Acoustic Thermometry to Assess Climate Change

How is climate change impacting the Arctic Ocean? It can be hard to track these changes, but researchers have been using acoustic signals transmitted beneath the ice to learn more. In this episode, we talk with Matthew Dzieciuch and Peter Worcester of Scripps Institution of Oceanography and Hanne Sagen of the Nansen Center about an international effort to use acoustic thermometry to better understand the changing ocean.

Associated paper: Matthew A. Dzieciuch, Hanne Sagen, Peter F. Worcester, Espen Storheim, F. Hunter Akins, Stein Sandven, John A. Colosi, John N. Kemp, and Geir Martin Leinebø. "Transarctic acoustic transmissions during the Coordinated Arctic Acoustic Thermometry Experiment in 2019–2020." J. Acoust. Soc. Am. 159, 1071–1085 (2026). https://doi.org/10.1121/10.0042233.

Read more from The Journal of the Acoustical Society of America (JASA).
Learn more about Acoustical Society of America Publications.

Music Credit: Min 2019 by minwbu from Pixabay. 

Transcript

ASA Publications (00:26)

How is climate change impacting the Arctic Ocean? It can be hard to track these changes, but researchers have been using acoustic signals transmitted beneath the ice to learn more about these changes. Today, I'm talking with Matthew Dzieciuch and Peter Worcester of Scripps Institution of Oceanography in California and Hanne Sagen of the Nansen Center in Norway about their article, “Trans-Arctic Acoustic Transmissions During the Coordinated Arctic Acoustic Thermometry Experiment in 2019 to 2020,” which is part of the JASA special issue on Climate Change and was also featured in an AIP Scilight.

Thanks for taking the time to speak with me today. How are you?

Hanne Sagen (01:05)

We are fine.

Peter Worcester (01:06)

Yeah, I'm great. Happy to be here.

Matthew Dzieciuch (01:09)

Yep, great to talk to you today.

ASA Publications (01:12)

Happy to have you guys. First, just tell us a bit about your research backgrounds.

Matthew Dzieciuch (01:16)

My name's Matt Dzieciuch, and I got into oceanography like a lot of people into it. You know, many people were not officially oceanographers. So I'm an electrical engineer who got into oceanography by actually just doing it, going out on experiments. And I’ve always been intrigued by applying, you know, the digital techniques I learned in EE to oceanography.

Hanne Sagen (01:42)

So my name is Hanne Sagen and I'm from Bergen in Norway. I have been educated in applied mathematics at the University of Bergen in the late 80s. And since then I have been working with underwater acoustics. And then in 2005, I started to work Peter and Matt on acoustics tomography and thermometry. And we started in the Fram Strait. And now we are so happy to be in the Central Arctic Ocean. It's exciting, and I enjoy what I'm doing.

Peter Worcester (02:25)

Good morning. My name is Peter Worcester. I'm an oceanographer at the Scripps Institution of Oceanography. My background, my undergraduate background’s in physics, but I came to Scripps as a student in the early 1970s and have been there ever since, semi-retired now. Going back all the way to my research for my PhD, I worked on using underwater sound as a tool to measure large scale ocean temperatures and ocean currents. And we'll talk more about how that works later, but I've used this tool all over the world from the Philippine Sea to the North Pacific, the North Atlantic, the Greenland Sea, Strait of Gibraltar, now up in the high Arctic, which I find very exciting, it's a very special place. And really happy that we're able to do this now, it's a critical time to be studying the Arctic.

ASA Publications (03:27)

Right, right. So let’s start off with a really big question. Why is it important to study the temperature of the Arctic Ocean? How has it typically been done? And what are the limitations of these methods?

