Massive earthquakes don’t just move the ground, they make adjustments to the speed of light in Earth’s gravitational field. Now the researchers have computers trained to identify these tiny gravitational signalsdemonstrating how signals can be used to mark the location and size of a strong earthquake almost instantly.
It’s a first step toward creating an early warning system for the planet’s most powerful earthquakes, scientists report May 11 in Nature.
Such a system could help solve a thorny problem in seismology: how to quickly pin down the true magnitude of a massive earthquake immediately after it strikes, says Andrea Licciardi, a geophysicist at the Université Côte d’Azur in Nice, France. Without that capability, it is much more difficult to quickly and effectively issue potentially life-saving hazard warnings.
When big earthquakes break, the jolts and shudders send seismic waves through the ground that show up as large ripples on seismometers. But current detection methods based on seismic waves notoriously have difficulty distinguishing between, say, a magnitude 7.5 earthquake and a magnitude 9 earthquake in the few seconds after such an event.
That’s because initial magnitude estimates are based on the height of seismic waves called P waves, which are the first to arrive at monitoring stations. However, for the strongest earthquakes, those initial P-wave amplitudes peak, making earthquakes of different magnitudes difficult to tell apart.
But seismic waves are not the first signs of an earthquake. All that mass moving around in a big earthquake also changes the density of the rocks in different places. Those changes in density translate into small changes in the Earth’s gravitational field, which produces “elastogravity” waves that travel through the ground at the speed of light, even faster than seismic waves.
Such signals were once thought to be too small to detect, says seismologist Martin Vallée of the Institut de Physique du Globe in Paris, who was not involved in the new study. Then, in 2017, Vallée and his colleagues were the first to report seeing these elastogravity signals on data from seismic stations. Those findings showed that “you have a window between the start of the earthquake and the moment you receive the [seismic] waves,” says Vallée.
But the researchers still wondered how to turn these elastogravity signals into an effective early warning system. Because gravity ripples are minute, they are difficult to distinguish from background noise in seismic data. When scientists looked back, they found that only six mega-earthquakes in the last 30 years generated identifiable signs of elastogravity, including the magnitude 9 one. Tohoku–Oki earthquake in 2011 that produced a devastating tsunami that flooded two nuclear power plants in Fukushima, Japan (Serial number: 03/16/11). (The initial AP wave-based estimate of the magnitude of that earthquake was 7.9.)
That’s where computers can come in, says Licciardi. He and his colleagues created PEGSNet, a machine learning network designed to identify “point signals of elastogravity.” The researchers trained the machines on a combination of real seismic data collected in Japan and 500,000 simulated gravity signals for earthquakes in the same region. Synthetic gravity data is essential for training, says Licciardi, because real data is so sparse and the machine learning model requires enough information to be able to find patterns in the data.
Once trained, the computers were given a test: to track the origin and evolution of the 2011 Tohoku earthquake as if it were happening in real time. The result was promising, says Licciardi. The algorithm was able to accurately identify both the magnitude and location of the earthquake five to 10 seconds earlier than other methods.
This study is a proof of concept and hopefully the basis for a prototype early warning system, says Licciardi. “Right now, it’s designed to work… in Japan. We want to build something that can work in other areas known for strong earthquakes, including Chile and Alaska. Eventually, the hope is to build a system that can work globally.
The results show that PEGSNet has the potential to be a powerful tool for earthquake early warning, particularly when used in conjunction with other earthquake detection tools, Vallée says.
Still, more work needs to be done. For one thing, the algorithm was trained to find a single point for the origin of an earthquake, which is a reasonable approximation if you’re far away. But up close, the origin of an earthquake no longer looks like a point, it’s actually a larger region that has ruptured. If scientists want an accurate estimate of where a breakup occurred in the future, the machines need to look for regions, not points, Vallée adds.
Bigger advances could come in the future as researchers develop much more sensitive instruments that can detect even smaller earthquake-caused disturbances in Earth’s gravitational field while filtering out other sources of background noise that could obscure the signals. Earth, Vallée says, is a very noisy environment, from its oceans to its atmosphere.
“It’s a bit the same as the challenge physicists face when trying to observe gravitational waves,” Vallée says. These ripples in space-time, caused by colossal cosmic collisionsare a very different type of gravity-driven wave (Serial number: 02/11/16). But gravitational wave signals are also dwarfed by noise from Earth, in this case, microquakes on the ground.