Woolfe, Katherine F., et al. “Monitoring deep‐ocean temperatures using acoustic ambient noise.” Geophysical Research Letters 42.8 (2015): 2878-2884. DOI: 10.1002/2015GL063438
The ocean is an important part of the Earth’s climate system; it takes up carbon and stores excess heat from the atmosphere. The deep ocean in particular acts as an efficient heat sink. The ocean is so good at taking up heat, it has been credited with the so-called “global warming hiatus.” Understanding how the deep ocean changes in response to this extra heat is a crucial factor in calibrating climate models and assessing global warming trends. But, as you might imagine, actually measuring the properties of the deep ocean is incredibly challenging.
Katherine Woolfe, and her colleagues at Georgia Tech, set out to see if the temperature of the deep ocean could be monitored using nothing but ambient noise. Their process builds on previous research in “ocean acoustic tomography.” The whole method is based on measuring the difference in how long it takes a sound to travel from one place to multiple receivers. Scientists can then extrapolate ocean temperature because the speed of sound in water is almost linearly related to the temperature. That means the time it takes a signal to reach a destination is directly correlated with the temperature of the water it passes through.
Ocean acoustic tomography was first used to measure temperature on a large scale during the 1996 Acoustic Thermometry of Ocean Climate (ATOC) study. The program placed a source in the SOFAR channel – the layer of the ocean where sound propagates most efficiently – just off Hawaii. Transmissions from the source were recorded at a variety of locations around the North Pacific Basin (figure 1). The experiment demonstrated that tomography could measure temperature between a source and receiver more accurately than any other existing method.
The ATOC set-up is an example of an active acoustic system – a known sound is generated at a known location. Woolfe wanted to show that this could be done passively by listening to sounds already bouncing around the ocean. She argues that eliminating the need for a sound source would enable global-scale monitoring of deep ocean temperature with only a network of microphones.
Woolfe and her team analyzed 8 years worth of passive sound recordings from two stations. The first one was located near Ascension Island in the Atlantic and the second near Wake Island in the Pacific (figure 2a). Each station consists of two triangular arrays of hydrophones (aka underwater microphones) oriented North to South (figure 2b). The Ascension and Wake Island hydrophone arrays were separated by 126 km and 132 km respectively. Since both stations had a direct line of sight to the poles, the group decided to listen for cracking ice. Breaking sea-ice in high latitudes has a distinct, low frequency acoustic signature making it easy to pick out of a long time series.
The sound of an individual sea-ice crack would travel from one of the poles to each station. Assume the noise came from the North Pole and arrived at the Wake Island site. The sound wave would then pass through the Northern hydrophone array and then travel to the Southern one. This would happen every time the ice cracked in the detection window of the station. By comparing the arrival times of the sound between the hydrophone arrays, averaged over 1-week intervals, the group was able to back out the change in ocean temperature over time.
Take a look at the plot in figure 3. In it, you’ll see a blue line illustrating the change in temperature between the hydrophone arrays at the Ascension Island station over the duration of the experiment. The temperature change is averaged over the whole depth of the SOFAR channel. The line is the result of recording the temperature change over weekly intervals.
To ground truth their measurements, Woolfe’s team compared their data to temperature estimates derived from the Argo Program. The Argo array consists of nearly 4000 deep-diving, free-floating drifters distributed all over the world’s oceans. Each drifter records temperature and salinity as it goes up and down in the water. It then relays the data via satellite to a server.
Each element of the Argo array provides a point measurement of the ocean environment. Scientists can then estimate what the ocean looks like between the drifters using a series of mathematical transforms. The gray dots in figure 3 are estimates of temperature change at the Ascension Island station derived from Argo. The numbers are generally in good agreement. But the ones derived using Woolfe’s method have much smaller error bars, meaning the measurement is more precise.
That discrepancy makes sense; the temperatures from Argo are derived from point measurements in space and time from the drifters. Data for the rest of the ocean is based on filling in the gaps between those recordings. Woolfe’s acoustic tomography process can make frequent, concrete measurements over a big area.
Neither Argo nor passive acoustic systems can tell us everything we need to know about the ocean; the physical processes that govern it happen on too a huge range of spatial and temporal scales for any one measurement to capture the whole thing. To really observe the ocean requires a diversity of technologies. To borrow a phrase from the ecologist Simon Levin, we are taking a low-dimensional slice of a high-dimensional cake. If we hope to understand our planet and what we are doing to it (and we do), then we will have to employ all kinds of new monitoring techniques.
Eric is a PhD student at the Scripps Institution of Oceanography. His research in the Jaffe Laboratory for Underwater Imaging focuses on developing methods to quantitatively label image data coming from the Scripps Plankton Camera System. When not science-ing, Eric can be found surfing, canoeing, or trying to learn how to cook.