Hanne Sagen (03:39)

So, the satellites have told us over several decades that the sea ice is declining. And that has been primarily due to the warmer air temperatures. And so far, the Arctic sea ice has been protected by the polar water just underneath the ice, a 100- to 200-meter-thick layer that protects it from some warmer water that stays below this layer. And this warmer water comes into the Arctic primarily through the Fram Strait and the Barents Sea area. It cools down and sinks below the polar water and it circulates for about 25 years before it leaves the Arctic Ocean again. So what is brought into the Arctic today, we will not feel until about 20 years forward in time. Therefore, it is very important that we start monitoring now. Or we had been monitoring earlier too, but it's important to intensify this monitoring. Because from now on, the climate model shows that there is an increasing increase in ocean heat below the ice. So that's one of the most important things to measure nowadays is how warm the water is under the ice. And also when the ice gets thinner, and there is more open areas, there will be more dynamic processes that starts to mix the cold water that stays under the ice with the warmer water that is further down, and as the stratification get weaker the ice might decline even faster. And how we measure this is with the ice-tethered platforms, for example, that drift with the ice and have lot of instruments hanging below it. Or you can have, for example, moorings, or seafloor installations, that also carries a lot of instrumentation. And this is what we also do in CAATEX, that we have a lot of other instruments than just acoustic thermometry instrumentation. And what I think that Peter and Matt will talk about is how these changes will influence the acoustic propagation conditions in such a way that we can observe it by acoustics. So I leave this to Peter to explain more about what acoustic thermometry is.

Peter Worcester (06:28)

Yeah, I might add to Hanne's answer  about how the measurements have been done. As you might imagine, the ice makes it very difficult to get measurements of the Arctic Ocean underneath the ice. Hanne mentioned a couple of ways it's done, one with instruments that are hanging from the ice and one with moorings that come up from the bottom. But just to install these things, you need icebreakers, very special ships. And a number of the techniques that we use to measure the ocean in mid-latitudes where there isn't ice are either very difficult or aren't really applicable. And so sound, which can travel underneath the ice, has a very special role to play in helping to measure what's happening to the ocean under the ice.

ASA Publications (07:20)

That's really, really interesting. So Hanne kind of alluded to this question for you guys, but what is acoustic thermometry?

Peter Worcester (07:28)

Acoustic thermometry is a technique for using sound as a way to measure ocean temperatures. The basic principle is actually quite simple. The speed at which sound travels in the ocean depends on temperature, salinity, and pressure, but mostly on temperature. So by measuring changes in the speed at which sound travels in the ocean, we're fundamentally measuring changes in the temperature. And to do this, we put acoustic sources in the ocean and acoustic receivers that receive the signals transmitted by the sources. And sound travels a very long way in the ocean. Low-frequency sound can travel distances of up to thousands of kilometers. So if your goal is to measure large-scale average temperatures, which is what you really want for studying climate, then sound is a very good way to do this. If we transmit a pulse of sound from a source and receive it at a point, say a thousand kilometers distant, and measure how long that takes very precisely, with an accuracy of thousandths of a second, and then measure how that changes with time, you can say how the average temperature over that whole thousand kilometers is changing with time. And this is very different from most measurements that we make. Most measurements are made at a point, whether from instruments lowered from ships or from buoys that drift in the ocean at depth or in other ways. So the two are really rather complementary. If you want fine-scale information, the point measurements are, of course, much better. But if you want large-scale average temperatures and currents, then sound is a very good way to do that.

ASA Publications (09:25)

That's really cool. So tell us about the US Norwegian Coordinated Arctic Acoustic Thermometry Experiment. What exactly did you do?

Matthew Dzieciuch (09:33)

So in this experiment, the CAATEX experiment, we thought that we would try to apply acoustic thermometry to measuring the temperature in the Arctic. And this had been done before, so we were inspired by that previous experiment. This previous experiment was done in 1994, and it was called the TAP experiment. So we thought we would try instruments along the same line as the TAP experiment. The TAP experiment transmitted sound from near Svalbard all the way to near Alaska. And we realized that we could do the same thing. And we tried to do even a little bit better, even instrument more than they did in that experiment. So that's what we set out to do. And we ended up putting out six moorings along the line from Alaska to Svalbard. And there was a transmitter sort of on the western side, near Alaska, and another transmitter on the eastern side, near Svalbard. And so that was on two of the moorings, and those moorings also had receivers. And then there were four other moorings that had just receivers. All the receivers are called hydrophones; they’re special sensors that record sound underwater. And we put them vertically, so they measured the distribution of sound in the vertical at a particular mooring. And we deployed those for a year. 

And they're quite complicated instruments because they have to be able to keep time very accurately underwater. You don't have access to GPS to keep an accurate clock. So we have to build atomic clocks into each of these instruments. And we also have to know exactly where they are. So they sort of navigate themselves with transponders on the bottom to compensate for the fact that they move around in the current as the current move over the year.

