A tomographic inversion of the Pn arrivals in Southern California yields new information about wave velocities and topography on the Moho discontinuity. We produce maps of Pn velocity and Pn station delays. The Pn velocities do not show the dramatic correlation with surface faults that is found for the shallower Pg arrivals (Hearn and Clayton, 1986). This implies that the lower crust and mantle are largely decoupled from the upper crust. Undoubtedly, this is due to the different responses of the brittle upper crust and the ductile lower crust to tectonic and isostatic stresses. Detachment faults must play an important role in separating the crust.
In general, velocities on the American plate are higher than on the Pacific plate, but no distinct transition is observed. The Colorado River region has extremely thin crust due to basin-and-range type extension. The Transverse Ranges have a small root as seen in the station delays and which also results in slightly lower Pn velocities there. The Peninsula Ranges also have slow Pn velocities, but they do not have late station delays. Any root to the Peninsula Ranges must be very narrow. Isostatic balance must be maintained primarily through lateral density contrasts.