Estimation of EMR Correction Factor in the Grenoble Basin; an Attempt to Establish a Simple Method to Get Earthquake HVR From Microtremors

Date: 4/26/2019

Time: 09:30 AM

Room: Pine

It is quite important to get velocity structures for site amplification evaluation. Kawase et al. (2018) proposed a method to calculate pEHVR (pseudo earthquake horizontal-to-vertical spectral ratios) from HVR of microtremors (MHVR) and EMR, which is the spectral amplitude ratio between HVR from earthquakes (EHVR) and MHVR at one hundred sites in Japan. They calculated EMRs for five categories based on their fundamental peak frequencies in MHVR. They found that pEHVR is much closer to EHVR than MHVR. They used their inversion code to invert one-dimensional S-wave velocity structures from EHVR based on the diffuse field theory (Nagashima et al., 2014). They found that velocity structures by pEHVR are much more similar to those by EHVR in comparison to those by MHVR.

However, when we apply this method to the data in other countries, EMR in Japan may not be directly applicable, because EMR should be a function of the velocity structures from the bedrock to the surface which may not be the same. If we want to calculate EMR in other countries in the same way as Kawase et al. (2018) did, we need to collect sufficient amount of data, which would not be easy in seismically not so active regions. This study is the first step to establish an easier way to get EMR correction factors outside Japan.

We calculated pEHVR at the earthquake observation sites in the Grenoble basin using the observed MHVR and EMR in Japan to check the difference between pEHVR and observed EHVR. We found that pEHVR seemed to be overestimated in almost all the frequency higher than the fundamental peak frequency.

In order to make pEHVR closer to the observed EHVR, we assumed EMR specific for the Grenoble basin depends on EMR in Japan and a modification factor α. Then we found α so that the logarithmic residuals of amplitude between pEHVR and observed EHVR became the minimum value at each station. After we got the optimal α to be around 0.3, new pEHVR with the optimal α have quite good match with the observed EHVR in a wide frequency range.

Presenting Author: Eri Ito

Authors