So that's what we did. We transmitted sound across the Arctic. I think it was every third day you know, between going back and forth across the Arctic.

ASA Publications (11:28)

Okay, okay. So it sounds like there was a lot of technology involved in these, as you mentioned.  What technological advances in thermometry made this work possible?

Matthew Dzieciuch (11:38)

There were some significant advancements in technology over the years that made this possible in a couple different areas. So one of the things is just the sound source itself. And there was a new sound source that was developed, actually for the oil industry for doing surveying, for when they try to map out where the oil is under the seafloor. And that sound source was developed by a company in Canada. And it, because it was the oil industry instrumentation, they did a lot of testing, they made it very reliable, which was very helpful for us, because we could trust it. And another thing that happened over the years was that atomic clocks got a lot better. So we were able to have an atomic clock on low power available underwater. So we could deploy these instruments for a year and keep the time measurement done very accurately over an entire year. And just the fact that we've been doing this for many years, you learn lot of lessons along the way about how to do this. Even though Peter said it sounds very simple making a travel time measurement, there’s a lot of little tricks and things that just come with experience that you learn how to do over the years.

ASA Publications (12:47)

Right, right, that totally makes sense. You were also concerned with ocean stratification and sea ice. Why, and what did you find?

Hanne Sagen (12:55)

So what I have learned about stratification and why it is important. Well, the ocean stratification, if that change in the Arctic Ocean, that the warmer water gets closer to the sea ice, then the melting of the sea ice will accelerate and it will go faster into a blue Arctic. And what we have seen in the data is events where the warm water comes up to the surface in the eastern Arctic, on the European side. So when that happens, it changes the acoustic propagation quite a lot. And we can see that in the recordings that we have on our side. In the Western Arctic, in the Beaufort Sea, it's different, and that is more something that Peter and Matt knows more about.

Peter Worcester (13:57

Yeah, I might add to that. As Hanne indicated earlier, the bulk of the water coming into the Arctic comes through the Fram Strait or through the Barents Sea on the other side of Svalbard. And that forms what's called the Atlantic water, which is warm, salty water that circulates around the entire Arctic Ocean at depth. And that's one of the aspects of what's changing, what Hanne was talking about. A small amount of water also comes in through the Bering Strait between Alaska and Russia, and that water has been changing as well. And it's created a subsurface temperature maximum, a temperature maximum that's below the surface in the Western Arctic, north of Alaska, in an area called the Beaufort Sea. And that too has greatly affected acoustic propagation. So there's multiple effects going on affecting the stratification in the Arctic, really how the different waters, waters of different temperatures and salinities are layered, if you will, in the Arctic.

 I might also comment on the ice. As Hanne indicated, the ice cover has been disappearing, and that's affecting the near surface waters. Disappears only in the summer, it's still there in the winter. But there used to be ice that would survive from one winter to the next. It was called multi-year ice. And as the Arctic is warm, that's almost entirely disappeared. So now we're up at the North Pole, for example, the ice used to be three or four meters thick. Now it's one meter, maybe one and a half meters. And so there's been tremendous changes in the ice thickness, and this has also changed the ice roughness. And all the sound in the Arctic as it travels interacts with the ice. It's all bent up toward the surface. And so as the ice gets thinner and less rough, sound can travel further because the ice isn't scattering it as much. So there's multiple effects going on here at the same time.

And all of this is important because when you use sound as a tool to measure ocean temperatures or ocean currents, you really have to understand exactly how that sound is traveling, how it's propagating, in order to accurately understand what the ocean is doing to it.

ASA Publications (16:38)

So you all mentioned a previous experiment from 1994. How did you expect the way the sound travels through the Arctic Ocean to have changed compared to the 1994 Trans-Arctic Acoustic Propagation Experiment?

Peter Worcester (16:52)

As Matt said, that experiment really inspired the CAATEX experiment. And I think all of us had really hoped that we'd be able to compare the travel times, and therefore the temperatures, measured in the 1990s with the temperatures that we measure now.  And I think what we half expected to see was that the ocean would have warmed further. In fact, it didn't quite turn out that way. There were some issues that really made that difficult to do. 

