TOPOGRAPHIC SITE RESPONSE AT HARD ROCK SITES

Site (material impedance) and topographic (geometric form) effects are known to be key factors that influence seismic ground motions. To characterize site effects, Yong et al. (2009) developed a terrain-based Vs30 prediction model using an automated classification method (Iwahashi and Pike, 2007) that relied on taxonomic criteria (slope gradient, local convexity and surface texture) developed from geomorphometry to identify 16 terrain types from a 1-km spatial resolution (SRTM30 data) digital elevation model of California. On the basis that the underlying framework for this model contains parameters (esp., local convexity) which aptly describe the geometry (i.e., base to height ratio) of relief features that are known to also control the behavior of ground motions (Bouchon, 1973), we extend our investigation to study topographic effects. Focusing on sites that would generally be considered “hard rock,” the classification scheme distinguishes 7 separate terrain types ranging from “moderately eroded mountains” to “well dissected alpine summits.” Observed 1-Hz amplification factors at Southern California Seismographic Network sites reveal a weak but systematic correlation with these 7 terrain types. Significant scatter is also found within each terrain type; typical standard deviations of logarithmic amplification factors are 0.2-0.3. Considering stations that have high amplification factors, we find some that have apparently been misclassified due to data resolution limitations. Many of the remaining stations with higher than expected amplifications are located on or near topographic peaks or ridges. The unusually high amplification factors at hard-rock sites, typically factors of 1.5-2, can most plausibly be explained as a topographic effect.