Within Station Variability and Uncertainty in Kappa Estimations: Insights From Various KiK-Net Downhole Arrays

Date: 4/24/2019

Time: 06:00 PM

Room: Grand Ballroom

The high-frequency spectral decay parameter Kappa (kappa_r) is one of the most used parameters to describe seismic attenuation at high frequencies, but it is still not well understood. There is a critical need to understand the physics behind the parameter, as well as potential factors that lead to large scatter in estimated values of kappa_r and its site component kappa_0. Such improvements would not only impact ground motion modeling, but the assessment of seismic hazards at large. However, the uncertainties in individual kappa_r estimations associated with different events at a selected site (which are hereafter referred to as the within station variability of kappa_r and kappa_0) remain uncharacterized and unquantified. Hence, robust estimates of kappa_0 and their interpretation remain a challenge. Recent research efforts have focused on studying epistemic uncertainties in kappa_r (and kappa_0) at specific site conditions (e.g., hard rock sites), and the variability in kappa_0 among different sites classes (e.g., the correlation between kappa_0 and the 30-m time-averaged shear wave velocity, Vs30). In order to understand the within station variability of kappa_r (and kappa_0), the uncertainties associated with individual kappa_r (estimated by the acceleration slope approach, kappa_r_AS) are categorized as site, path, and source components. Factors contributing to uncertainties in each component include: the type of seismicity, epicentral depth, directionality of the ground motion, and azimuths. In this paper, we use the Japanese Kiban-Kyoshin network (KiK-net) database to investigate how these factors influence the within station variability of kappa_r (and kappa_0). Finally, this paper provides recommendations for kappa estimations and record selection.

Presenting Author: Chunyang Ji

Authors