Search

Ambient noise oceanography is proving to be an efficient and effective tool for the study of ice-ocean interactions in the bays of marine-terminating glaciers. However, obtaining quantitative estimates of ice melting or calving processes from ambient noise requires an understanding of how sound propagation through the bay attenuates and filters the noise spectrum. Measurements of the vertical structure in sound speed in the vicinity of the Hans Glacier in Hornsund Fjord, Spitsbergen, made with O(130) CTD casts between May and November 2015, reveal high-gradient, upward-refracting sound speed profiles created by cold, fresh meltwater during summer months. Simultaneous recordings of underwater ambient noise made at depths of 1, 10 and 20 meters in combination with propagation model calculations using the model Bellhop illustrate the dominant role these surface ducts play in shaping the underwater soundscape. The surface ducts lead to a higher intensity and greater variability of acoustic energy in the near-surface layer covered by glacially-modified waters relative to deeper waters, indicating deeper zones as most appropriate for inter-seasonal acoustic monitoring of the glacial melt. Surface waveguides in Hornsund are relatively shallow and trap sound above O(1 kHz). Deeper waveguides observed elsewhere will also trap low-frequency sounds, such as those generated by calving events for example. Finally, the ambient noise field in Hornsund is shown to be strongly dependent on the distribution of ice throughout the bay, stressing the importance of performing complementary environmental measurements when interpreting the results of acoustic surveys. This article is protected by copyright. All rights reserved.

Air-entraining breaking waves form oceanic whitecaps and play a key role in climate regulation through air-sea bubble-mediated gas transfer, and sea spray aerosol production. The effect of varying sea surface temperature on air entrainment, subsurface bubble plume dynamics, and surface foam evolution intrinsic to oceanic whitecaps has not been well studied. By using a breaking wave analog in the laboratory over a range of water temperatures (Tw = 5°C to Tw = 30°C) and different source waters, we have examined changes in air entrainment, subsurface bubble plumes, and surface foam evolution over the course of a breaking event. For filtered seawater, air entrainment was estimated to increase by 6% between Tw = 6°C and Tw = 30°C, driven by increases of about 43% in the measured surface roughness of the plunging water sheet. After active air entrainment, the rate of loss of air through bubble degassing was more rapid at colder water temperatures within the first 0.5 s of plume evolution. Thereafter, the trend reversed and bubbles degassed more quickly in warmer water. The largest observed temperature-dependent differences in subsurface bubble distributions occurred at radii greater than about 700 μm. Temperature-dependent trends observed in the subsurface bubble plume were mirrored in the temporal evolution of the surface whitecap foam area demonstrating the intrinsic link between surface whitecap foam and the subsurface bubble plume. Differences in foam and plume characteristics due to different water sources were greater than the temperature dependencies for the filtered seawater examined.

A number of remote and in situ acoustical sensors were mounted in the surf zone off Scripps Pier, La Jolla, California, during March 1997 and used to characterize rip currents and their ability to transport bubbles offshore from the surf zone. The bubble size distributions and air fractions at the injection point were measured using an acoustical resonator and a conductivity sensor right in the surf zone. The offshore moving bubble plume was tracked with a multibeam 100 kHz Doppler sidescan sonar, giving the spatial extent and offshore variability in the rip events as well as information about the surface wave field. Furthermore, three acoustical resonators were mounted at 2.5 m depth 55–60 m farther offshore to probe the bubble field 300–400 s after injection. These results are interpreted using models that include bubble buoyancy and dissolution, rip current advection, and the turbulent boundary layer, which is represented by the Christoffersen and Jonsson [1985] formulation.

[1] Whitecap foam decay times for 552 individual breaking waves determined from digital images of the sea surface are reported. The images had sub-centimeter pixel resolution and were acquired at frame rates between 3 and 6 frames per second at the Martha's Vineyard Coastal Observatory over a 10-day period in 2008, subdivided into 4 observation periods. Whitecap foam decay times for individual events varied between 0.2 s to 10.4 s across the entire data set. A systematic positive correlation between whitecap foam decay time and maximum whitecap foam patch area was found for each observation period. For a given whitecap size within each observation period, the decay times varied between a factor of 2 and 5, with the largest variation occurring during unsteady environmental forcing conditions. Within observation periods, bin-averaged decay times varied by up to a factor of 4 across the range of foam patch areas. Between observation periods, the effective whitecap foam decay time, which we define as the area-weighted mean decay time, varied by a factor of 3.4 between 1.4 s and 4.8 s. We found a weak correlation between decay times and individual event-averaged breaking wave speeds. The variation in the active breaking area across all 4 observation periods was small, indicating relatively uniform surface whitecap area generating potential. We speculate that the variation in the foam decay times may be due to (i) the effect of surfactants on bubble and foam stability, and (ii) differences between bubble plume characteristics caused by a variation in breaking wave type.