In 1991, scientists lowered large subwoofers into the water at Heard Island, a snowcapped volcanic island in the Indian Ocean. The speakers emitted low-frequency sounds that, like whale song, rumbled across entire oceans. Picked up by receivers off the coasts of California and Bermuda, the signals contained a crucial piece of information about the water they had traversed: how hot it was. It was a promising way to monitor Earth’s warming oceans, but concerns about how the underwater noise might affect marine life soon sidelined it, with only a few dedicated scientists keeping the technique alive. Now, it is back—only this time, Earth itself is providing the noise.
A team of seismologists and oceanographers has shown that small earthquakes repeatedly emanating from the same spot beneath the ocean floor can take the place of the subwoofers. The quakes generate reliable acoustic signals for measuring ocean temperatures, including at depths below 2000 meters, beyond the reach of other techniques. If validated, the approach, published today in Science, could open an entirely new ocean observation system for understanding past and future climate change, says Frederik Simons, a geophysicist at Princeton University unaffiliated with the study. “There’s a potential treasure trove of data waiting to be analyzed.”
The oceans absorb more than 90% of the energy trapped by global warming, and any change in the rate at which they soak up heat would have an outsize impact on how fast the atmosphere warms. Two decades ago, robotic floats from the international Argo array began to monitor the warming of the ocean to a depth of about 2000 meters. But the float array, now 4000 strong, could not probe the large volume of water at greater depths. “The inability to determine what is going on in the deep water is a major obstacle to understanding the ocean and climate, even today,” says Carl Wunsch, a retired oceanographer from the Massachusetts Institute of Technology.
In 1979, Wunsch and Walter Munk, an oceanographer at the Scripps Institution of Oceanography who died last year, first proposed using sound waves to measure the ocean’s heat and structure. Sound travels faster as water grows hotter or denser, making its travel time a reliable gauge of temperature and density if the sound source and receiver are at fixed locations.
The technique did not require especially loud sources. At a depth of about 1000 meters, the speed of sound hits a minimum, forming a conductive channel between warm waters above and dense water below. This waveguide enables sound waves to coast across entire ocean basins, says Bruce Cornuelle, a Scripps oceanographer who worked with Munk. “It’s like a 5-year-old grabbing a wrapping paper tube and yelling in his brother’s ear.”
Besides probing the entire width of an ocean, the sound waves—with vertical amplitudes of thousands of meters—capture conditions from shallow waters all the way down to the abyss. As a result, they average out smaller scale natural temperature fluctuations, revealing basinwide changes of just a few thousandths of 1° per year. “That makes it much easier to extract the global warming signal,” says Jörn Callies, an oceanographer at the California Institute of Technology (Caltech) and co-author on the new study.
After the 1991 demonstration at Heard Island, Munk won Department of Defense funding for a follow-up experiment in the Pacific Ocean, called Acoustic Thermometry of Ocean Climate (ATOC). But it became mired in controversy over its two human-size speakers, placed off the coasts of Hawaii and California in prime whale territory. “It became a political nightmare,” says Brian Dushaw, a retired oceanographer who worked on ATOC. ATOC’s signals were no louder than whale calls and ship traffic, but much of its $35 million budget went to studies of the sound’s impact on marine mammals.
Military secrecy also got in the way. To hear the signals, the project relied on classified Navy hydrophones normally used to detect submarines. The scientists couldn’t even publish the receivers’ locations, Wunsch says. “We didn’t tell the Navy that if you published the signal, which we did, then you could figure out where the receivers were,” Wunsch adds. The Hawaiian source, off Kauai, ran until 2006, providing 10 years of warming data. But by then, oceanographers had left acoustic thermometry behind and were relying on Argo, Dushaw says.
That was until 1 year ago, when Wenbo Wu, a Caltech seismologist, realized that repeating earthquakes on slowly creeping faults below the sea floor could provide an alternative sound source. When earthquakes shake the ocean floor, some of the energy is transformed into acoustic waves. Wu and his co-authors just had to find the right source.
Their search went back to the Indian Ocean. In earthquake records, they identified more than 4000 earthquakes from faults in the ocean floor west of Sumatra in Indonesia from 2004 to 2016, many of them between magnitude 3.5 and 5. Triangulating on the source, the team identified patches of fault less than 100 meters apart that ruptured repeatedly, says Sidao Ni, a co-author and seismologist at the Institute of Geodesy and Geophysics of the Chinese Academy of Sciences. The resulting sound waves traveled through the ocean unfettered to Diego Garcia, a remote atoll south of India, where they hit land and turned back into seismic waves, picked up on the island’s seismometer.
Converting those travel times to temperatures, Wu and his colleagues found that the eastern Indian Ocean warmed 0.044°C over the decade. The annual fluctuations matched up well with Argo measures from the same time, but the warming signal was nearly double what the Argo floats detected. The disparity suggests Argo is missing some heat, Callies says, at least for this basin over this short span of time. Some 40% of their heat measurement came from water below 2000 meters, suggesting some warming is working its way deeper into the ocean, out of Argo’s current reach.
This work is “quite extraordinary and very promising,” says Susan Wijffels, an Argo leader at the Woods Hole Oceanographic Institution. If extended globally, it could provide an independent check on Argo measurements, especially when production of a new line of Argo floats that can descend 6000 meters, currently deployed only in the dozens, ramps up. Even more alluring for Wijffels is the possibility of extending global warming trends back in time, before Argo, by detecting repeaters in old seismic records. “What a gift to the climate community that’d be,” she says.
The team thinks it can capture the earthquake-generated sounds more cleanly with hydrophones than with land-based seismometers. That will let them use lower power earthquakes, and by using the global network of hydrophones deployed as part of the Comprehensive Nuclear Test Ban Treaty, they should be able to pick up signals from repeaters throughout the world’s oceans.
Hydrophones deployed under Arctic sea ice might gauge water temperatures in a place Argo floats can’t reach. It might even be possible to use the crash of collapsing ice in nearby Greenland—glacial earthquakes, as they’re known—as the sound source. “It’s free data,” Dushaw says. “There’s no question someone will implement a system to take advantage of this.”
The newly bright prospects for ocean acoustic thermometry are also a validation for Munk, who was deeply saddened when his global acoustic dreams were muted, Cornuelle says. “I wish Walter had been around to see it. He’d be overjoyed.”
*Correction, 17 September, 4:25 p.m.: A previous version of this story stated that 40% of the measured warming came from below 2000 meters. Although 40% of the measured temperature came from water below 2000 meters, the technique cannot yet say where in the water column the warming occurred.