One was that the experiments in the 1990s had difficulties with their clocks. And so the travel time measurements were only accurate to about a second. And it didn't have anything to do with the acoustic propagation. They basically had trouble setting their clocks accurately to GPS. And so this means that the travel times measured then were only good to kind of plus or minus a second. The travel times we measured were good to thousandths of a second. And they would have been in the 1990s too, if it weren't for this clock setting issue. But what this meant was it was very hard to be able to conclude that the ocean had in fact warmed or cooled or changed because the error bars on those early measurements were so large, the uncertainties were so large.

And then there was a second part to it. As we got into this, we started looking at other data that was available. There's the thing called World Ocean Database and the World Ocean Atlas, which is a compilation of all the measurements of ocean temperatures, salinities, and other properties that have been measured. And when we simulated our experiment in the World Ocean Atlas, we could do that decade by decade, the 1990s, 2000 to 2010, 2010 to 2020. And what we found when we did that was that the predicted travel time changes for our path were not very large. And we've since looked at other compilations of ocean data and they all give essentially the same answer that for our path and our measurements, the way we sample the ocean, the travel time changes were not that large. And so I think the effect we'd hope to see, in fact, if anything, we showed that the compilation of ocean temperatures that were available, in fact, were not too bad. They predicted our current measurements pretty well, but they also predicted that there hadn't been that much change over the decades.

I think there's one other thing to say about that is really these measurements in the 90s were just made for a few days in April. They were made from instruments suspended from the ice. And so there was maybe a dozen measurements of travel time. Whereas we measured the changes in temperature over the whole year. And what we found was that the temperatures over a year could actually vary by close to a second. And so this again made trying to compare our data, which showed much more variability on short time scales than we expected, made it hard to compare with these earlier measurements. So I think what we showed really was that our measurements were very precise, very accurate. We could measure changes in large-scale temperature over the year on time scales that were virtually inaccessible using other methods. But we didn't really make a very satisfying comparison to the earlier measurements.

ASA Publications (20:48)

Okay, so it really speaks to the technological advances involved in this, like you said, with the atomic clock, making it so you can actually time things more precisely and so on.

Peter Worcester (20:58)

Yes,

Matthew Dzieciuch (20:58)

You mentioned  that our experiment seemed that the TAP measurements were, I guess, somewhat compromised by the inability to get the clock set correctly. So that's why we compared to the World Ocean Atlas and World Ocean Database. But it's also important to note that our measurements are probably much more accurate than those things also. And the comments he made about the accuracy of our experiment, compared to the World Ocean Atlas, are also valid. And the time series we're getting over the season is much better than is available in the World Ocean Atlas and Database. So I just wanted to emphasize that also.

Peter Worcester (21:38)

Yeah, I might add to that. I mean, to give people some feeling for what happens in the World Ocean Atlas, it's basically in the Arctic, made up of relatively rare measurements made from icebreakers, which follow, you know, limited paths in the Arctic Ocean. It's made up from data with instruments hanging from the ice called, ice tethered profilers, that Hanne mentioned earlier.

But you can't control where these drift, they go with the ice. And so it's not like you have a measurement that stays in the same place. You're mixing time and space in the measurement. And then you have moorings. And in CAATEX, we made a lot of moored measurements of temperature, but there's very few of those in the Arctic. It's so hard to get them up there because you have to use an icebreaker, and you have to get them back through the ice that there's, there's just not a lot of other data and to try to construct how the temperature changes over years is exceedingly difficult with those conventional measurements.

Hanne Sagen (22:42)

Yeah, I think what you say shows clearly how important it is that we continue to measure with accurate methods, like the acoustic thermometry, over several years, so that we can look at both the inter-annual variability but also to look at the long-term trends which is needed in climate models, for example, and to have data available for constraining these models in a good way is also very important. And as I mentioned earlier, this is the time to do it because there is an increase in the increasing temperature. So it goes much faster from now on, and the next decades, according to the models. And even if the models are different from each other, they have the same trend, that the increase is increasing. So I really hope that we can continue with CAATEX experiments, or similar, in the Arctic because it's the only way that you get synoptic measurements of large ocean volumes and that you maintain sections, and not as Peter says drifting buoys is fine but you have the geographical and the temporal variability mixed up in the same measurement, and it's hard to decouple. So I say yes for moorings and acoustic thermometry for the climate monitoring in the future.

