UCLA

The time-averaged shear-wave velocity of the upper 30 m of the crust (Vs30) is the primary parameter used to predict local ground-motion amplification. When measurement-based methods for determining Vs30 are not available, proxy-based prediction methods are employed instead. The principally used proxy for Vs30 prediction is based on topographic slope, although there have been developments in hybrid slope-geology proxies in recent years. While slope exerts a first-order control on Vs30, there are still considerable deviations between measured Vs30 and predicted Vs30 from the slope proxy, especially in rock sites. In this study, we compiled 218 sites of measured Vs30 in southern California to examine whether the use of high-resolution DEMs or combined topographic metrics of slope, curvature, relief, and distance to the San Andreas Fault can produce better predictions of Vs30. We find that positive residuals, or faster-than-predicted Vs30 from the slope proxy, tend to lie in valley landforms for rock sites. Our results indicate that Vs30 is best predicted with the single metric of slope calculated from a 30 m resolution DEM smoothed over a radius of 900 m. However, Vs30 is also well-estimated using a multivariable model combining slope calculated from a 30 arc-second DEM with curvature calculated smoothed over radii of 700 m and 900 m from a 30 m DEM. The influence of curvature on Vs30 may be a result of bedrock weathering processes, which predict thicker weathered zones beneath ridgetops and thinner weathered zones beneath valleys. Therefore, ridgetops are expected to correspond with thicker slow-velocity zones, while valleys are expected to correspond with thinner slow-velocity zones.