Slip rates and off‐fault deformation in Southern California inferred from GPS data and models

Abstract

[1] In Southern California, fault slip rate estimates along the San Andreas fault (SAF) and Garlock fault from geodetically constrained kinematic models are systematically at the low end or lower than geologic slip rate estimates. The sum of geodetic model slip rates across the eastern California shear zone is higher than the geologic sum. However, the ranges of reported model and geologic slip rate estimates in the literature are sufficiently large that it remains unclear whether these apparent discrepancies are real or attributable to epistemic uncertainties in the two types of estimates. We further examine uncertainties in geodetically derived slip rate estimates on major faults in Southern California by conducting a suite of inversions with four kinematic models. Long-term rigid elastic block models constrained by the geologic slip rates cannot fit the present-day GPS-derived velocity field. Deforming (permanent off-fault strain) elastic block models and viscoelastic earthquake cycle block models constrained by geologic slip rates can fit the present-day GPS-derived velocity field with 28-33% of the total geodetic moment rate occurring as distributed deformation off of the major faults. Models incorporating viscoelastic mantle flow predict systematically higher slip rates than purely elastic models on many of the major Southern California faults with ranges of (elastic/viscoelastic) 29-34/30-37 mm/yr for the Carrizo SAF segment, 20-24/20-32 mm/yr for the Mojave SAF segment, 14-17/18-22 mm/yr for the Coachella SAF segment, 13-19/14-22 mm/yr for the San Jacinto fault, and 5-11/5-11 mm/yr for the western Garlock fault.

Recovery of secular deformation field of Mojave Shear Zone in Southern California from historical terrestrial and GPS measurements

“…Several mechanisms have been invoked to interpret the geologic‐geodetic slip rate discrepancy in the southern Mojave ECSZ (Text S6 in the supporting information), including earthquake cycle effects [ Chuang and Johnson , ; Johnson , ], long‐range fault interactions [ Dolan et al ., ], transient weakening of a ductile shear zone at depth [ Oskin et al ., ], and off‐fault deformation [ Herbert et al ., , ]. Our block motion models cannot discriminate between these candidate mechanisms, as we do not consider earthquake cycle effects or evolution of fault zone rheology, and we do not attempt to solve for distributed off‐fault deformation.…”

“…Geological studies [e.g., Oskin et al ., ] show the presence of several closely spaced and low slip rate (≤2 mm/yr) faults in the Mojave ECSZ, which define the potential block boundaries. Due to the uncertainties in slip rates and fault geometry, uncertainties arise as to the use of mapped surface traces of active faults to define discrete block boundaries, as shown in previous geodetic studies [ McClusky et al ., ; Miller et al ., ; Becker et al ., ; McCaffrey , ; Meade and Hager , ; Spinler et al ., ; Loveless and Meade , ; Johnson , ; Zeng and Shen , ]. Here the block models within the Mojave Block are objectively crafted on the base of the cluster analysis of the CMM4 GPS velocities rather than subjectively constructed by an arbitrary selection of mapped fault traces as block boundary faults.…”

“…The NEMD block is marked by west‐striking left‐lateral faults [ Garfunkel , ; Dokka and Travis , , ], in contrast to other blocks within the Mojave Desert that are cut by northwest trending right‐lateral faults, and has been already included in previous model studies [ Johnson , ]. However, comparing to the one in Johnson [], the western boundary of the NEMD is shifted eastward to the Goldstone Lake fault.…”

Analogue earthquakes and seismic cycles: experimental modelling across timescales

“…tectonic) deformation through seismic cycles (e.g. Wang and Hu, 2006;Johnson, 2013). Observation both in nature (e.g.…”

Bounding the moment deficit rate on crustal faults using geodetic data: Methods

“…The majority of these models assume an elastic crust, and discretize faults with either rectangular [ Okada , ] or triangular [ Maerten et al , ; Meade , ] elements (dislocations) in a half‐space [e.g., Segall and Harris , ; Johnson et al , ; Jónsson et al , ; McCaffrey , ; Fialko , ; Loveless and Meade , ; Xue et al , ]. Others have modeled the lower crust and/or upper mantle as viscoelastic [e.g., Hashimoto et al , ; Chuang and Johnson , ; Johnson , ]. SDR inversions have often been done within the context of elastic block models by estimating and subtracting deformation due to steady motion on block boundaries, then solving for back slip [e.g., Murray et al , ; Prescott et al , ; Meade and Hager , ; Fukuda and Johnson , ; Minson et al , ; Xue et al , ].…”