This study reanalyzes a refraction seismic profile from NTS, Nevada, to San Francisco transverse to the high Sierra Nevada. Three models of the Earth's crust, with an asymmetrical mountain root, constructed by different authors are tested by the use of ray tracing: model CQM (Carer et al., 1970), with a maximum crustal thickness of 35 km centered about 50 km west of the crest of the Sierra; models A and B (Pakiser and Brune, 1980), with a mountain root extending to a depth of 55 km below the Sierra. In the latter, guided and diffracted waves are used to explain seismic Pn energy in the shadow zone of the root. The preferred model A has a high-velocity layer, dipping from west to east in the upper crust, while model B does not.
The present study demonstrates how fallacies in interpretation of complicated structures can occur if all ray types are not quantitatively traced. The influence of the P-wave gradient in the upper mantle on the length of the shadow zone is shown to be important, and it is demonstrated that a modified CQM model explains the observed travel times when diffracted waves are included. (CQM does not, however, explain why the Pn travel times on a separate refraction profile along the Sierra Nevada from a Truckee earthquake are so great.) Model B interprets the Pn signals at Δ > 200 km as diffracted waves, but the computed travel times do not fit the observations. The dipping wave guide in model A does not work as expected and perhaps should be ruled out.