Sound waves traveling thousands of kilometers through the ocean may help scientists monitor climate change.

As greenhouse gas emissions warm the planet, the ocean is absorbing vast amounts of that heat. To monitor the change, a global fleet of about 4,000 devices called Argo floats is collecting temperature data from the ocean’s upper 2,000 meters. But that data collection is scanty in some regions, including deeper reaches of the ocean and areas under sea ice.

So Wenbo Wu, a seismologist at Caltech, and colleagues are resurfacing a decades-old idea: using the speed of sound in seawater to estimate ocean temperatures. In a new study, Wu’s team developed and tested a way to use earthquake-generated sound waves traveling across the East Indian Ocean to estimate temperature changes in those waters from 2005 to 2016.

Comparing that data with similar information from Argo floats and computer models showed that the new results matched well. That finding suggests that the technique, dubbed seismic ocean thermometry, holds promise for tracking the impact of climate change on less well-studied ocean regions, the researchers report in the Sept. 18 Science.

Sound waves are carried through water by the vibration of water molecules, and at higher temperatures, those molecules vibrate more easily. As a result, the waves travel a bit faster when the water is warmer. But those changes are so small that, to be measurable, researchers need to track the waves over very long distances.

Fortunately, sound waves can travel great distances through the ocean, thanks to a curious phenomenon known as the SOFAR Channel, short for Sound Fixing and Ranging. Formed by different salinity and temperature layers within the water, the SOFAR channel is a horizontal layer that acts as a wave guide, guiding sound waves in much the same way that optical fibers guide light waves, Wu says. The waves bounce back and forth against the upper and lower boundaries of the channel, but can continue on their way, virtually unaltered, for tens of thousands of kilometers (SN: 7/16/60).

In 1979, physical oceanographers Walter Munk, then at the Scripps Institution of Oceanography in La Jolla, Calif., and Carl Wunsch, now an emeritus professor at both MIT and Harvard University, came up with a plan to use these ocean properties to measure water temperatures from surface to seafloor, a technique they called “ocean acoustic tomography.” They would transmit sound signals through the SOFAR Channel and measure the time that it took for the waves to arrive at receivers located 10,000 kilometers away. In this way, the researchers hoped to compile a global database of ocean temperatures (SN: 1/26/1991).

But environmental groups lobbied against and ultimately halted the experiment, stating that the human-made signals might have adverse effects on marine mammals, as Wunsch notes in a commentary in the same issue of Science.

Forty years later, scientists have determined that the ocean is in fact a very noisy place, and that the proposed human-made signals would have been faint compared with the rumbles of quakes, the belches of undersea volcanoes and the groans of colliding icebergs, says seismologist Emile Okal of Northwestern University in Evanston, Ill., who was not involved in the new study.

Still, Wu and colleagues have devised a work-around that sidesteps any environmental concerns: Rather than use human-made signals, they employ earthquakes. When an undersea earthquake rumbles, it releases energy as seismic waves known as P waves and S waves that vibrate through the seafloor. Some of that energy enters the water, and when it does, the seismic waves slow down, becoming T waves.

Those T waves can also travel along the SOFAR Channel. So, to track changes in ocean temperature, Wu and colleagues identified “repeaters” — earthquakes that the team determined to originate from the same location, but occurring at different times. The East Indian Ocean, Wu says, was chosen for this proof-of-concept study largely because it’s very seismically active, offering an abundance of such earthquakes. After identifying over 2,000 repeaters from 2005 to 2016, the team then measured differences in the sound waves’ travel time across the East Indian Ocean, a span of some 3,000 kilometers.

The data revealed a slight warming trend in the waters, of about 0.044 degrees Celsius per decade. That trend is similar to, though a bit faster than, the one indicated by real-time temperatures collected by Argo floats. Wu says the team next plans to test the technique with receivers that are farther away, including off of Australia’s west coast.

That extra distance will be important to prove that the new method works, Okal says. “It’s a fascinating study,” he says, but the distances involved are very short as far as T waves go, and the temperature changes being estimated are very small. That means that any uncertainty in matching the precise origins of two repeater quakes could translate to uncertainty in the travel times, and thus the temperature changes. But future studies over greater distances could help mitigate this concern, he says.

The new study is “really breaking new ground,” says Frederik Simons, a geophysicist at Princeton University, who was not involved in the research. “They’ve really worked out a good way to tease out very subtle, slow temporal changes. It’s technically really savvy.”

And, Simons adds, in many locations seismic records are decades older than the temperature records collected by Argo floats. That means that scientists may be able to use seismic ocean thermometry to come up with new estimates of past ocean temperatures. “The hunt will be on for high-quality archival records.”