Regional variations in the crustal structure of southern California are defined by travel-time data from natural and artificial events.We show that the crust of the Mojave, northeastern Peninsular Ranges, eastern Transverse Ranges and Colorado Desert is dominated by a velocity of 6.2±0.1 km/sec. The western Transverse Ranges and the western portion of the Peninsular Ranges are typified by a crustalvelocity of 6.7 km/sec. The data indicate that the Transverse Ranges do not have a distinct crustal root. As the topography is not supported isostatically, the Range must be sustained by major north-south compression. A composite profile extending north from the southern end of the Salton Sea defines a crustal thickness for theCoachella Valley of less than 20 km. Through the inversion of Rayleigh wave dispersion data obtained from the analysis of teleseismic surface waves recorded across southern California, we have obtained average S-wave models for the southern Mojave-central Transverse Ranges and the Peninsular Ranges. The observed P-wave velocities and the calculated Poisson's ratio from both P- and S-wave data require a quartz rich crust for the Mojave and a more mafic crust for the Peninsular Ranges.All S-wave models suggest a slight mid- crustal velocity reversal that is approximately coincident with the bottom of the seismiczone.
Regional variations in P_n velocities are obtained from several reversed refraction profiles . These data show that P_n varies from 7.7 to 8.2 km/sec. The high P_n values, 8.2 km/sec, are observed in the eastern Mojave, the western Transverse Ranges and the Coast Ranges.The 7.8 km/sec P_n velocity extends from the Imperial Valley, through the central Transverse Ranges, and across the western Mojave. P_nprofiles indicate that the Moho beneath the eastern Transverse Ranges and the southeastern Mojave dips 2-3° west.
P-delay studies of a vertically incident PKP phase indicate that a high velocity, 8.3 km/sec structure exists within the shallow upper mantle beneath much of the geomorphic Transverse Ranges. This feature is not offset by the San Andreas fault. We suggest that the continuity of this anomaly across the plate boundary indicates that if the upper mantle participates in plate motion, the mantle plate boundary must be laterally displaced from the crustal boundary. We suggest that the mantle boundary may extend northwest from the Salton Trough and across the eastern end of the velocity anomaly, in the vicinity of the active Helendale-Lenwood-Camprock faults. We propose that thehorizontal decoupling between the crust and mantle, required by the lateral displacement at depth of the plate boundary, is accommodated, in part, within the 7.8 km/sec layer.
