Science and TechNatural Disasters


Predicting earthquakes is unlikely, but could we detect them earlier?

Some scientists are pushing more research into early earthquake detection, while others believe we should first fix our building and alert systems.
Posted at 8:25 PM, Mar 01, 2023

After a major earthquake, like the ones that recently hit Turkey and Syria, one question often looms: How did we not see this coming?

A judge from L'Aquila, Italy, even sentenced six scientists and a government official to several years in prison for the crime of not properly warning L'Aquila citizens of a devastating earthquake that would go on to kill hundreds. This was after the team assessed smaller tremors in the area, and one official said a bigger quake was "improbable." The scientists were charged with manslaughter.

Although some argue the conviction was mostly for poor communication, others felt the judge misunderstood the role of the scientists. After the original indictments, nearly 4,000 scientists from around the world signed a letter of protest to the president of Italy. The letter argued the team was charged for failing to do the scientifically impossible: predict earthquakes. 

Seven years later, all convictions were eventually appealed and overturned. But the underlying debate remains: Is there still really no reliable way to predict earthquakes and no better way to warn citizens ahead of time?

Let's look at why earthquakes are so hard to predict in the first place.

The tectonic plates underneath the Earth's surface are always moving, but because of the immense weight on the plates, there is a huge amount of friction along the boundaries. The friction gets the plates stuck together, while pressure builds as they continue moving. Eventually, the pressure overcomes the friction, and the plates snap out of place and move suddenly. This sends out a ripple effect through the layers of rock above — in other words, an earthquake.

While we can detect these ripples once they're sent out, we can't predict that initial moment when the friction holding the plates together gives way.

"I wrote a book saying, basically saying, 'We can't predict earthquakes,'" said Susan Hough, a seismologist and lead scientist with the United States Geological Survey. "One of my colleagues took me to task and said, 'We can predict earthquakes in the way that matters. We know where they're expected, likely to happen on average.' 

Hough also led the USGS team installing earthquake monitoring equipment and reporting methods in Haiti after the catastrophic 2010 earthquake there. 

Here's a look at the deadliest earthquakes in the past 500 years
A man sits in the rubble of collapsed collapsed building in Antakya, Turkey

Here's a look at the deadliest earthquakes in the past 500 years

Some earthquakes are never felt and others have devastating and deadly consequences.


"People do look at the statistics of earthquake triggering by looking at all the earthquakes that have happened in the given region over time," Hough said. "You can say, 'We just had a five in this area. What's the probability that we're going to have another earthquake larger than five?'"

For example, the USGS can estimate there is a 72% chance that an earthquake with a magnitude of 6.7 or greater will hit the San Francisco Bay Area sometime over the next 30 years. This is how we end up with forecasts of quakes like the so-called "Big One" expected to hit the West Coast of the U.S. but can't predict when exactly it's going to hit.

"We usually talk about hazard and numbers over 30 years because that's a meaningful length of time for people," Hough said. "That information feeds into building codes, and with proper building codes, buildings can withstand pretty much the most severe shaking that that we're ever going to experience."

In this debate, there seems to be two general approaches: There's one camp that believes predicting earthquakes is basically a scientific pipe dream and distracts from the more attainable goal, which is improving our engineering and alert systems. But another believes that with proper research and experience, we'll be able to eventually detect a quake much earlier, giving evacuations a crucial head start.

And there may be new methods of prediction we just haven't fully considered yet.

For example, some early research has found a correlation between electrical disturbances and ionization in the atmosphere with an incoming earthquake. But there's still some doubt over whether the phenomena are truly related.

Back down on Earth, some scientists are even looking to more ancient methods of sensing earthquakes.

Humans may be stumped, but throughout history, various animals have shown signs they can sense an earthquake is coming — at least on certain occasions. 

How the strongest earthquakes can rewrite our maps
An uplifted fault in New Zealand

How the strongest earthquakes can rewrite our maps

When the shaking and tsunamis subside, the coastline can look a bit different.


One of the earliest recorded references to this was when the Greek historian Thucydides wrote of rats, snakes, weasels, and dogs fleeing the city of Helice days before a huge earthquake.

In 1805, animals reportedly panicked before a major quake in Naples. A century later, horses allegedly fled before the San Francisco quake. Then, in 1975, Chinese authorities evacuated a city partially because of the sudden change in behavior from snakes and other animals, saving lives before a major quake struck.

It's far too soon to say whether there's a link here; there are just as many earthquakes without unusual animal behavior beforehand. But some researchers at the Max Planck Institute for Animal Behavior did try to put this to the test. 

In a recent study on farm animals in Italy over several years, researchers found they could detect unusual behavior patterns as early as 20 hours before an earthquake. The closer the animals were to the epicenter, the earlier their behavior changed.

Still, many experts would note it's more effective to focus on preparation than chasing the dream of earlier predictions. Because where science stands now, it's still impossible to guess when exactly an earthquake is going to hit. 

"People want earthquakes to be predicted because it's terrifying when they happen out of the blue," Hough said. "Whether or not they're predicted and whether or not you get 30 seconds of warning, life safety demands, building safety, you're building a house to withstand the shaking, and 10 seconds of warning isn't going to help you if your building collapses. Forget short-term prediction; figure out how buildings need to be built to withstand the earthquakes that we know are going to happen eventually."