RELATIONS BETWEEN ELASTIC WAVESPEEDS IN QUATERNARY ALLUVIUM IN THE SANTA CLARA VALLEY, CALIFORNIA

Deep borehole logs in the Santa Clara Valley provide an important test of relations between elastic wavespeeds in Quaternary alluvium. There, and elsewhere, strong ground motion calculations require 3-D models of compressional-wave velocity (Vp), shear-wave velocity (Vs), and density. For such models Vs is more important than Vp because strong ground motions are generated chiefly by shear and surface wave arrivals. Borehole logs, and seismic tomography and refraction studies, however, often report only Vp so a direct relation between Vp and Vs is highly desirable. We compare empirical Vp/Vs versus Vp curves to Vp and Vs measurements obtained in five suspension logs sampling Quaternary alluvium in the Santa Clara Valley (CCOC, GUAD, MCGY, STGA, and STPK). For these wells Vs predicted from the observed Vp logs using the so-called “mudline” curve reported by Castagna and others (1985), Vs (km/s) = (Vp - 1.36)/1.16, match the observed Vs logs to within an average standard deviation of 100 m/s. The predicted Vs logs provide a close fit to the peaks and troughs on the observed logs that respectively correspond to coarser and finer grained units within the alluvium. These close fits were achieved using Vp suspension logs starting at about 40 m depth to total well depths of 410 m, where Vp is larger than 1.49 km/s. We also compared the observed Vs logs to Vs predicted from the observed Vp logs using an empirical relation (Ludwig et al., 1970) for Poisson's ratio = -0.0014Vp³ + 0.0316Vp² - 0.226Vp + 0.769, where Vp is in units of km/s. This relation consistently overestimates the observed Vs by an average of 96 m/s at depths greater than 40 m. We therefore use the Castagna and others “mudline” to calculate Vs from Vp for several sonic logs in the Santa Clara Valley for which we have only Vp measurements (Palo Alto (ELNR), EVGR, MOIT, and WILLOW).