Oskar Glowacki – oglowacki@igf.edu.pl
Institute of Geophysics, Polish Academy of Sciences
Ksiecia Janusza 64
01-452 Warsaw, Poland

Popular version of paper 2pAO6 “An acoustic study of sea ice behavior in a shallow, Arctic bay”
Presented Monday afternoon, June 26, 2017
Session in Honor of David Farmer
173^rd ASA Meeting, Boston

Glacial bays are extremely noisy marine environments, mainly because of the melting of marine terminating glaciers [1-3]. Tiny air bubbles bursting explosively from the ice during contact with warm ocean waters are responsible for these signatures. One of the most noisy and spectacular phenomena are also detachments of large icebergs at the ice-ocean boundary, called glacier calving events [4-5].

Both processes are particularly active during warm conditions in the Arctic summer and early autumn. When the air temperature drops, the water cools down and after some time a thin layer of sea-ice appears. But even then, the underwater environment is not always a quiet place. Researchers found it a few decades ago during field measurements far in the north.

A large number of acoustical studies concerning sea-ice processes appeared in the 1960s. Results from field campaigns clearly showed that underwater noise levels recorded below the ice depend strongly on environmental conditions and the structure of ice itself. For example, sea-ice cover cracks during abrupt decrease in air temperature and deforms under the influence of wind action and ocean currents [6-8].

The noise levels measured in winter were often similar to those observed at open sea with wave heights reaching up to 1.25 meters [6]. Conversely, when the ice is strongly consolidated and thick enough, recorded noise levels can be much lower than those typically observed during completely calm conditions [9]. However, most of these findings based on acoustic recordings carried out very far away from the ocean shore. The question is: Should we even care about sea-ice conditions in much shallower regions, like small Arctic bays?

Now, we are all experiencing climate shifts that lead to disappearance of sea-ice. Without ice formed close to the shores, coastlines are directly exposed to the destructive action of ocean waves [10]. This, in turn, poses a serious threat to settlements and infrastructure. It is therefore important to monitor sea-ice evolution in shallow areas, including both the degree of consolidation and phases of transformation.

I am addressing these questions by showing the results of several experiments, conducted in Hornsund Fjord, Spitsbergen, in order to find acoustical characteristics of different types of ice. Sea-ice was present in various forms during the whole field campaign, from a thin layer through rounded chunks (pancake ice) and finally consolidated ice cover (Fig. 1).

Fig. 1. Different forms of sea-ice have different sound signatures. A photograph taken at the study site, in Hornsund Fjord, Spitsbergen, close to the Polish Polar Station.

Recorded underwater noise levels changed periodically together with a tidal cycle. For consolidated ice cover, the highest noise levels occurred suddenly at low water, when underwater rocks are crushing the ice (Mov. 1; Rec. 1). Another scenario takes place for relatively thick ice pancakes. They are packed together when the water is low, but the spaces between them begin to grow during the high tide. With additional wind or current stress, chunks of ice can easily collide and thus produce low-frequency, transient noise (Rec. 2). Finally, for thinner pancakes or freshly formed ice cover, we can hear the loudest sounds when the water is going down. Chunks of mechanically weak ice are squeezed together, leading to deformations and consequently highest underwater noise levels at low frequencies (Fig. 2; Rec. 3). In some cases, stresses acting on ice are not crushing it, but produce sounds resemble a creaking door (Rec. 4).

The results prove that different types of sea-ice react differently for tidal movement, and we captured these differences by acoustic recorders. This relationship can be used for long-term studies of sea-ice conditions in the shallow Arctic bays. The environments, where ocean tides serve as a conductor in the underwater icy concerts.

Fig. 2. Noise levels at low frequencies are much higher when the water is going down (see red frames). Mechanically weak sea-ice cover is squeezed and leads to large deformations and break-up events. The upper plot presents a spectrogram of the acoustic recording lasting more than 15 hours. Brighter color indicates higher noise levels. Time is on the horizontal axis, and frequency in logarithmic scale is on the vertical axis. A value of 3 is a frequency of 1000 Hz, while 2 equates to 100 Hz. The lower plot presents modeled data, corresponding tidal cycle (water level change) for the study site.

