The relationship between seismic velocity and internal fault structure was investigated through a shallow seismic refraction experiment across the Punchbowl fault, Devil's Punchbowl Los Angeles County Park, California. The Punchbowl fault is a northwest-striking, large-displacement fault of the San Andreas system that is exhumed to several kilometers depth and places crystalline basement against arkosic sandstone of the Punchbowl Formation. Seismic refraction profiles using hammer and impulsive shear-wave sources along a 300-m-long line reveal the velocity structure of the fault zone beneath a thin deposit of alluvium. We determine a velocity model assuming the alluvial layer is fairly uniform in velocity and thickness consistent with geologic observations and P-wave travel times. Raytracing with damped least-squares inversion of travel times of P and S waves indicate that the Punchbowl fault is best modeled as a zone several tens of meters wide with velocities reduced by 10%-25% from wall-rock velocities (V p ‫ס‬ 3.2 km/sec for granitic basement and V p ‫ס‬ 2.9 km/sec for Punchbowl Formation). Thickness of the low-velocity zone and the variation in seismic velocity across the zone are qualitatively consistent with expectations based on the observed distribution of fault-related fracturing and alteration. Apparent crack densities calculated from measured seismic velocities using O'Connell and Budiansky (1974) formulation for a cracked medium range from about 0.4 in the core to a background crack density of 0.1 in the host rock. The variation in calculated crack density across the fault is similar to observed variations in microfracture density in the Punchbowl Formation sandstone along traverses across the fault. An estimate of the Poisson's ratio near the fault is about 0.25, suggesting that open cracks in the shallow part of the Punchbowl fault zone are dry, consistent with the geologically inferred location of the groundwater table. Although the seismic data do not completely constrain the velocity structure, the seismic velocity model determined by raytracing and inversion of travel times is admissible on the basis of structural data.

A 600-m borehole vertical seismic profiling (VSP) survey conducted in crystalline rock in the Mojave Block stress province, southern California, yielded 100 oriented three-component seismograms from a fan-shooting geometry with radius of 100 m using both vertical and horizontal vibrators and a vertical impact source. The seismograms show up to 12 ms of shear wave splitting and 5% lateral velocity heterogeneity. First-and second-order ray tracing was used to model the travel times and amplitudes of P-, SV-and SH-waves in a three dimensional mildly heterogeneous medium with elastic anisotropy induced by aligned fractures. A least squares inverse procedure was used to improve model parameters determined by forward modeling; the final root mean square travel time residual was +_2 ms. For purposes of amplitude computation we assume that the wavelets observed in a VSP section derive from a smooth intermediate interface (base of weathering layer or sedimentary/basement interface) that is illuminated by a narrow range of ray angles originating at a point source on the surface. Source-derived amplitude complications are thus m'mimized. Our results strongly indicate a population of vertical cracks with an average crack density of 0.035 oriented in the N31.5øW direction. Such a fracture population is consistent with regional principal strain and stress determinations. Ray series geometric amplitudes were consistent with the observation that wave motion parallel to the aligned fractures decayed more slowly than wave motion transverse to the aligned fractures. We estimate that Q > 75-100; we were unable to measure a polarizationdependent Q. et al., 1970; Nur, 1971; Gupta, 1973; Anderson and Whitcomb, 1973; Kanamori and Hadley, 1975; Crampin, 1978]. Subsequently, the existence of fractures in basement rock was hypothesized by Smithson et at. [1980] and directly studied in the shallow basement by Green and Mair [1983]. Recently, shear wave splitting observations were obtained by threecomponent borehole vertical seismic profiling (VSP). Oriented three-component VSP data have provided a direct means of characterizing a fracture population in crustal rock [Martin et a/.,1986; Corrigan et a/.,1986; Leary et at., 1987; Majer et al., 1988; Li et al., 1988]. For instance, a crack density profile for oriented fractures was determined for a section of an active normal fault zone at Oroville, California [Leary et a/.,1987] and a Vsv/VSH ratio about 1.12 was estimated in the Geysers geothermal field, CA [Majer et al., 1988] from shear wave splitting observations in the plane of the local prevailing fracture set. The methods of three-component borehole seismic investigation of aligned fracture populations in crystalline rock are further explored in this paper. In contrast to the threecomponent borehole VSP work at Oroville [Leary et a/.,1987] Copyfight 1990 by the American Geophysical Union Paper number 89IB01170 0148-0227/90/89 IB-01170505.00 and at the Geysers geothermal field [Majer et al.,1988], where the principal fracture trends are ...