Some 70% of Earth's surface is covered by water, and yet nearly all earthquake detectors are on land. Aside from some expensive battery-powered sensors dropped to the sea floor and later retrieved, and a few arrays of near-shore detectors connected to land, seismologists have no way of monitoring the quakes that ripple through the sea floor and sometimes create tsunamis.
Now, a technique described online in Science this week promises to take advantage of more than 1 million kilometers of fiber optic cables that criss-cross the ocean floors and carry the world's internet and telecom traffic. By looking for tiny changes in an optical signal running along the cable, scientists can detect and potentially locate earthquakes.
The technique requires little more than lasers at each end of the cable and access to a small portion of the cable's bandwidth. Crucially, it requires no modification to the cable itself and does not interfere with its everyday use. The method "could be a game-changer," says Anne Sheehan, a seismologist at the University of Colorado in Boulder who wasn't involved in the work. "More observations from oceanic regions could fill in some pretty big gaps."
It began with an accidental discovery, says Giuseppe Marra, a metrologist at the National Physical Laboratory in Teddington, U.K., who works on the fiber optic links that connect atomic clocks in labs across Europe. He was testing a connection on a 79-kilometer buried cable that runs from Teddington to Reading, U.K., and relies on a stable, resonating loop of laser light. Vibrations near the cable—even the noise of traffic above—can bend it imperceptibly. That can shorten or lengthen the light's travel distance by less than the width of a human hair, shifting the resonating light beams slightly out of phase.
Marra was accustomed to background noise on his fiber optics. But when he reviewed data from October 2016, he saw more than the average amount of noise. It turned out to be the local effects of 5.9- and 6.5-magnitude quakes that struck central Italy late that month. "It was quite a revealing moment," Marra says. That noise, he realized, pointed to a new way to detect earthquakes.
Marra wondered whether the technique could be extended to the ocean, where the environment might be quieter. Using a 96-kilometer submarine cable connecting Malta and Sicily in Italy, he and his colleagues detected a magnitude-3.4 earthquake in the Mediterranean Sea. They couldn't pinpoint it. But by shooting lasers down a cable from both ends, he says, scientists could detect differences in the travel times of the out-of-phase signals, which would reveal just where the earthquake first caused a disruption along the cable. With three or more cables outfitted this way, he says, the earthquake's exact location in the crust could be triangulated.
By filling in the "seismic desert" in the ocean crust and showing where seafloor earthquakes occur and how often, the method could illuminate new fault structures and regions where tectonic plates are colliding or rifting apart, says Charlotte Rowe, a seismologist at Los Alamos National Laboratory in New Mexico. It could also help with tsunami warning systems, she says, provided the strength of the optical signal reveals an earthquake's size.
Besides mapping earthquakes, Rowe thinks the cable networks could sharpen pictures of Earth's interior. Like x-rays in a computerized tomography (CT) scan, seismic waves from big earthquakes carry clues to the density of rock they pass through. From crisscrossing waves received by multiple sensors, seismologists can construct 3D pictures of mantle convection, in which hot plumes well up and cold tectonic plates plunge toward Earth's core. Data from seafloor cables could fill in blind spots in these seismic CT scans. But Rowe says investigators will have to get better at interpreting the cable signals before using them to peer into deep Earth.
Marra says the new technique is sensitive enough to work across ocean basins thousands of kilometers wide. It requires adding a small cabinet of lasers and optical equipment that costs about $50,000 at each end of the cable, and access to just one of the hundreds of channels in a typical cable. Renting a dedicated channel might cost about $100,000 a year on a transpacific cable, and much less on one between North America and Europe, says Stephen Lentz, who works with the cable industry as director of network development for Ocean Specialists, Inc., based in Stuart, Florida. "Frankly, this is the kind of thing where the cable owner could donate the service and take the tax write-off. It costs them little or nothing to share unused wavelengths."
That's significant, says Bruce Howe, a physical oceanographer at the University of Hawaii in Honolulu, who leads a task force exploring how to stud new ocean cables with seismic, pressure, and temperature sensors, every 50 to 100 kilometers. Although the add-on sensors, at roughly $200,000 apiece, are cheaper than operating stand-alone ocean bottom detectors, cable owners have been wary of affecting cable performance. The new technique offers a cheaper and less disruptive way to listen to the ocean floor. Howe calls the results "intriguing" and says his task force will advocate for a longer test. "It should absolutely be pursued."