Bias and Uncertainty of Depth Parameters Extracted From a Regional Velocity Model and Its Implications on Ground Motion Prediction

Date: 4/26/2019

Time: 11:45 AM

Room: Pine

Site depth is utilized as a site proxy, in addition to the shear-wave velocities averaged over the first 30 m, VS30, to parameterize a site in ground motion models and seismic regulations, e.g., Next Generation Attenuation of Ground Motions models and the revised European building code EC8. However, site-specific depth measurements are often unavailable at sites of interest. Considering that subsurface velocity models have been developed for some regions, e.g., Japan and California, can site depth extracted from these structural models be utilized instead? To address this issue, we examine a subsurface model for Japan, the J-SHSIS model, proposed by the National Research Institute for Earth Science and Disaster Prevention (NIED). We firstly identity KiK-net sites with reliable velocity profiles of which empirically-derived site periods are consistent with theoretical ones calculated from the velocity profile. Then we establish site-specific depth measurements from these reliable velocity profiles, and then we compare these depth measurements with depth data extracted from the J-SHIS model. The comparison indicates that the J-SHIS depths are fraught with large bias and uncertainty. The J-SHIS model underestimates site depths at shallow sites and overestimates depths at deep sites, which can have a potentially significant impact on ground-motion prediction, e.g., overestimation of ground motion by a factor up to 1.4. Besides, our results also show that correlation models for estimating depth parameters from VS30 have considerably large scatter. Therefore, one needs to be cautious about inferring site depth from a regional velocity model or from a known VS30.

Presenting Author: Chuanbin Zhu

Authors