Mov. 1. Ocean tides lead to huge deformations and break-up of the sea-ice cover. Time-lapse video from Isbjornhamna Bay, Hornsund Fjord, Spitsbergen.

Rec. 1. The sound of sea-ice brake-up caused by underwater rocks during low water.

Rec. 2. Transient noise of colliding chunks of ice during high water.

Rec. 3. The sound of deforming ice, which is squeezed when the water is going down.

Rec. 4. Sometimes sea-ice processes sound like a creaking door.

The work was funded by the Polish National Science Centre, grant No. 2013/11/N/ST10/01729.

[1] Tegowski, J., G. B. Deane, A. Lisimenka, and P. Blondel, Detecting and analyzing underwater ambient noise of glaciers on Svalbard as indicator of dynamic processes in the Arctic, in Proceedings of the 4th UAM Conference, 2011: p. 1149–1154, Kos, Greece.

[2] Pettit, E. C., K. M. Lee, J. P. Brann, J. A. Nystuen, P. S. Wilson, and S. O’Neel, Unusually loud ambient noise in tidewater glacier fjords: A signal of ice melt, Geophys. Res. Lett., 2015. 42(7): p. 2309–2316.

[3] Deane, G. B., O. Glowacki, J. Tegowski, M. Moskalik, and P. Blondel, Directionality of the ambient noise field in an Arctic, glacial bay, J. Acoust. Soc. Am., 2014. 136(5), EL350.

[4] Pettit, E. C., Passive underwater acoustic evolution of a calving event, Ann. Glaciol., 2012. 53: p. 113–122.

[5] Glowacki, O., G. B. Deane, M. Moskalik, P. Blondel, J. Tegowski, and M. Blaszczyk, Underwater acoustic signatures of glacier calving, Geophys. Res. Lett., 2015. 42(3): p. 804–812.

[6] Milne, A. R., and J. H. Ganton, Ambient Noise under Arctic-Sea Ice, J. Acoust. Soc. Am., 1964. 36(5): p. 855-863.

[7] Ganton, J. H., and A. R. Milne, Temperature- and Wind-Dependent Ambient Noise under Midwinter Pack Ice, J. Acoust. Soc. Am., 1965. 38(3): p. 406-411.

[8] Milne, A. R., J. H. Ganton, and D. J. McMillin, Ambient Noise under Sea Ice and Further Measurements of Wind and Temperature Dependence, , J. Acoust. Soc. Am., 1966. 41(2): p. 525-528.

[9] Macpherson, J. D., Some Under-Ice Acoustic Ambient Noise Measurements, J. Acoust. Soc. Am., 1962. 34(8): p. 1149-1150.

[10] Barnhart, K. R., I. Overeem, and R. S. Anderson, The effect of changing sea ice on the physical vulnerability of Arctic coasts, The Cryosphere, 2014. 8: p. 1777-1799.

Grant B. Deane – gdeane@ucsd.edu
M. Dale Stokes – dstokes@ucsd.edu
Scripps Institution of Oceanography, UCSD,
La Jolla, CA 92093-0206

David M. Farmer – farmer.david@gmail.com
School of Earth and Ocean Sciences,
Victoria BC, V8P 5C2, Canada

Eric D’Asaro – dasaro@apl.washington.edu
Zhongxiang Zhao – zzhao@apl.washington.edu
Applied Physics Laboratory, University of Washington,
Seattle, WA 98105

Popular version of paper 2pAO10
Presented Monday afternoon, June 26, 2017
173th ASA Meeting, Boston

Waves breaking on the ocean, often called “whitecaps,” limit the growth of ocean waves, transfer momentum between the atmosphere and ocean, generate marine aerosols, increase ocean albedo and enhance the air-sea transport of greenhouse gasses. Despite their importance for understanding weather and climate, they remain poorly understood.

