Detection and localization of ice cavitiy using ambient seismic noise

Subglacial cavities may trap a considerable quantity of liquid water, causing devastating outburst floods in densely populated mountain areas. Dedicated studies aimed at identifying such intraglacial cavities at an early stage of their formation (1-2) to prevent and mitigate potential subsequent hazards. Both active and passive geophysical methods are employed for the glacier-bedrock interface and intra-glacial characterization e.g., (3), including Ground Penetrating Radar (GPR), refraction seismic, borehole measurements, and surface nuclear magnetic resonance (SNMR). 

Ambient seismic noise can be collected by light surveys at a relatively moderate cost, and allows to access some mechanical properties of the glacier, including the detection and localization of ice cavities. The horizontal-to-vertical-spectral ratio (HVSR) technique is highly sensitive to impedance contrasts at interfaces, especially the ice/bedrock interface, thus allowing to estimate the glacier thickness (but with limited resolution compared to GPR).

In contrast to the classical Horizontal to Vertical Spectral Ratio (HVSR), Saenger et al. (4) proposed analyzing the (opposite) V/H spectral ratio (VHSR) for spectral anomalies characterization. Specifically, a peak in the VHSR indicates a low impedance volume beneath the surface. As a simple picture, we can refer to the “bridge” vibrating mode, where the vertical displacement in the middle of the bridge largely dominates other components of the movement.  Antunes et al. (5) furthermore noticed that the VHSR gives information about seismic energy anomalies generated by fluids in reservoirs since the wavefield is polarized mainly in the vertical direction.

In this work, we apply the HVSR and VHSR techniques to characterize the Tête Rousse glacier (Mont Blanc area, French Alps) and a subglacial water-filled cavity. We analyze the HVSR and VHSR results from 60 temporary dense seismic array installed on the glacier for 15 days (May 2022). Mapping the VHSR over the free surface evidences areas where the main cavity (or secondary cavities) is (are) expected. We perform an elastic modal analysis based on numerical simulations obtained with Comsol Multiphysics finite element numerical scheme to reproduce the observed field data and confirm some geometrical and physical features of the cavity(ties).

References:

• (1) Haeberli, W. et al: Prevention of outburst floods from periglacial lakes at Grubengletscher, Valais, Swiss Alps. Glaciol., 47 (156), 111–122 (2001).
• (2) Vincent, C. et al : Origin of the outburst flood from Glacier de Tête Rousse in 1892 (Mont Blanc area, France), Journal of Glaciology, 56 (198), pp 688–698 (2010).
• (3) Petrenko, V. F, and R.W. Whitworth: Physics of ice. Oxford University Press, New York, 373 (2002).
• (4) Saenger, E-H. et al: A passive seismic survey over a gas field: Analysis of low-frequency anomalies, Geophysics, 74 (2), O29–O40 (2009).
• (5) Antunes V. et al: Insights into the dynamics of the Nirano Mud Volcano through seismic characterization of drumbeat signals and V/H analysis. Journal of Volcanology and Geothermal Research, 431 (2022).