Geodetic measurements provide the best constraints on the rate of strain accumulation on faults within the San Francisco Bay area. Together with paleoseismic and other geologic observations, geodetic data are central in long-term forecasts of earthquake hazards. An enormous amount of relevant geodetic data has been collected over the last century and a half, the vast majority in the last two decades. In collaboration with P. Segall (Stanford), we are integrating existing geodetic data to elucidate the spatial and temporal patterns of crustal deformation within the Bay Area. The focus of our effort is on refining the interseismic deformation field and three-dimensional strain accumulation models that are consistent with it, and on investigating spatio-temporal variations, such as postseismic deformation.
Since the early 1970's, three geodetic techniques have been used to measure horizontal crustal deformation in the Bay Area. Geodolite trilateration (EDM) measurements conducted by the USGS during 1972-1989 show that San Andreas fault (SAF) system in northern California accommodates approximately 34 mm/yr (e.g., Lisowski et al., 1991), but the EDM networks did not span the entire active zone of deformation, which may extend to the Sierra Nevada-Great Valley (SNGV) block (Argus and Gordon, 1991). VLBI measurements in the 1980's at 5 stations between the coast and the SNGV spanned the broader deformation field, but were too sparse to assess its spatial distribution (e.g., Ward, 1990). Since 1987, the USGS, and many of the local universities, including Stanford and UC Berkeley, have conducted approximately annual Global Positioning System (GPS) surveys of several networks and profiles across the SAF system, including a broad-scale geodetic network from the Sierras to the Farallon Islands, as well as five dense profiles in the Bay Area (e.g., Williams et al., 1994; Bürgmann et al., 1997). Since 1991, the Bay Area Regional Deformation (BARD) permanent network, currently consisting of 40 continuously monitoring GPS receivers, has provided continuous, nearly real-time monitoring of deformation and a regional framework for other GPS surveys in northern California.
Despite this extensive data collection and modeling studies, important questions remain unanswered. Although the cumulative deformation across the Bay Area is reasonably well constrained, it is still not clear how that motion is partitioned amongst the various Bay Area faults. In part, this is due to the non-uniqueness inherent in the inverse problem, in part because rigorous analyses of extremal models have not been conducted, and in part due to lack of knowledge of the fault geometry below the seismogenic crust. There is still a clear need to improve our understanding of the kinematics and mechanics of faulting in the San Francisco Bay area.
To better address such issues, we are currently finalizing a determination of a self-consistent velocity field for northern California in collaboration with many of the investigators that have made measurements in the past (Murray et al., 1998). For this task, we have obtained solutions of GPS site positions and their covariance derived from observations collected by USGS and Stanford University, as well as the BARD daily site positions estimated by the USGS and UC Berkeley since 1993. These solutions were determined by each of the investigators using different analysis packages and using different processing strategies. For example, 1879 daily solutions for the BARD permanent site positions were estimated by M. Murray at UC Berkeley using GAMIT, which assumes loose a priori constraints on all site positions, satellite orbits, and earth orientation parameters. 176 solutions for Bay Area GPS profile sites were estimated by W. Prescott's group at the USGS using two different GIPSY strategies (fiducial-free network and precise point positioning) that apply loose a priori station constraints but fixed orbits and possibly satellite clocks. 203 solutions for sites near Loma Prieta were estimated by R. Bürgmann using Bernese, many of these using tight or fixed a priori station constraints. We have converted these estimates to the SINEX solution format used by the international GPS community. We have also obtained VLBI position SINEX solutions estimated by the Goddard Space Flight Center, and Geodolite line length changes collected in the Bay Area and near Cape Mendocino.
We have found the task of combining the different GPS solutions to be more challenging than originally envisioned. Our original approach was to use solutions with loosely constrained a priori site coordinates that can be combined by Kalman filter techniques which allow additional constraints to be applied a posteriori to rigorously define a consistent reference frame (e.g., Herring et al., 1990; Dong et al., 1998). This approach, which has been applied to southern California, successfully, as part of a very large multi-year multi-institutional effort (Shen et al., 1997), minimizes biases introduced by using tight or absolute constraints on inaccurate fiducial site coordinates. In practice, we have found the different analysts and analysis packages introduce fundamentally different definitions of the reference frame that hinder rigorous combinations. For example, GAMIT loose constraint allows complete freedom to define the reference frame a posteriori, whereas GIPSY fiducial-free network and precise point positions solutions do not reflect uncertainties in satellite orbits. We are currently investigating an alternative approach, which uses GIPSY precise point positioning solutions (Zumberge et al., 1997) provided by the USGS, that allows for rapid reprocessing of all the primary GPS observations in a self-consistent manner.
