Characterising ice sheet properties using Rayleigh wave ellipticity

The physical properties of the ice column are fundamental to the deformation and flow of glaciers and ice sheets. With a warming climate, surface meltwater is ever increasingly being routed and distributed throughout the ice column changing the mechanical and hence thermal properties of the ice and leading to accelerated ice flow and ice mass loss. Since the early 1990s, ice mass loss from the Greenland Ice Sheet (GrIS) has contributed ~10% of the mean global sea level rise. Seismic waves have routinely been used to study the physical characteristics of glaciers and ice sheets due to their sensitivity to both mechanical and thermal properties of ice. Traditionally, reflection seismic surveys have been chosen as the primary seismic approach but this survey method can suffer from difficult logistics in polar regions. Recent advancements in ambient noise methods and the permanent installation of a seismic network in Greenland now permit the long term study of the ice properties of the GrIS.

Rayleigh wave ellipticity measurements (the horizontal-to-vertical ratio of Rayleigh wave particle motions) are particularly sensitive to the subsurface structure beneath a seismic station. Using the polarisation properties of seismic noise, we extract Rayleigh wave ellipticity measurements from the Earth’s ambient noise for on-ice stations deployed in Greenland from 2012-- 2018. For wave periods sensitive to the ice sheet (T ≤ 3.5 s), we observe significant deviation between ellipticity measurements extracted from noise and synthetic fundamental mode calculations using a single ice column. Using a forward modelling approach we show: (1) a slow seismic shear-wave velocity at the near surface, (2) seismic attenuation, quantified as the quality factor Q, is sensitive to the temperature, water content and density of the ice and (3) the excitation of Rayleigh wave overtones plays a leading role in perturbing the ellipticity. Our results highlight how the inclusion of Q and overtone information can fill important gaps in our knowledge of ice sheet temperature, density and water content, which are important for predictions of the future evolution of the GrIS.