T43C-3012
Imaging spatially variable afterslip using stress-drive models governed by rate and state friction and constrained by geodetic observations following the South Napa earthquake

Thursday, 17 December 2015
Poster Hall (Moscone South)
Jessica R Murray, Sarah E Minson and Jerry L Svarc, U.S. Geological Survey, Earthquake Science Center, Menlo Park, CA, United States
Abstract:
Geodetic observations after the 24 August 2014 M6 South Napa earthquake provide insight into the manifestation of fault frictional properties in the interseismic, coseismic, and postseismic periods. The West Napa fault (WNF) was not known to creep, although data to confirm this are lacking. Absence of creep could indicate a fault is locked interseismically and characterized by velocity weakening (VW) friction. A VW fault would not exhibit afterslip. However, field and geodetic data record a months-long postseismic signal starting within hours of the mainshock. The spatial pattern is consistent with afterslip on the WNF, implying the presence of velocity strengthening (VS) fault zone material.

Marone et al. (1991) presented a one degree of freedom spring-slider model governed by VS friction in which shallow afterslip depends on the depth extent of the VS zone (h) and a rate and state friction variable (a-b). We fit this model to records from four alinement arrays on the WNF. For (a-b) in the laboratory range of 0.005 – 0.05, the data are consistent with h = 1 – 4 km, thinning to the northwest.

GPS data are sensitive to deeper afterslip than near-fault measurements. While frictional properties may be optimized by fitting the space-time distribution of afterslip inferred from kinematic inversion of GPS data, stress-driven models encompass more of the underlying physics. We use the first month of postseismic GPS time series to constrain a steady-state VS friction model driven by coseismic slip inferred from seismic data (Wei et al., 2015). The optimal uniform value of (a-b), 0.021, adequately fits many time series, but significant misfit exists for some. We are investigating variations in frictional properties along-strike and with depth that can better explain the data. Such variations, if correlative with lithology or other fault zone features, could inform assessment of expected coseismic slip and damaging afterslip on faults that do not exhibit surface creep.