← Back to Sessions

Development of a New Methodology for Evaluation of Site Fundamental Frequency and Its Associated Uncertainty Based on H/V Spectral Ratio Using an NGA-West2 Dataset

Session: How Well Can We Assess Site Effects So Far? IV

Type: Oral

Date: 4/21/2021

Presentation Time: 05:45 PM Pacific

Description: 

A new automated methodology is developed for estimating the site fundamental frequency (f₀), its corresponding amplitude (A₀), and their associated uncertainty using Horizontal-to-Vertical Spectral Ratio (HVSR) of ground motion records. Over the last decades, many studies have been carried out for developing ground motion prediction equations (GMPEs). In the GMPE, the local site effect is mainly addressed by V_S30. However, recent studies show that site fundamental frequency (f₀) is another useful proxy in reducing the uncertainties associated with the GMPEs. Using HVSR is an efficient and inexpensive measure to estimate f₀ as it does not require a reference rock site. It has been found that the peak in the HVSR curve is associated with f₀. In general, there are more than one peak in the site's HVSR curve. Therefore, peak selection relies on the analysts’ decisions while they are only using the average HVSR curve of the site, so a single number for the f₀ is estimated. In other words, f₀ might be estimated subjectively, and it does not incorporate uncertainty. This study uses selected sites in the NGA-West2 dataset to evaluate the performance of a new fully automated method in estimating the site characteristics (i.e., f₀ and A₀). In addition to using the site's average HVSR curve, individual HVSR curves from the recorded events are also considered for estimating the site proxies. In this methodology, the average HVSR curve is used to find a rough estimation of the site fundamental frequency (f_0,m), which is assumed to be the lowest frequency peak in the average HVSR curve. Then, the peaks in individual events that are more consistent with f_0,m are selected as flagged peaks. In the next step, the statistical parameters (i.e., mean, standard deviation) are calculated using the flagged peaks. The whole process is programmed in a way to be completely automated, hence unbiased. Overall, this study presents this new automated methodology for estimating the site fundamental frequency, its corresponding amplitude, and their associated uncertainty using HVSR.

Presenting Author: Mohammad Yazdi

Student Presenter: Yes

Authors