Predicting earthquakes is one of the holy grails of the geosciences. And researchers are getting creative in their efforts to predict future tremors—from the use of data mapping to radon gas to the behavior of toads. But they've been met with varying degrees of success, and have yet to come up with a foolproof method.
Now, as Paul Voosen at Science reports, scientists are proposing a new idea: use tiny changes in the Earth's rotation to predict temblors magnitude 7 and higher. The most remarkable part of the proposed method is that it predicts quakes five years ahead of time.
The idea was recently presented by Roger Bilham of the University of Colorado at Boulder and Rebecca Bendick at the University of Montana in Missoula at a meeting of the Geological Society of America. They began their work examining the repeated buildup and release of forces on an earthquake fault, discovering it resembles an oscillator—a system that releases energy over time, like a clock pendulum or even heart beat. So they drilled into the data for major earthquakes over the last 100 years to see if they could pick out patterns.
What they found is that earthquakes cluster through time, but their location around the world was seemingly random. And overall, activity peaked on a 32-year cycle. They published their results in August in the journal Geophysical Research Letters.
In their new presentation, the researchers connect this 32-year cycle of peak earthquakes, which also includes shorter five-year earthquake cycles, with another cyclical phenomenon: the millisecond slowdown of the Earth's rotation.
Though researchers aren't sure exactly why, every few years the Earth's rotation decelerates by several milliseconds—with the peak slowdown occurring every 30 years or so, reports Voosen. The researchers found that this change, which affects the length of the day, seems to correspond to the uptick in earthquakes five years after the deceleration. Bilham and Bendick propose that those two cycles could be powered by the same geologic process, and might be related to one another.
As Bilham tells Kate Sheridan at Newsweek, there's a possibility that instability in the Earth's rotating core might cause these slow downs in rotation. And those tiny changes could cause the Earth’s tectonic plates to bump into one another, similar to a the way people on a subway car or bus bump into each other when it stops abruptly.
"[The core is] very unstable. And these instabilities, over the course of decades, are able to influence the rotation rate of the Earth,” Bilham tells Sheridan. “So what it means is that the core may well be responsible for modulating earthquakes at the Earth's surface. It doesn't produce them, but it looks as though the core may indirectly influence the motion of the plates by changing the rotation rate.” But it takes time for the impacts of the rotation changes to take effect. In their model, the bumping of the plates that triggers earthquakes doesn’t happen until five years after the millisecond deceleration occurs.
Bilham admits the idea might be a little crazy. But other researchers agree that the concept is compelling. “It might be nonsense,” Michael Manga of the University of California, Berkeley tells Voosen, but it’s not out of the question. “I’ve worked on earthquakes triggered by seasonal variation, melting snow. His correlation is much better than what I’m used to seeing.”
If it turns out to be true, it will be a big step forward in Earthquake casting. “Five years is a great amount of warning,” Bilham tells Sheridan. “Earthquake prediction is known to be either impossible or beyond our means. But here is the Earth saying—perhaps—you're in for something more. It’s now time to do something about it.”
We won’t have to wait too long to see if Bilham and Bendick are right. As Voosen reports, the last millisecond slowdown of the Earth’s rotation took place four years ago, meaning that 2018 might have more major quakes than expected.