The first version of the velocity field will focus on Bay Area deformation excluding the transient effects of the 1989 Loma Prieta earthquake from a combination of the VLBI and Geodolite data collected prior to 1989 and GPS data collected after 1993. A preliminary version of this field, from a combination of BARD and Bay Area profile GPS solutions, shows nearly stable Sierran block motion indicative of Basin and Range spreading, radial deformation associated with the Long Valley Caldera volcanic unrest, predominantly right-lateral strike-slip motion across the Bay Area faults with a suggestion of compression across the East Bay hills (Figure 11.1). We are currently assessing how to best define a "stable" North America velocity reference frame from a combination of globally distributed VLBI and GPS stations, which we suspect may be responsible for some of the non-fault parallel motions.
When the velocity field is finalized, we will begin to explore 3-D interseismic strain accumulation models that allow for along-strike variations in slip rate and the effects of surface creep and variable locking depth, and we will apply methods we developed for a study of the northern California Coast ranges ( Freymueller et al., 1999) to assess how well the fault slip-rates are resolved. In collaboration with S. Kenner (Stanford), we will also apply physical constraints based on finite element calculations to help distinguish between possible fault models.
Argus, D. F., and R. G. Gordon, Current Sierra Nevada - North America motion from very long baseline interferometry: Implications for the kinematics of the western United States, Geology, 19, 1057-1152, 1991.
Bürgmann, R., P. Segall, M. Lisowski, and J.P. Svarc, Postseismic strain following the 1989 Loma Prieta earthquake from Repeated GPS and Leveling Measurements, J. Geophys. Res., 102, 4933-4955, 1997.
Dong, D. N., T. A. Herring, R. W. King, Estimating regional deformation from a combination of space and terrestrial geodetic data, J. Geod., 72, 200-214, 1998.
Freymueller, J. T., M. H. Murray, and P. Segall, Kinematics of the Pacific-North America plate boundary zone, northern California, J. Geophys. Res, 104, 7419-7442, 1999.
Herring, T. A., J. L. Davis, and I. I. Shapiro, Geodesy by radio interferometry: The application of Kalman filtering to the analysis of very long baseline interferometry data, J. Geophys. Res., 95, 12,561-12,581, 1990.
Lisowski, M., J.C. Savage, and W.H. Prescott, The velocity field along the San Andreas fault in central and southern California, J. Geophys. Res., 96, 8369-8389, 1991.
Murray, M. H., W. H. Prescott, R. Bürgmann, J. T. Freymueller, P. Segall, J. Svarc, S. D. J. Williams, M. Lisowski, and B. Romanowicz, The deformation field of the Pacific-North America plate boundary zone in northern California from geodetic data, 1973-1989 (abstract), EOS Trans. AGU, 79(45), Fall Meeting Suppl., F192, 1998a.
Shen, Z.-K., D. Dong, T. Herring, K. Hudnut, D. Jackson, R. King, S. McClusky, and L.-Y. Sung, Crustal deformation measured in southern California, EOS Trans AGU, 78, 477, 1997
Ward, S. N., Pacific-North America plate motion: New results from very long baseline interferometry, J. Geophys. Res., 95, 21,965-21,981, 1990.
Williams, S.D.P., J. L. Svarc, M. Lisowski, and W. H. Prescott, GPS measured rates of deformation in northern San Francisco Bay region, California, 1990-1993, Geophys. Res. Lett., 21, 1511-1514, 1994.
Zumberge, J. F., M. B. Heflin, D. C. Jefferson, M. M. Watkins, and F. H. Webb, Precise point positioning for the efficient and robust analysis of GPS data from large networks, J. Geophys. Res., 102, 5005-5017, 1997.