A Texas-specific [Formula: see text] map that uses geostatistical kriging integrated with a region-specific geologic proxy, field measurements of [Formula: see text], and P-wave seismogram estimates of [Formula: see text] is developed. The region-specific geologic proxy is used first to predict [Formula: see text] from the surface geologic conditions across the state, and then geostatistical kriging with an external drift is used to incorporate the local [Formula: see text] measurements/estimates into the map. Compared with the [Formula: see text] map of Texas developed from a topographic slope proxy, the Texas-specific [Formula: see text] map predicts larger [Formula: see text] values across much of Texas, except for the Gulf Coast region where the values are similar. The utilization of kriging brings the Texas-specific [Formula: see text] map into better agreement with the in situ measurements and estimates of [Formula: see text]. The sensitivity of predicted ground motions by ShakeMap to changes in [Formula: see text] values is evaluated with a scenario earthquake in the Dallas–Fort Worth area. The results suggest smaller predicted ground motions due to the generally larger values of [Formula: see text] in the Texas-specific [Formula: see text] map as compared to the [Formula: see text] from the topographic proxy.

No 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.