Vs30 From Multi-Method Site Characterization Approach at Seismograph Locations in the Fairview, Oklahoma Region

To better constrain ground motion models for induced seismicity in the Oklahoma region, we acquired active- and passive-source seismic data for time-averaged shear wave velocity to 30 m depth (Vs30) at 13 seismograph locations near Fairview, Oklahoma. Each seismograph had recorded at least one M4+ earthquake within 20 km hypocentral distance. We collected active-source P- and SH-wave body and surface wave data on 72-sensor linear arrays, with collocated 4.5 Hz vertical and single-component horizontal sensors. The sensors were spaced at 1.5 m with off-end source locations, using a 4.5 kg sledgehammer source, for each dataset. In addition, we collected linear-array passive refraction microtremor (ReMi) data with vertical sensors. The active source data were analyzed for P- and SH-wave refraction traveltimes and Rayleigh- and Love-wave phase velocity dispersion. The resulting observations were then modeled in a nonlinear least-squares joint inversion to obtain a best-fit one-dimensional shear wave velocity-versus-depth profile for each site. The ReMi passive-source data were independently forward-modeled for comparison. At many of the sites the preferred joint inverse model did not optimally fit each of the Love, Rayleigh and SH traveltime datasets; therefore, two of the three sets were typically fit preferentially. Taken as a whole, we estimate five sites to be NEHRP class D while the remaining eight are class C. Site amplification estimates were next calculated from 1D SH transfer functions (program "nrattle"). The active-source models suggest amplifications up to 7 within the 2-7 Hz frequency band relative to modeled bedrock depth and Vs at a given site. The ReMi models tend to be less variable in Vs30 but also indicate amplification in the 3-6 Hz range. Horizontal-to-vertical spectral ratios from the seismograph data are also amplified in the 2-10 Hz range. The Vs30 values and site classes are broadly comparable with estimates from other data sources in this area.