Key points are not available for this paper at this time.
This paper attempts to construct a preliminary shear velocity model compatible with both travel times and wave forms of observed seismograms. Transversely polarized long-range seismic measurement and World-Wide Standard Seismograph Network observations are used in this study. Vertical secondary wave signatures are demonstrated to be inadequate in identifying triplications because of contamination by PL coupled shear waves and other P-SV interactions. Well-located west coast earthquakes, modeled as shear dislocations, are used as sources. The observed SH travel times are considerably slower than the Jeffreys-Bullen values out to 30°. Select profiles of observed wave forms in conjunction with the travel times provide the data for model determinations by fitting the observed wave forms with synthetics. Synthetics are computed for a number of current models as preliminary attempts at fitting the data. Models containing large sharp transitions predict strong second arrivals beyond 30°. Models containing low-velocity zones between 400 and 700 km predict complicated wave shapes between 20° and 25°. None of the above features are apparent in observed SH wave forms. The final model is relatively smooth and exhibits the same characteristics as the HWB P model (Wiggins and Helmberger, 1973), except that the percentage velocity jump is much less at 400 km and slightly larger at 500 km. The 600-km transition is subdued similar to that of the P model.
Helmberger et al. (Tue,) studied this question.