Project Abstract
|
Interseismic deformation measurements are generally interpreted in terms of steady slip on buried
elastic dislocations. While such models often yield slip rates that are in reasonable accord with geologic
observations, they are inconsistent with our expectation of fault structure at depth, and can not
explain transient postseismic deformation following large earthquakes. An alternate
two-dimensional model of repeating earthquakes that break an elastic plate of thickness $, overlying a
Maxwell viscoelastic half-space with relaxation time $\tau_{R}$ [Savage and Prescott, 1978] involves five parameters; , \tau_{R},
t, T,$ and $\dot{s}$, where $ is the time since the last quake, T is the earthquake repeat time, and
$\dot{s}$ is the slip rate. Many parts of the San Andreas fault (SAF) system involve multiple parallel
faults, which further increases the number of parameters to be estimated. All hope is not lost, however,
if we make use of geologic constraints on slip rate, as well as the measured
time-dependent strain following the 1906 earthquake, in addition to the present-day spatial
distribution of deformation rate.
GPS data from the Carrizo Plain segment of the SAF imply a considerably larger relaxation time
than inferred from the post-1906 strain rate transient recorded by repeated triangulation surveys. This
may indicate that the simple Savage-Prescott model is deficient, or that the crustal structure and/or thermal profile differs
significantly between northern and central California. The data are consistent with
an average recurrence interval of $\sim 280$ years, which agrees reasonably well with paleoseismic estimates.
To test the
effect of regional variations in $ and $\tau_{R}$, I analyze GPS derived velocities from the northern San Francisco Bay area
[Prescott \etal,
2001], and include the SAF, the Hayward-Rodgers Creek (HRC), and Concord-Green Valley faults (CGV).
Optimization using constrained non-linear least squares yields: \sim 18$ km, $\tau_{R} \sim$ 35 years, {SAF} = 280$ years,
$\dot{s}_{SAF} \sim 25 $ mm/yr, {HRC} = 225$ years, {HRC} = 429$ years, and
$\dot{s}_{HRC } \sim$ 7 mm/yr. The effect of the Concord - Green Valley faults is modeled kinematically; the optimal model
involves 8 mm/yr of slip beneath 2 km. While neither the average recurrence interval on the SAF or HRC
is well resolved, the optimal inter-event time for the SAF is in reasonable agreement with paleoseismic results
which indicate that the most recent event prior to 1906 occurred in the mid 1600's. The geodetic data also favor the
latest date allowed by
historical data ($\sim 1775$) for the most recent HRC earthquake, a result not inconsistent with the paleoseismic finding
that the most recent event there occurred in the interval 1670-1776. The conclusion of this study is that by
combining the present-day deformation field, post-1906 strain data, and geologic bounds on slip rate and maximum
earthquake slip, it is possible to estimate parameters of considerable geophysical interest, including time since the most recent
earthquake and average recurrence intervals. |
