Ocean-Based Seismic Noise Sources Recorded by a Moderate Aperture Array in Antarctica

Abstract:

The deployment of a temporary, 60 km aperture, broadband seismic array on the Whillans Ice Stream (WIS), West Antarctica provides an opportunity to analyze ocean-derived seismic noise sources. The location of Antarctica, surrounded by the Southern Ocean and the seasonal effect of sea ice on shallow water noise production, allows for an intriguing experiment as to the production of primary and secondary microseisms. The WIS array was deployed for 2 months between December 2010–January 2011 with its primary objective to study WIS stick–slip events and glacial microseismicity. However, daylong stacks of station-to-station correlograms show directionality of the ambient noise field within the frequency bands of the primary and secondary microseisms. Although the WIS array is located close to the grounding line, it lies 700 km from the nearest open water at the end of the austral summer. The array consists of 17 broadband stations arranged in a series of offset concentric circles that minimizes spatial artifacts with regards to the array response. We use beamforming analysis to show that primary microseisms (~15 s) are sourced from three azimuthal directions with some ice-free coastline: Antarctic Peninsula, Victoria Land, and Dronning-Maude Land. Long-period secondary microseisms (~10 s) appear to be sourced in the deep Southern Ocean and track storm systems. Short-period secondary microseisms (~6 s) show much more dependence on the continental shelf and possibly coastal reflections. This is consistent with year-long noise spectra showing diminishment in the 15 s and 6 s bands [Grob et al. 2011]. Modeling of secondary microseism sources [Ardhuin et al. 2011] provides insight on the sources of surface wave noise at higher frequencies. We backproject daily P and PKPbc body wave microseism signals found at lower ray parameters sourced at distances of ~20–90° and ~145–155° respectively. The ocean sources for these arrivals remain fairly consistent, suggesting a bathymetric control, although the signal power increases with the passing of storm systems. We conclude that both 15 s primary microseisms and 6 s secondary microseisms at surface wave slownesses are generated in shallow coastal regions, but 10 s microseisms and noise with body wave slownesses are largely generated in deeper oceanic regions.