Large Scale Site Response variation in the Central and Eastern U.S.

Abstract:

Central and eastern United States (CEUS) is considered to be less seismically active compared with western United States; however, these few earthquakes are felt over much larger distances and have the potential to cause damages over a very large area. The earthquake ground motions vary due to the site conditions that effect strongly the propagation and localized amplification of regional seismic waves. Lg, is the most destructive regional phase for stable continental paths due to its large amplitudes due to low attenuation. The United States combined seismic network that includes EarthScope Transportable Array (TA), offer excellent station coverage for all of the CEUS. The availability of regional and local seismic records provided the resources to study and understand the characteristics of the site amplification of high frequency regional waves. The goal of this project to understand the physical characteristics of regional high frequency phases and the factors effecting these phases. We have used 40 earthquakes with magnitudes between 4 and 5 Mw and distances range between 20 – 2000 km. We collected approximately 8000 seismograms recorded on Transport Array (TA) component of USArray. We employed to methods to estimate the site effect: Nakamura technique (Nakamura, 1989), Horizontal over vertical spectral ratio (HVSR), applied on the shear wave part of seismic event and on the ambient noise. We also employed a Reverse two station method (RTS), which eliminates the source effect and includes the attenuation factor Q in calculating the site response (Chun et al., 1987). In order to correlate the site effect results, the shear wave velocity was calculated for each TA stations by modeling Horizontal/ vertical ratio of the first part of the P waves. The comparison between the two methods suggests we are able to estimate regional variations in site amplification of regional waves. Comparisons of this site amplification with Vs30m suggest that there is only a weak correlation with site amplification. The assumption that Vs30m provides a robust estimate of the site response was based on the data coming from western United States (Borcherdt, 1992, 1994); however, we do not see as good of correlation between site affect and Vs30.