Turkey and Syria’s 7.8-magnitude quake on Monday is a brutal reminder that deep down, planet Earth still hides secrets. Scientists know full well that faults are prone to earthquakes, but they can’t tell when a shaker will strike or how big it’ll be. If they could, the death toll wouldn’t stand at over 20,000 so far—and rescuers are still scrambling to find survivors

Still, in recent years scientists have made progress in developing early earthquake warning systems, in which seismometers detect the beginnings of rumblings and send alerts directly to people’s phones. That alarm comes not days or hours before the quake strikes, but seconds. The planet’s seismic strikes are just too sudden for scientists to provide substantial warning times.

A novel technique, though, could one day boost those early warning systems, providing extra time for people to prepare for incoming quakes—although it’d still be on the order of a few seconds, depending on how close a person is to the epicenter. It’s called distributed acoustic sensing, or DAS. Though the field is still in its infancy, DAS could tap into the fiber optic cables buried under our feet as a sprawling, ultra-sensitive network for detecting seismic waves. These cables are used for telecommunications, but they can be repurposed for sensing earthquakes and volcanic eruptions because the ground’s movement slightly disrupts the light traveling through the cable, creating a distinct signal.

DAS can’t predict earthquakes; it just detects early tremors. “Any system, whether it is a seismometer or fiber optic cable, cannot detect things before they happen at the sensor,” says geoscientist Philippe Jousset of the German Research Centre for Geosciences, who has used DAS to detect volcanic activity on Italy’s Mount Etna. “We have to have the sensor as close as possible to a source so that we can detect early. There are a lot of cables everywhere. So if we could monitor them all at once, then we would get information as soon as something is happening.” 

When a fault ruptures, it fires off different kinds of seismic waves. The primary ones, P-waves, travel at 3.7 miles per second. These aren’t super damaging to homes and other infrastructure. Secondary waves, or S-waves, are much more damaging, traveling at 2.5 miles per second. Even more destructive are surface waves, which move at about the same speed as S-waves or maybe a bit slower. These rip along Earth’s surface, leading to dramatic deformation of the ground. (They’re especially destructive because their energy is concentrated on a relatively flat plane along the surface, whereas P-waves and S-waves spread out more three-dimensionally underground, distributing their energy.)

Existing earthquake early warning systems, like the United States Geological Survey’s ShakeAlert, use seismometers to exploit the differing speeds of seismic waves. ShakeAlert consists of about 1,400 seismic stations across California, Oregon, and Washington, with plans to add nearly 300 more. These monitor for fast-moving P-waves, which forewarn of more damaging S-waves and surface waves on the way. If an earthquake strikes and at least four separate stations detect the event, that signal is sent to a data center. Should the system’s algorithms determine that the tremor will be above a magnitude 5, it’ll trigger an emergency alert to be sent to the cell phones of local residents. (Thanks to a ShakeAlert partnership with Google, it goes out to Android users if the magnitude is above 4.5.)

All this shuttling of data through modern telecommunications equipment happens at the speed of light—around 186,000 miles per second—which is much, much faster than destructive seismic waves travel. But how much warning a resident gets depends on how far away they are from the epicenter. If they’re right on top of it, there just isn’t enough time to get the alert before they feel shaking. Think of it like a thunderstorm: The closer you are to the lightning, the sooner you hear the thunder.

“Everything happens super fast,” says Robert-Michael de Groot, a member of the ShakeAlert operations team at the USGS Earthquake Science Center. “If you’re far enough away, you may get a few seconds. And that’s better than before earthquake early warning existed, where basically the only signal that you knew that something was going on was the ground was shaking.” 

With those few seconds, people can gather up their kids and get under a table. ShakeAlert basically outruns the earthquake, at least the bits of it that humans experience on the surface as intense shaking. “It’s a race,” says de Groot. “People may feel a bump or something like that, but then, when the heavy shaking arrives, hopefully the alert would have been delivered and people would have been in position.”

