Introduction
Collaborating with Danan Dong of JPL and Bob King and Simon McClusky of MIT, we have made significant progress in 1997 to update the SCEC crustal deformation velocity map using reprocessed and newly processed GPS data. The total number of velocity vectors is 361, an increase of 74 vectors compared to the Version 1.0 velocity field, representing the most complete velocity map to date. New GPS data includes observation from Inter-county 92, Lander 93, LABS 97 and STRC 97. Reprocessed GPS data includes STRC 88, 89, 93, 95 (from MIT), HPGN 91-94, Inter-county 93, Gorman 92 and Lander 92. We are currently working with our colleagues towards a future release of version 2.0 of the velocity map.
Estimation procedures
GPS daily solutions were obtained using the GAMIT software and combined together using the GLOBK and QOCA (being developed by D. Dong) software to solve for station positions, velocities, and coseismic displacements. Velocities of a subset of fiducial sites were constrained using GSFC VLBI solution GLB1014. A priori coseismic displacement constraints for Joshua Tree, Landers, and Northridge earthquakes are obtained from independent coseismic studies. The EDM data were adjusted at JPL with loose constraints and then combined with the GPS velocities using QOCA software. For details of the estimation procedures and discussions about the estimation uncertainties, please see http://scec.ess.ucla.edu/velmap/welcome.shtml.
Results
Our GPS solution includes interseismic velocities for 204 stations with one-sigma uncertainties of both horizontal components less than 5 mm/yr. Figure 1 shows horizontal velocity vectors (relative to three VLBI stations on the North American Plate) and error ellipses for 361 sites after GPS and EDM combination. The error ellipses are regions of 95% confidence. The velocity field reveals about 48 mm/yr dextral shear motion across southern California, which is consistent with NUVEL-1A model prediction, suggesting a steady gross deformation across the plate boundary. This result is the same as what we obtained from the v1.0 velocity map solution (Shen et al., 1997a, b).
Figure 2 shows maximum strain rate interpolated from the velocity solution (see Shen et al., 1996 for interpolation method). It demonstrates that most of the rapid deformation occurs off the San Andreas fault, in the regions of past large earthquakes. For example, the epicentral areas of the 1992 M7.3 Landers, 1940 M6.9 and 1979 M6.4 Imperial Valley earthquakes show greater than 0.3 micro-radian/yr deformation. About 0.2 micro-radian/yr shear motion is detected in the 1952 M7.5 Kern County epicentral region. The only high strain concentration on the San Andreas is along the Parkfield-Carrizo Plain section. This result suggests that although earthquakes probably would not change the overall rate of relative plate motion across a plate boundary, they could change local deformation rates over a period of time from years to decades, depending on the magnitudes of the earthquakes (Jackson et al., 1997).
Pre- and Post- Landers velocity solutions
Figure 1 is based on the assumption that the velocity field did not change due to the Landers earthquake. We also compared station velocities before and after the Landers earthquake to determine whether the deformation rates have changed since that earthquake. While the pre-Landers solution is constrained using VLBI global velocities from GSFC, two post-Landers solutions are constrained using PGGA/DGGA velocities from SIO and slots of global IGS yearly solutions from MIT respectively. The two post-Landers solutions give virtually the same result.
The results are shown in Figure 3. Sites located south of the epicentral region show a few mm/yr NE motion with respect to their pre-Landers velocities. Sites NW of the epicentral region show a few mm/yr additional NW motion. Discrepancies are less for sites located further to the west. Such pattern is consistent with the sense of the Landers post-seismic deformation. Therefore, this preliminary result seems to confirm significant change of deformation rates after the Landers earthquake, although the pre-Landers station velocities might suffer from systematic errors related to the global reference frame due to relatively sparse coverage of global sites.
Reference
Shen, Z.-k., D.D. Jackson, B. Ge, Crustal deformation across and beyond Los Angeles basin from geodetic measurements, JGR, Vol. 101, pp. 27957-27980, 1996
Publications resulting from this funding
Jackson, D.D., Z.-k. Shen, D. Potter, X.-B. Ge, and L.-y. Sung, Southern California Deformation, Science, Vol. 277, pp. 1621-1622, 12 September 1997.
Shen, Z.-k., D. Dong, T. Herring, K. Hudnut, D. Jackson, R. King, S. McClusky, L.-y. Sung, Crustal deformation measured in Southern California, EOS, Vol. 78, No. 43, 1997a.
Shen Z.-k., D.D. Jackson, L.-y. Sung, Geodetic strain rates and fault deformation modeling in Southern California, AGU Fall Meeting abstract, EOS, 1997b.