ASA Publications (24:32)

Yeah, totally. What did you end up observing in the received signals?

Peter Worcester (24:36)

I think in the signals, what we observed was really two things. One of the questions and what we were doing is we were transmitting signals at a much higher frequency in CAATEX, at 35 Hertz, compared to those in the 1990s, which had been done at about 20 Hertz. And one of the questions was how well would sound propagate at that higher frequency; higher frequencies are scattered more by the ice cover than lower frequencies. And it turned out that we could transmit all the way across the Arctic at this higher frequency because of the changes to the ice cover, that it's less rough. There's fewer of what are called “ice keels”. Back in the day, as the ice was pushed by winds and currents, the ice would collide and pile up and ridges and big keels that would go down underneath the ice for tens of meters, sometimes 30, 40 meters. And this really scattered the sound a lot. That doesn't happen much anymore. So one thing we found was we could do it at this frequency, which lets one use much more manageable sources than the ones needed for the lower frequency. 

The other thing we found is that as we measured the changes in the travel times across the Arctic, that the travel times, I think it's fair to say, changed more over the year than we'd really quite anticipated. It seemed that the Arctic was more variable on those shorter time scales than we’d anticipated. And, you know, we would see travel time changes on some of the paths of up close to a second, which is a huge change compared to the accuracy which we measured the travel times with a few thousandths of a second. So I think those are really kind of the two things that we found in this experiment.

Yeah, I'm not sure this is the time to add it, but I might add one other thing. The question of the role of acoustics in the Arctic, as I indicated earlier, sound has the unique property of being able to travel underneath the ice. And that makes it useful for acoustic thermometry, which we've talked about it some length here. But it also means that sound can be used to position instruments under the ice. So example, if one put in drifters under the ice that profile the temperature and salinity profiles, we can do that now, but then we don't know where they are because they can't come to the surface to get their position from GPS. You could use sound so that they could go under the ice and make measurements and then know where they were. So it becomes an underwater navigation system, an underwater GPS, if you will.

And also the same instruments, the same technology lets one record just the sound that's there, the sound from the ice, the sound from marine life. And this is called passive acoustic monitoring. And it's a powerful tool for monitoring both the ice, marine life, and to see what's going on. So I think our vision really is a system that uses sound in multiple ways in order to improve our understanding of the Arctic.

ASA Publications (28:12)

So how can this work be used to further our understanding of the impact of climate change on the Arctic Ocean?

Hanne Sagen (28:18)

It is important with these long-term observations at fixed positions and also along fixed sections, which is the power of acoustic thermometry that you repeat the same section over and over again. And that I think something unique that you can't get in any other way. A ship that goes, I don't know how long time it takes for 1000 kilometers. It's several days. So I think that just this synoptical measurements are something that we need and also that the public need to understand and to believe in the changes that is going on in the Arctic.

Peter Worcester (29:10)

Yeah, I might give a little more background here. It's important to realize that the Arctic is changing at a rate far faster than the rest of the planet. The surface air temperatures in the Arctic are increasing at something like four times the rate of the rest of the planet. The ice is disappearing in the summertime and we fully expect that sometime in the next 10 years that the Arctic will become ice-free in the summertime, not in the winter, but in the summertime with huge implications for what people can do with the Arctic. So the Arctic is just changing it in ways difficult to grasp and are just incredibly important for the rest of the planet.

Matthew Dzieciuch (30:00)

Yeah, and just to go on from that, we know a lot about the atmosphere because we have satellites and we can see the ice from satellites, but it's hard to measure under the ice in the wintertime. And you know, measuring the water is something, under the ice is something that acoustic thermometry can do. So that's why we're doing it.

Hanne Sagen (30:20)

And also with the passive acoustics, I mean, you can observe, if you start monitoring or you keep on monitoring for several years, you can see how marine life is changing. Because with the higher temperature, the ecosystems are changing. And I think that we will hear it in our data if we get long-term records. In fact you can listen to the climate change in the Arctic.

ASA Publications (30:52)

Right, right. So what is the value of modern acoustic thermometry for scientists?

