Fault Rerupture during the July 2019 Ridgecrest Earthquake Pair from Joint Slip Inversion of InSAR, Optical Imagery, and GPS

Abstract

The Ridgecrest earthquake pair ruptured a previously unknown orthogonal fault system in the eastern California shear zone. The stronger of the two, an Mw 7.1 earthquake that occurred on 6 July 2019, was preceded by an Mw 6.4 foreshock that occurred 34 hr earlier. In this study, distinct final slip distributions for the two earthquakes are obtained via joint inversion of Interferometric Synthetic Aperture Radar (InSAR), optical imagery, and Global Positioning System (GPS) measurements. Special attention is paid to the merging of dense (e.g., InSAR and optical imagery) and sparse geodetic (e.g., GPS) datasets. In addition, a new approach is introduced for data and model discretization through intermittent model- and data-space reconditioning that stabilizes the inversion, thus ensuring that small changes in the data space do not cause disproportionate large changes to the model space. Although the coseismic slip of the Mw 6.4 earthquake was complex, involving three distinct asperities distributed among an intersecting orthogonal set of faults, the coseismic slip of the Mw 7.1 earthquake was limited to the main northwest-striking fault. In addition to the Mw 7.1 earthquake, that northwest-striking fault plane also hosted one of the Mw 6.4 asperities. Slip on this coplanar foreshock asperity increased the shear stress at the future site of the Mw 7.1 hypocenter, and triggered a vigorous aftershock activity on the main northwest fault that culminated in its rupture. This, in turn, reactivated the coplanar foreshock asperity. In addition to failing twice within 34 hr, we find that the reruptured asperity slipped about six times more during the Mw 7.1 than during the Mw 6.4 earthquake. This repeated failure is indicative of an incomplete stress drop and premature rupture arrest during the Mw 6.4 foreshock, requiring an efficient frictional strengthening and emphasizing the causal link between highly rate-dependent friction, dynamic frictional restrengthening, and partial stress drop that has been observed in numerical studies of frictional sliding.

Coseismic Surface Displacement in the 2019 Ridgecrest Earthquakes: Comparison of Field Measurements and Optical Image Correlation Results

“…A curated inventory of the ground observations of measurable surface displacement form the basis for the near‐field observations considered in this study (DuRoss et al., 2020a) (Figure 1). Satellite observations from a mixture of optical (e.g., WorldView, Sentinel, Planet Labs, Pleiades) and radar sensors (e.g., Sentinel), yield insight into the resulting surface displacements (Barnhart et al., 2019b, 2020b; Fielding et al., 2020; Magen et al., 2020; Milliner & Donnellan, 2020; Xu et al., 2020). We use previously published optical pixel correlation results (Barnhart et al., 2020b) to measure net lateral displacement (Figure 2).…”

Structural Controls Over the 2019 Ridgecrest Earthquake Sequence Investigated by High‐Fidelity Elastic Models of 3D Velocity Structures

“…Such elevated in situ rock strength could favor stress‐accumulation and brittle rock failure (Ma et al., 2016; Romano et al., 2014). The vertical profile shown in Figure 10b further reveals that most slip (up to 7.4 m at 3.6 km depth) is resolved at a depth of 2–10 km where most aftershocks occurred (Barnhart et al., 2019; Chen et al., 2020; Feng et al., 2020; Goldberg et al., 2020; Jin & Fialko, 2020; Li et al., 2020; Magen et al., 2020; Pollitz et al., 2020; K. Wang et al., 2020). However, the lateral profile shows that aftershocks (of density >150 per 1,000 km 3 ) cluster beyond the major slipping areas along the NWF (Figure 10a).…”

“…(2020) (supporting information, Figure S6). Localized prediction misfits could be associated with the unaccounted fault segmentation (Hudnut et al., 2020) and minor geometrical deficiencies (Barnhart et al., 2019; Magen et al., 2020) (supporting information, Figures S6c and S6f). The above validations reaffirm that our InSAR‐resolved model sufficiently reflects the coseismic slip distribution and satisfactorily recovers the horizontal displacement field.…”

“…Validating the simulated displacement field predicted by our optimum slip model with the nearby GPS stations' coseismic displacements, we find a satisfactory agreement with minimal residuals <0.025 m (Figure 4), given the maximum GPS displacement is as high as 0.7 m. Our optimum model can also reproduce the overall displacement fields mapped by pixel-offset tracking SAR images provided by Fielding et al (2020) (supporting information, Figure S6). Localized prediction misfits could be associated with the unaccounted fault segmentation (Hudnut et al, 2020) and minor geometrical deficiencies (Barnhart et al, 2019;Magen et al, 2020) (supporting information, Figures S6c and S6f). The above validations reaffirm that our InSAR-resolved model sufficiently reflects the coseismic slip distribution and satisfactorily recovers the horizontal displacement field.…”

Repeating Nontectonic Seasonal Stress Changes and a Possible Triggering Mechanism of the 2019 Ridgecrest Earthquake Sequence in California

“…We do not consider dynamic or static stress changes during or following the M w 6.4 and the M w 7.1 events, although the magnitude of such stress changes must be orders of magnitude higher than the preseismic seasonal stress changes we present in this manuscript (e.g., Barnhart et al, 2019;Chen et al, 2020;Goldberg et al, 2020;Jin & Fialko, 2020;Lozos & Harris, 2020;Magen et al, 2020;Mancini et al, 2020;Qiu et al, 2020;Ramos et al, 2020;Toda & Stein, 2020;Wang et al, 2020). Our interest is rather to quantify and characterize preseismic nontectonic stress changes, which we argue in this manuscript may have provided preseismic stress increases on right-lateral faults within the area of rupture.…”