Characterization of a Strain Rate Transient Along the San Andreas and San Jacinto Faults Following the October 1999 Hector Mine Earthquake.

We are continuing work on developing and refining a tool for recognizing strain rate transients as well as for quantifying the magnitude and style of their temporal and spatial variations. We determined time-averaged velocity values in 0.05 year epochs using time-varying velocity estimates for continuous GPS station data from the Southern California Integrated GPS Network (SCIGN) for the time period between October 1999 and February 2004 [Li et al., 2005]. A self-consistent model velocity gradient tensor field solution is determined for each epoch by fitting bi-cubic Bessel interpolation to the GPS velocity vectors and we determine model dilatation strain rates, shear strain rates, and the rotation rates. Departures of the time dependent model strain rate and velocity fields from a master solution, obtained from a time-averaged solution for the period 1999-2004, with imposed plate motion constraints and Quaternary fault data, are evaluated in order to best characterize the time dependent strain rate field. A particular problem in determining the transient strain rate fields is the level of smoothing or damping that is applied. Our current approach is to choose a damping that both maximizes the departure of the transient strain rate field from the long-term master solution and achieves a reduced chi-squared value between model and observed GPS velocities of around 1.0 for all time epochs. We observe several noteworthy time-dependent changes. First, in the Eastern California Shear Zone (ECSZ) region, immediately following the October 1999 Hector Mine earthquake, there occurs a significant spatial increase of relatively high shear strain rate, which encompasses a significant portion of the ECSZ. Second, also following the Hector Mine event, there is a strain rate corridor that extends through the Pinto Mt. fault connecting the ECSZ to the San Andreas fault segment in the Salton Trough region. As this signal slowly decays, shear strain rates on segments of the San Andreas fault, just east of Palm Springs, and the San Jacinto fault increase during 2001-2004. During this period shear strain rates increase by roughly 20 nanostrain per year on the San Andreas fault and 20-30 nanostrain per year on the San Jacinto fault (over a zone approximately 20 km wide). Lastly, a further investigation into this strain rate recovery reveals a power law flow mechanism during the first six months after the earthquake for the Anza segment, after which strain rates appear to reach a steady state for the remainder of the data. Moreover, seismicity rates increase along these segments following the period of shear strain rate increase. These results quantify the spatial coverage of the strain rate changes and provide some bounds on their magnitude and confidence, as well as constraints on the associated regional rheology and interseismic cycle strain rate pattern. The compiled epoch solution "movies" may be viewed at the additional resources site.