The reason for this is clear: breaking waves are the product of storms at sea, they are a source of intense turbulence and they can destroy the sensitive instruments we might use to measure them. This makes them tricky to study in their natural ocean environment, and has encouraged the development of various remote sensing techniques using aircraft and satellites. While we have learned much about breaking waves from above, we still need to understand what is happening in the turbulent core. Here we probe the whitecaps’ inner structure from beneath using the natural sound they create.

The video (link broken) shows a breaking wave seen from above and below during a storm of Point Conception, California in 2000. Credit: Deane

The mass of bubbles that give the whitecap its bright appearance come from the air entrained as the wave breaks. The breaking process generates intense turbulence that fragments the trapped air cavity into a mass of small bubbles. These bubbles create underwater noise. The sounds of crashing surf, the tinkling fountain and the babbling brook are all made by bubbles, which emit a musical pulse of sound when they are first formed.

Each pulse of sound has its own tone that is determined by the size of the bubble making it. So, wave noise intensity and frequency contains information about the numbers and sizes of bubbles entrained by a wave. By measuring the sound safely beneath the fury of the ocean surface, we can learn what is going on within its turbulent interior.

Wave noise has been used over the years to learn many interesting things about breaking waves, including their intensity, how frequently they break and their movement across the sea surface. Wave noise has been used to probe the properties of recently formed bubbles left after a wave breaks and even to infer wind speed, which is closely related to the overall intensity of noise in the ocean.

We have been using wave noise to probe fluid turbulence in whitecaps. Our interest in whitecap turbulence is motivated by its relationship to bubble entrainment and breakup. Fluctuating pressure within the breaking wave driven by fluid turbulence can rupture bubbles by distorting them from their spherical form into irregular shapes.

Small bubbles are stabilized against rupture by surface tension, but large bubbles get ripped apart. These two forces are balanced at a spatial scale, the Hinze scale, which is related to the intensity of the turbulence. The Hinze scale plays a key role in setting the bubble size distribution in breaking waves. An important question is how does the Hinze scale, and therefore the bubble size distribution, change as the wind grows from a gentle breeze to a tropical cyclone?

We might reasonably expect the turbulence to increase with increasing wind speed. If this were true, the bubble distribution created by wave breaking would lead to smaller bubbles at higher wind speed. Surprisingly, this turns out not to be the case. Our experiments on breaking waves in a laboratory show that turbulence intensity in breaking waves, measured by both bubble sizes and a quite different method, reaches a maximum value, relatively independent of the size of the wave.

This leads us to suspect that the Hinze scale, and therefore the bubble size distribution, should be the same for a wide range of wind speeds. We call this phenomenon “turbulence saturation,” and it has important implications for transport processes linking the ocean and atmosphere. But, does this result translate from the laboratory to the open ocean?

Field measurements support this hypothesis. Wave noise was measured along 7 transects across 3 different tropical cyclones. Figure 1 shows measurements of wave noise as it depends on frequency for different wind speeds (colored lines) varying from 15 to 40 meters per second. Notice that all spectra change slope between 2000-4000 Hertz, annotated with the vertical, grey box. The frequency of this break point is thus nearly independent of wind speed.

Figure 1. Measurements of wave noise for wind speeds ranging from 15 to 40 meters per second. The black curves show model calculations of the wave noise under conditions of changing turbulence.

Since we expect this frequency to be related to the Hinze scale, these data suggests that the Hinze scale, and therefore the bubble size distribution, is the same across the entire range of wind speeds. We support this conclusion with a model of sound generation by bubbles (yellow/black lines). The model predicts a peak near the Hinze frequency. Sound generation at lower frequencies is due to other physics and not modeled here. Changing the turbulence dissipation rate by a factor of 5 moves the location of the peak by about a factor of 3, suggesting that if the turbulence intensity did change then we would see evidence of it in the wave noise.

This combination of laboratory and field measurements with theory provide us with evidence of “scale invariance” of turbulence within breaking waves in the open ocean up to 40 meters per second wind speeds, supporting the turbulence saturation hypothesis and demonstrating the unique contributions that ambient sound measurements can make under severe conditions.

[Work supported by ONR, Ocean Acoustics Division and NSF. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation or the Office of Naval Research].