Seismic noise interferometry and Distributed Acoustic Sensing (DAS): measuring the firn layer S-velocity structure on Rutford Ice Stream, Antarctica

Abstract

Distributed acoustic sensing (DAS) is a rapidly growing seismic technology, which provides near-continuous spatial sampling, low maintenance, long-term deployments, can exploit extensive cable networks already deployed in many environments. Here, we present a case study from the Rutford Ice Stream, Antarctica, showing how the ice-sheet firn layer can be imaged with DAS and seismic interferometry, exploiting noise from a power generator and fracturing at the ice stream margin. Conventional cross-correlation interferometry between DAS channels yields an unstable seismic response. Instead, we present two strategies to improve interferograms: (1) combining signals from conventional seismic instruments with DAS; (2) selective-stacking crosscorrelation. These steps yield high-quality Rayleigh wave responses. We validate our approach with a dataset acquired using a sledgehammer-and-plate source, and show an excellent agreement between the dispersion curves. The passive results display a lower frequency content (˜3Hz) than the active datasets (˜10Hz). A 1D S-wave velocity profile is inverted for the top 100m of the glacier, which contains inflections as predicted by firn densification models. Using a triangular DAS array, we repeat the noise interferometry analysis and find no visible effect of seismic anisotropy in the uppermost 80 meters of our study site. Results presented here highlight the potential of DAS and surface wave inversions to complement conventional refraction surveys, which are often used for imaging firn layer, and the potential in near-surface imaging applications in general.

Seismic Noise Interferometry and Distributed Acoustic Sensing (DAS): Inverting for the Firn Layer S‐Velocity Structure on Rutford Ice Stream, Antarctica

“…Seven hours of continuous DAS data (decimated to 10 m channel distance and 100 Hz sampling rate), and continuous geophones data (3 collocated geophone, A000, R102, R104, vertical component) have been made available through Zenodo (Zhou et al., 2022), which could be used to reproduce the SI and surface wave inversion for S ‐wave velocity profile.…”