Estimation of Kappa (κ) for Rock Sites in the Nga-East Database and Implications on Design Motions

Abstract eng:
At high frequencies, the amplitude of the Fourier amplitude spectrum (FAS) of acceleration decays rapidly. Anderson and Hough (1984) [1] introduced the spectral decay factor (κ) to model the rate of decay in log-linear space. Its site-specific component, κ0, comes from damping over the top few km of the crust. κ0 is an important input parameter in the simulation and prediction of ground motion. It is the principal site parameter controlling the limitation of high frequencies (>5 Hz) at close-in distances (out to ~50 km). Thus, its range of values is important in characterizing strong ground motions for engineering design, particularly in regions of sparse seismicity. Current uncertainty in the estimation of κ0 is very high (Ktenidou et al., 2014 [2]). In practice, this can have significant implications on seismic risk: 1) for safety-related equipment in nuclear facilities, and 2) for the seismic behavior of small concrete dams. In October 2014, the NGA-East project published a metadata flatfile for data from Central-Eastern North America (CENA). Those data are important for the creation or updating of current ground motion prediction models (GMPEs). We estimate κ0 for CENA focusing on rock sites (Vs30>1000 m/s) at short distances (<100 km), so as to avoid path attenuation. We use two band-limited methods to estimate κ0, the acceleration spectrum approach (AS) and the displacement spectrum approach (DS) as summarized in [2], and compare the two approaches. Due to the lack of seismic records on hard rock sites, site effects for hard-rock sites are typically computed using analytical models for κ0. In current practice and consistent with its original definition, it is assumed that all of the measured κ0 is due to attenuation beneath the site. We raise two issues: 1. the idea that new datasets that are richer in hard-rock recordings will allow us to evaluate the scaling for hard-rock sites (soft-to-hard-rock correction factors), and 2. the concern that the total κ0 measured from records may incorporate additional effects that are not directly related to attenuation (e.g. site-specific shallow resonance patterns), which may affect the apparent measured values of κ0. We use the NGA-East dataset as an example of a new dataset rich in hard-rock data, and study residuals with GMPEs. The high-frequency response spectra residuals are weakly correlated with κ0, in contrast to the strong scaling with κ0 in the analytical models. We propose that this may be due to amplification marking attenuation at high frequencies, causing overestimation of the actual attenuation effect on ground motion, and leading to possible (unconservative) overcorrections in GMPE adjustments in current practice. The coupling of site amplification and site attenuation may become a key limitation in measuring and interpreting κ. For this reason, an empirical model is developed for the combined Vs30 and κ0 scaling for hard-rock sites relative to a reference site condition of 760 m/s. This model shows high-frequency amplification that is more similar to the analytical prediction corresponding to a hard-rock κ0 of 0.020 s rather than the typical value of 0.006 s commonly used for hard-rock sites in the Central-Eastern US. This leads to a reduction of high-frequency scaling of about a factor of 2, compared to the traditional analytical approach.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket