Near-Surface San Andreas Fault Location and Dip Near Woodside, California From Tomographic Vp, Vs, and Vp/Vs Ratios

Abstract:

The slip history of the San Andreas Fault (SAF) on the San Francisco peninsula is not well determined, and paleoseismic investigations to date yield inconsistent results, possibly because previous measurements were made on differing near-surface traces of the SAF. Furthermore, ground-shaking modeling requires accurate shallow-depth S-wave velocities across the fault zone and accurate fault dips; the existing Bay Area 3-D velocity models lack such measurements. To locate all near-surface traces of the SAF within 150 m of the 1906 surface rupture, to determine near-surface shear-wave velocities (V_S30 to V_S100), and to determine the fault dip, the USGS acquired a set of 300-m-long, high-resolution, P- and S-wave seismic imaging profiles across the SAF near Woodside, California, centered on the 1906 surface rupture zone. Sources (seisgun and hammer) and receivers (40-Hz P-wave and 4.5-Hz S-wave) were spaced at 3-m intervals. We developed independent P- and S-wave tomographic velocity models to depths of ~ 120 m (P-waves) and 80 m (S-waves). P-wave velocities vary widely from near the surface (Vp = 800 m/s, Vs = 250) to 100 m depth (Vp > 3000 m/s, Vs > 500 m/s). The 1906 surface rupture zone forms P- and S-wave low-velocity zones (Vp = 1600 to 1800 m/s; Vs = 250 to 350 m/s) within the fault zone relative to outside the fault zone (Vp = 2000 to 2200 m/s; Vs = 300 to 400 m/s). Vp/Vs ratios range from about 2.8 to about 5.7, with higher ratios on either side of the 1906 surface rupture zone and on the uphill (southwest) side. Southwest-dipping Vp/Vs contours suggest a 60^o-southwest dip for one of the fault traces. However, a CDP reflection stack shows a near-vertical dip of the main 1906 rupture zone to 1.2 km depth, and the stack shows evidence for multiple fault traces. Collectively, these new data show a complex fault structure and highly variable velocities across the fault zone, which should aid paleoseismic investigations and improve existing ground-shaking models.