GSA Annual Meeting in Phoenix, Arizona, USA - 2019

Paper No. 62-3
Presentation Time: 2:00 PM


PAVLIS, Gary1, ATTERHOLT, James1 and BROWNLEE, Sarah J.2, (1)Department of Earth and Atmospheric Sciences, Indiana University, 1001 East 10th St., Bloomington, IN 47405, (2)Department of Geology, Wayne State University, 0224 Old Main, Detroit, MI 48208

A fundamental issue for any seismic experiment is the recording density necessary to sample the wavefield adequately. Since the 1950s, it has been known that the weathered layer impacts all surface data and can limit wavefield coherence at high frequencies. Seismic data collection in underground mines provides a way to collect data unaffected by the weathered layer. We analyzed data at the Sanford Underground Research Facility (SURF) located in the old Homestake Mine in the Black Hills of South Dakota. The experiment used a 92 m long profile of three-component sensors spaced at 4 m intervals deployed at the 4850-level of SURF. We focused on two shot profiles with a vertical hammer source fired at 1 m intervals along two profiles, one perpendicular to the array and the other slightly oblique to the array. Unaliased records show strong variations from sensor-to-sensor indicating surprising amounts of scattering given the geologic setting (phyllite to garnet bearing schist of the Homestake and Poorman formations). We find clear evidence that the rocks are strongly anisotropic. We measured velocity variations with azimuth from a fan shot profile and found the variations were well matched by a model based on lattice preferred orientation of a generic schist. The model used a Christoffel solution for an elastic tensor derived from electron backscatter diffraction data collected from an analogous garnet-biotite-muscovite schist (qtz: 32%, plag: 22%, bio: 19%, mus: 22%, gar: 6%), where the tensor was rotated into the average foliation orientation for the Homestake mine (strike: 334.8 dip: 67.3). We also measured P wave particle motion from the longitudinal direct wave using three different techniques and found large, frequency-dependent variations. Variation magnitude also seems to be dependent on the length of the sampling window, with shorter sampling windows displaying higher particle motion deviations (as much as 50 degrees). The particle motion deviations show an oscillatory pattern consistent with the model fit to the slowness data, but with a larger range of deviation values than the model predicts. The rocks at the site are best characterized as strongly anisotropic with frequency dependent scattering likely induced by structure smaller than a wavelength and/or fracturing linked to the mine drift.