Results

Figure 13.11 shows the best-fitting coseismic slip and afterslip distributions. The cumulative afterslip after 9 months is shown. The afterslip is largely localized within two patches. One patch is in the upper 5 km, just above the larger coseismic slip patch. The other patch of afterslip occurs above the hypocenter. The fit to the GPS time-series is shown in Figure 13.12. The model reproduces the rapid postseismic velocities during the first 0.1-0.2 years and the less rapid velocities during the later time periods. There is a tendency for the model to under-fit the displacements at later times indicating that the model relaxes and returns to the pre-earthquake rate too quickly. This may indicate that another (un-modeled) deformation mechanism, such as deep distributed flow, may dominate the signal at later times.

Figure 13.11: A. Best-fitting coseismic slip distribution and, B. corresponding 9-month cumulative rate-state afterslip.
\begin{figure}
\begin{center}
\epsfig{file=kajjohnson_05_01.eps, width=8cm}
\end{center}
\end{figure}

Figure 13.12: Fit to the postseismic GPS time series. Only a subset of data is shown because of space limitations. Data shown with 95% confidence error bars. Solid line is model. Vertical axis is displacement in mm.
\begin{figure}
\begin{center}
\epsfig{file=kajjohnson_05_02.eps, width=8cm}
\end{center}
\end{figure}

Table 1 shows the approximate 95% confidence intervals on the frictional parameters. The values for A and B assume a uniform effective normal stress on the fault of 50 MPa. The parameters vary linearly with depth, so we report the values at the center of the top and bottom rows of patches. A-B is on the order of $10^{-4}$ to $10^{-3}$, which is about an order of magnitude lower than experimental values for granite at conditions well above or below the transition from potentially unstable (negative A-B) to nominally stable (positive A-B) friction. It is also an order of magnitude lower than an estimate from an afterslip inversion of the Tokachi-oki, Japan, earthquake (Miyazaki et al., 2004). However, the estimated A-B values fall within a wide range of experimental values reported for Serpentinite, which crops out along the San Andreas fault zone and is frequently mentioned as an important factor allowing for velocity-strengthening behavior of some faults (Reinen et al., 1994). The low values of A-B might be indicative of a transition zone from velocity-weakening (negative A-B) to velocity-strengthening (positive A-B). Estimates of $d_c$ are of the order 0.01-0.1 m, in reasonably good agreement with the seismic estimate near Parkfield (Fletcher and Spudich, 1998). The scaling relationship of Marone and Kilgore (1993) infers a fault gouge thickness of 1-10 meters.

Table 13.1: Table 1. 95% Confidence intervals on frictional parameters
parameter top of fault bottom of fault
A 0.0004-0.0052 0-0.024
B 0-0.0045 0-0.0468
A-B 0.0006-0.0018 0.0001-0.0021
$d_c$ (meters) 0.020-0.250 0-0.650

Berkeley Seismological Laboratory
215 McCone Hall, UC Berkeley, Berkeley, CA 94720-4760
Questions or comments? Send e-mail: www@seismo.berkeley.edu
© 2005, The Regents of the University of California