Empirical Horizontal Site Amplification Factor (HSAF) from Observed Earthquake Horizontal-to-Vertical Ratio (EHVR) and Vertical Amplification Correction Function (VACF)

Session: Near-Surface Effects: Advances in Site Response Estimation and Its Applications
Type: Oral
Date: 4/28/2020
Time: 09:30 AM
Room: 110 + 140
Description: 

It is important to deduce the horizontal site amplification factor (HSAF) in the main S-wave part for quantitative evaluation of strong motion. When we use an empirical method such as the method of spectral ratios for soil sites to a near-by reference rock site, there is an issue of whether the reference site can really be a site without site amplification, that is the seismological bedrock.

Using the generalized spectral inversion technique (GIT), Nakano et al. [BSSA, 2015] obtained HSAFs with respect to an appropriately corrected reference site as an outcrop of the seismological bedrock for the date of K-NET, KiK-net and JMA network in Japan. We can refer to these HSAFs by GIT. In order to expand the empirical evaluation method for HSAF, we focus our attention on the horizontal-to-vertical spectral ratio of earthquakes (EHVR) since we only need three components of earthquake ground motions at a single station. Although Nakamura [RTRI, 1989] proposed that EHVR can be a direct substitute of HSAF under the hypothesis of no vertical amplification, our GIT results for vertical component revealed that there exists a reasonable amount of the vertical site amplification factor (VSAF) at almost all the sites.

We propose here a simple method to estimate HSAF from EHVR. As vertical amplification is included in the denominator of EHVR, the ratio cannot be viewed as HSAF. We deduced the average VSAF from a total of 1,678 sites where ten or more earthquakes were observed. We called the resultant empirical factors as the vertical amplitude correction function (VACF). VACF can be categorized for peak frequency and amplitude of EHVR, but VACF averaged over all the sites can also be used for simplicity. By multiplying EHVR with VACF, we can easily obtain HSAF. We have verified the effectiveness of this approach by using the sites not used in the averaging. On the average 75 % sites can be successfully reproduced by the proposed method and 15 % sites do not show large HSAF. We need to carefully investigate the remaining 10 % sites why their VSAFs are not close to the average VACF.

Presenting Author: Hiroshi Kawase

Authors