DAS works on the same principle as ShakeAlert, only instead of seismometers monitoring for P-waves it uses vast spans of fiber optic cables. Scientists can get authorization to attach a device called an interrogator to unused cables. (Telecom companies often laid down more than they ended up needing.) This device fires laser pulses down the wire and analyzes tiny bits of light that bounce back when the fiber is disturbed. Because scientists know the speed of light, they can pinpoint disturbances based on the time it took for the signal to get back to the interrogator. 

Instead of taking seismic measurements at a single point, like a seismometer does, DAS is more like a miles-long string that forms one giant earthquake sensor. If there are a bunch of cables zig-zagging across a region, all the better. “One of the big advantages of DAS is actually a lot of those cables are already there, so it’s readily available,” says Sunyoung Park, a seismologist at the University of Chicago.

DAS may also be able to gather data where there aren’t any proper seismic stations, like rural areas that have fiber optic cables stretching out beneath them. Because those cables are also under the sea—running along coastlines and connecting continents across oceans—they can pick up earthquakes there too. For those longer spans, researchers use “repeaters,” devices already placed every 40 miles or so along the cables that boost signals. In this case, instead of analyzing the light that bounces back to an interrogator, they analyze the signal that reaches each repeater. 

Last year, scientists described how they used a cable stretching from the United Kingdom to Canada to detect earthquakes all the way down in Peru. The technique was so sensitive that the cable even picked up the motion of the tides, meaning it could potentially be used to also detect tsunamis spawned by underwater earthquakes.

And last month in the journal Scientific Reports, a separate team of researchers described how they used undersea cables off the coasts of Chile, Greece, and France to detect earthquakes. They compared this data to seismometer data that monitored the same events, and they matched well. “We can, in real time while the earthquake is happening, analyze the signals recorded using optical fibers and estimate the magnitude of the earthquake,” says Itzhak Lior, a seismologist at Israel’s Hebrew University and lead author of the paper. “The game changer here is we can estimate the magnitude every 10 meters along the fiber.” 

Because a traditional seismometer measures at a single point, it can get thrown off by localized data noise, like that caused by large vehicles rolling by. “If you have fibers, you can actually quite easily distinguish an earthquake from noise, because an earthquake is almost instantaneously recorded along hundreds of meters,” says Lior. “If it’s some local noise source, like a car or train or whatever, you only see it on a few tens of meters.”

Basically, DAS significantly bumps up the resolution of seismic data. That’s not to say that it would be a replacement for these highly accurate instruments—more of a complement to them. The overall idea is just to get more seismic detectors closer to earthquake epicenters, improving coverage. “In that sense, it doesn’t really matter if you have seismometers or DAS,” says Lior. “The closer you are to the earthquake, the better.”

And DAS research has a few challenges to contend with, notably that fiber optic cables weren’t designed to detect seismic activity—they were designed to shuttle information. “One of the issues with DAS cables is they’re not necessarily what we call ‘well coupled’ to the ground,” says Park, meaning the lines may just be laid loosely into piping, while a proper seismometer is finely tuned and situated to detect rumblings. Scientists are researching how a cable’s data-gathering might change depending on how it’s laid underground. But because there are so many miles of fiber optics out there, especially in urban areas, scientists have plenty of options. “Since it’s so dense, you have a lot of data to play with,” Park says.

Another obstacle, says geophysicist Ariel Lellouch, who studies DAS at Tel Aviv University, is that constantly firing laser pulses down fiber optics and analyzing what returns to interrogators creates an enormous amount of information to parse. “Just the sheer amount of data that you acquire, and the processing, means you’re going to need to do a lot of it probably on site,” says Lellouch. “Meaning, you cannot afford to upload all the data to the internet and then process it in some centralized location. Because by the time you upload, the earthquake would have been way, way past you.”

In the future, that processing might actually happen in the interrogators themselves—creating a network of continuously operating detectors. The same fiber optics that bring you the internet could well bring you precious seconds of extra warning to prepare for a quake.