Monitoring Submarine Glacier Melt Using Hydroacoustics: The Role of Timescale in the Signal of Bubble Release

Abstract

With global temperatures rising, glaciers are losing mass at substantially elevated rates (Intergovernmental Panel on Climate Change, 2014). One of the main processes by which glaciers lose mass is submarine melt. This process plays a key role in determining glacier stability and driving other mass loss processes, but direct, near-terminus measurements cannot be taken due to occupational hazards (Luckman et al., 2015; Deane et al., 2019). Current methods of estimating submarine melt using far-field measurements, such as ice-ocean thermodynamic modeling, are costly and have large associated uncertainties (Sutherland et al., 2019). Passive acoustics, or the monitoring of the ambient underwater soundscape, is a novel method that has been proposed for studying tidewater glaciers from afar (Schulz et al., 2008). The goal for future acoustic monitoring systems is to use the total acoustic energy produced from bubbles bursting out of glacier ice to determine the number of bubbles that were released in a given amount of time. Then, using the known density of bubbles in glacier ice, melt rates can be determined. However, these calculations are complicated by the fact that different bubbles produce different amounts of acoustic energy (Deane et al., 2019). This study advances the literature on passive acoustic monitoring by investigating the physical factors that cause bubbles to produce energy heterogeneously. Acoustic data from melting glacier ice in Svalbard, Norway was analyzed to see if the timescale over which a bubble is released from the ice impacts the amount of acoustic energy that it produces. Timescales of release and acoustic energy values were determined for over 200 bubble release events. Two types of events were found in the data: low energy and high energy events. For low energy events, timescales were determined solely based on the bubbles’ acoustic signals. For high energy events, a numerical model of bubble release was developed and timescales of release, initial internal pressures, and energy conversion efficiency values were determined from this model. Timescale of release and acoustic energy were shown, overall, to be negatively related. However, the two different types of events formed two distinct clusters of data points, each with their own unique trends. Additionally, for the high energy release events, it was determined that timescale of release was linearly related with neither initial internal pressure nor energy conversion efficiency. This work identifies timescale of release as one of the physical mechanisms controlling varying acoustic energy production for glacier ice bubbles and highlights the need for further research into the environmental factors controlling different bubble release mechanisms.