An evaluation of empirical earthquake record-based versus array-based site characterization
Tara NyePublished September 11, 2022, SCEC Contribution #12221, 2022 SCEC Annual Meeting Poster #223
We compare site empirical transfer functions (ETFs) estimated from earthquakes recorded at stations of interest with active, array-based theoretical transfer functions (TTFs) to evaluate the value of including less-intensive studies for characterizing local site conditions. Seismic site characterization often involves active- or passive-source array-based field surveys to estimate shear-wave velocity (VS) profiles which can be costly, arduous, and time consuming. Furthermore, VS profiles obtained from these surveys are non-unique, and necessary approximations used to transform the survey into TTFs (layer density and damping) introduce added uncertainties that undermine the initial measurement precision. We aim to evaluate the utility and effectiveness of performing single-station earthquake record-based site characterization in regions with existing high seismicity by comparing ETFs obtained using earthquake records with TTFs obtained using array-based methods. We compute transfer functions for 10 sites within the Anza (AZ) and USGS/Caltech (CI) seismic networks, located near the Imperial Valley. Our array-based dataset consists of VS profiles and the time-averaged VS of the upper 30 m from the surface (VS30) measurements for each site from Yong et al. (2013). We use these profiles in equivalent linear site-response analysis where ground motion is propagated through 1D soil columns to obtain TTFs. We also examine TTFs from proxy VS30-based soil profiles (Shi and Asimaki, 2018) for comparison. For the earthquake record-based approach, we use a dataset consisting of over 1,000 broadband and accelerometer recordings for the 10 sites of interest from earthquakes ranging in moment magnitude 3.5≤ M ≤5.5 and occurring between 2008-2022. We apply a modified version of the Klimasewski et al. (2019) spectral decomposition approach to isolate the site spectra and use station AZ.KNW as a reference site to obtain the ETFs. We present a comparison of these ETFs with the array-based and proxy VS30-based TTFs by evaluating their frequency peaks and average spectral amplitudes at various frequency bins. We also correlate the spectral amplitudes from each method with ground-motion model site residuals to evaluate their utility for ground motion analyses in the region.
Key Words
site response, transfer function, Imperial Valley
Citation
Nye, T. (2022, 09). An evaluation of empirical earthquake record-based versus array-based site characterization. Poster Presentation at 2022 SCEC Annual Meeting.
Related Projects & Working Groups
Ground Motions