Matthew Dzieciuch (30:57)

One of the most important things is that establishing a baseline measurements in 2019 t0 2020, right? Time advances quickly and we'll, you know, it's already five years later and I'm certain that the Arctic has changed. So by having this nice, precise measurement over that time period, doing it again, lets us go back and compare to those measurements. In terms of, you know, both travel time, ambient sound, and the structure of the measurements for the stratification, the structure of the pattern we measure for the stratification. Those are all important things. Also having this data set to work on. You that work doesn't just stop once you make one measurement. You learn how to use these measurements in different ways. So we're learning new methods, new what we call inverse methods, how to interpret the what we call acoustic modes, how they propagate in the sound channel. We're learning how to interpret those measurements ocean models. And ocean models are, they’re the models that people use to understand how the ocean circulates, and in the Arctic because they have little data, they're a little bit blind, so in this state it's really important for those models to make them work better. So I think that there's a of value in these measurements that still is coming out. And as we hopefully make more of them, we'll get even more value.

ASA Publications (32:24)

So that's kind of a good segue. What are the next steps in this research?

Hanne Sagen (32:28)

Well, it's five years since CAATEX, and we have already taken some more steps where we use acoustic thermometry in combination with ocean observations and passive acoustics. But we have two nice sister projects ongoing, the HiAOOS and the HiAATS project. Where we have set up… Tell it.

Matthew Dzieciuch (32:58)

Yeah,  so we haven't been just sitting here analyzing the measurements. We also went back out with the equipment and redeployed the instruments and put them on four different moorings in the Eastern Arctic, sort of perpendicular to the trans-Arctic path that we had before, in sort of a box actually. So those instruments are out there right now. And from what we learned in CAATEX, that informed what we're doing now in that experiment. So we're hoping to use these to learn even more. And it's in a slightly different area. And the emphasis is a little bit more on getting the navigation data that Peter alluded to that, you know, it's actually possible to navigate a buoy underwater. So we're learning how to do that. And we're gonna pick those moorings up later this year, in August of 2026. So we'll have a lot more data to compare to previous measurements, as infrequent as they were in the past. We are setting a baseline in a different part of the Arctic now. And then we're hoping to continue. It doesn't really make any sense to bring these instruments back to the lab at the end of the experiment and let them sit on the shelf. We're hoping that we can redeploy them and keep them out there because that's the only way you learn something. So that’s what we're really hoping to do is put them back out and keep that going.

Hanne Sagen (34:20)

Yes and, because we, this time have three sources out and that makes it possible for people who have floats or drifting stuff under the ice capable to position their devices. It has been a little hard to get people, I think, to benefit from this system, and that makes it even more important that we get this out now and we can show that it's actually working and that they have this tool to geoposition their data. So that is my hope, that we get much more attention to the benefit of having an acoustic system out in the Nansen and Amundsen basins in this time, but maybe we can expand, like Peter Mikhalevsky suggested a long time ago to cover the whole Arctic with a geopositioning system.

ASA Publications (35:20)

So do you have any closing thoughts?

Matthew Dzieciuch (35:22)

I want to say thanks for having us on your program today. And I'm really, you know, looking forward to the future to you know keep these measurements going. I think think they are going to be very useful for people in the future, both in terms of, you know, having a baseline and learning to understand what is really going on up there, but also as Hanne said, just for, you know, having a dual-use system that lets people do navigation is going to be very interesting, and it’s going to expand the amount of data available in the Arctic.

So thanks for having us.

Hanne Sagen (35:57)

And also what I want to say is that I hope that Norway, Europe, and US can collaborate on this in the future, because it's expensive and it's demanding. As you said earlier, we need icebreakers, we need equipment and stuff. So I really cross my fingers for continued collaboration in the future.

ASA Publications (36:23)

Yeah, it really seems like the international collaboration is essential to this type of research, you know.

Hanne Sagen (36:28)

It is.

Matthew Dzieciuch (36:29)

Absolutely. Absolutely. Really. Everybody benefited from that. We couldn't do this without that collaboration.

ASA Publications (36:37)

I'm so glad to have you guys on. It's really cool that you were able to learn so much about the changing ocean temperature over the course of a year. And hopefully the techniques you used in this work will allow us to better understand how the environment is changing in the Arctic Ocean. Thank you again for taking the time to speak with me today. And I wish you all the best of luck in your future research.