A 30 km‐long N‐S seismic reflection line was shot by California Consortium for Crust Studies (CALCRUST) across the southern Mojave Desert and onto the northern flank of the San Bernardino Mountains in southern California. On the northern end of the seismic section, the reflectivity increases markedly in the midcrust at a depth corresponding to a two‐way travel time of 4 to 5 s (12–15 km), suggesting a transition between nonreflecting brittle upper crust and reflecting ductile lower crust. The high reflectivity disappears at about 8 s (24 km) and may be correlated with a change in seismic velocity in the lower crust from 6.3 km/s to 6.8 km/s. A band of reflectivity between 9.5 and 10 s (27–30 km) is believed to represent the Moho. The midcrustal relectivity transition and Moho both deflect downward toward the San Bernardino Mountains uplift over the entire length of the profile. The deflection of the midcrustal transition (12°) appears greater than that of the Moho (6°), resulting in a thinning of the lower crust to the south beneath the uplift. In addition, the midcrustal transition coincides with the base of the seismogenic zone (brittle‐ductile transition?) which is also dipping southward beneath the San Bernardino Mountains, while the Moho deflection is consistent with elastic flexure resulting from edge loading by the San Bernardino Mountains which have been thrust over the Mojave block. It is suggested that the thinning of the lower crust beneath the San Bernardino Mountains is a result of north directed ductile flow in response to loading by the over thickened upper crust. Since a portion of the load is transmitted through the lower crust to the Moho, the time constant for flow equilibrium must be of the order of or greater than that for the time of uplift (≥2 m.y.).
In situ stress measurements on the San Andreas fault are modeled in terms of a solution to the equations of three‐dimensional elastic equilibrium in a layered medium. Crustal stresses are expressed in terms of surface deformations. In particular, vertical stress gradients are shown to be proportional to horizontal gradients of vertical surface deformation. In light of constraints derived by this relation it is proposed to reinterpret the observed near surface in situ vertical stress gradients on the San Andreas fault as partly a manifestation of a layered crust with vertically increasing rigidity. Extrapolation to depth of the revised stress gradients suggests that the San Andreas fault can be regarded as a low‐stress phenomenon compatible with estimates of shear stress obtained from heat flow data.
Seismic waves travelling through the crust generate trains of scattered waves called the seismic coda. Coda motion u(t) excited by a source pulse σ(t) and recorded at times t > ≈2t0 by a sensor located at traveltime t0 from the source can be written as the convolution of σ(t) with the gradient c′(r) = ∂rc(r) of spatial velocity fluctuations c(r) encountered by the wave front at r ≈ tc0/2: u(t) ≈ 4πu0(t0/t)〈c′(tc0/2)〉 *σ(t), where t0/t accounts for source wavelet spherical divergence, c0 is the mean seismic velocity of the medium, and 〈〉 denotes the scattering amplitude c′(r) averaged over the 4π solid angle of the wave front. It follows that the coda frequency spectrum u(f) is proportional to vc(v)σ(f), where v is the spatial frequency v= 2f/c0, vc(v) the spectrum of c′(r) and σ(f) the source spectrum. Sonic-velocity fluctuations cBH(r) logged at 15 cm intervals over a 1.5 km length of deep borehole in crustal rock show that the major cause of seismic velocity fluctuations c(r) in the brittle crust are fracture distributions with a power-law spectral dependence on fracture spacing, c(v)≈v−0.4. The resulting velocity gradient power-law spectrum is vc(v)≈v0.6. In accordance with the above scattering expressions, the power-law scattering amplitude vc(v) causes the high-frequency power-law enrichment of the coda spectrum relative to the source spectrum u(f)/σ(f)≈f0.6 as observed in coda waves recorded in a borehole at a depth of 2.5 km in the crust. The coda wave phenomenology observed in deep crustal borehole data does not emerge from treatments of seismic scattering based on correlations between randomly distributed elastic heterogeneities. This is because scattering is proportional to the derivative of fluctuations in material properties; for power-law-distributed velocity fluctuations in crustal rock, differentiation enhances scattering while correlation smoothing of fluctuations reduces scattering.
P224 DOWNHOLE ORBITAL VIBRATOR AS SOURCE FOR 1KM- OFFSET CROSSWELL SEISMIC WAVE-GUIDE SURVEYS 1 L. WALTER 1 and P. LEARY 1 1 Geospace Engineering Resources International 12750 S. Kirkwood Stafford Texas 77477 USA The downhole orbital vibrator (DOV) provides a stable acoustically-coupled linear-physics swept-frequency far-field seismic wavelet with frequencies in 50-350Hz band and a cylindrically-symmetric angular power-band at angles < 20 o -30 o from the plane normal to the host borehole. Tube wave generation is negligible. DOV-excited acoustic waves have a uniform displacement amplitude � 1-3� across the DOV axial cross-section. Because the acoustic wave motion is spatially limited
S‐waves generated by mode conversion of P‐waves from surface compressional sources and recorded by oriented geophones at depths between 1300 and 1820 meters in the Cajon Pass scientific drillhole arrive at higher apparent velocities when polarized along the direction N70°E than when polarized in the orthogonal direction. The observed polarization dependent velocities are consistent with seismic anisotropy due to in situ fractures or cracks that are preferen tially aligned N70°E (subnormal to the San Andreas fault). This result is consistent with the maximum principal horizontal stress direction at Cajon Pass also being subnormal to the San Andreas fault.
The feasibility of locating fracture zones and estimating their crack parameters was examined using an “areal well shoot” method centered on Utah State Geothermal Well 9‐1, Beaver County, Utah. High‐resolution traveltime measurements were made between a borehole sensor and an array of shotpoints distributed radially and azimuthally about the well. The directional dependence of velocity in the vicinity of the well was investigated by comparing traveltimes from different azimuths. Velocity anisotropy was detected; this condition is consistent with, but not required by, the existence of fractures in the basement rock having orientations subparallel to the latest episode of local faulting. The interpretation is complicated by a variable thickness of overburden around the well. Traveltime delays also suggest the presence of a low‐angle fracture zone intersecting the well at a depth of ∼1 500 ft.
