Evaluation of the P-wave seismogram approach to estimate VS30 in California

Abstract

The P-wave seismogram method is utilized to estimate the V S30 at 194 ground motion recording stations in California. Comparison with the measured V S30 at these sites shows an average overestimation of 9%, which is similar to values reported by other studies for the P-wave seismogram method. The estimated V S30 are within ±50% of the measured V S30 for 85% of the sites. The V S30 residuals are interpreted with respect to the peak frequency obtained from horizontal-to-vertical spectral ratio (HVSR) measurements, and the results indicate that stations with estimated V S30 greater than 500 m/s may be systematically overestimated by the P-wave seismogram method when a distinct peak is identified in the HVSR. Thus, the peak frequency from HVSR can be used to identify sites with potential bias and also to correct this bias. The P-wave seismogram V S30 estimates from this study are compared to estimates from a geology/slope proxy, and the results suggest that the P-wave seismogram method achieves similar accuracy and less variability relative to the measured V S30 compared with the geology/slope proxy-based method.