismic moment accumulation are revealed when considering spatial variations in crustal rheology.

Vertical Postseismic Deformation of the 2019 Ridgecrest Earthquake Sequence

“…We chose to compare a mid‐field and far‐field station because both have observations before and after the EMC event, which we consider until 1 January 2019 (roughly 9 years after the rupture), while most near‐field station observations started after the EMC event and had lower resolution. Using a previously published heterogeneous elastic plate rheology model for the Salton Trough (Ward et al., 2021), we first explored three time‐averaged viscosities (1e18, 3e18, and 1e19 Pa·s) (i.e., Dickinson‐Lovell et al., 2018; Hines & Hetland, 2016; Rollins et al., 2015; Spinler et al., 2015; Tang et al., 2020) for postseismic displacement of the EMC event (Figure 7). In the mid‐field, we found that up to 67% of the cumulative vertical displacements and 24% of the cumulative horizontal displacements could be explained by our modeled viscoelastic relaxation when a viscosity of approximately 1e18 Pa·s is adopted.…”

“…Including the restoring force of gravity is essential; our calculations show that it reduces the far‐field postseismic vertical amplitude by up to 25%, with only very minor effects on the horizontal components (Smith & Sandwell, 2004). A laterally varying elastic plate rheology was also included in the model (Ward et al., 2021), based on surface heat flow and depth of the lithosphere asthenosphere boundary, where the resulting average crustal rigidity is 33 GPa. This region has an average plate thickness of 66 km, which defines the average transition depth between the elastic layer and underlying viscoelastic half‐space.…”

“…At longer times, the half‐space undergoes a linear viscoelastic response of the half space. The model includes lateral variations in shear modulus and viscosity to simulate large‐scale variations in rheology (Ward et al., 2021). Furthermore, the model includes long‐term earthquake cycle deformation related to past surrounding deep slip and paleoseismic earthquake ruptures so over thousand‐year timescales the deep and shallow fault slip are balanced.…”

“…However we found that a reduced Coso zone viscosity does not have a significant effect on the overall surface displacements, consistent with findings from Penney et al (2021), with only slight improvements of model fit at select GNSS stations within the Coso area. Further maxwell model specifications are described in the Supporting Information of Ward et al (2021).…”

“…We parametrically fit the time series for postseismic deformation using the Hector software package (Bos et al., 2013). We then compare the data‐derived deformation to a viscoelastic fault deformation model, maxwell (Sandwell & Smith‐Konter, 2018; Smith & Sandwell, 2004; Ward et al., 2021), and attempt to distinguish viscoelastic versus nonviscoelastic (i.e., commonly afterslip and poroelastic rebound) contributions of postseismic deformation. Given that a majority of vertical far‐field earthquake cycle deformation can be attributed to viscoelastic relaxation processes at timescales >1 year since rupture (as nonviscoelastic contributions significantly reduce after weeks to months following an earthquake), this analysis also focuses on the role of upper mantle viscosity transients and the consistency of such transients to previous regional events, as demonstrated by postseismic deformation of the 2010 Mw7.2 El Mayor‐Cucapah earthquake.…”