Geologists show why we still have no idea how to predict earthquakes

At 3:00 a.m., when everyone was sleeping, a violent earthquake struck 65 miles from Turkey’s capital in August 1999. The 7.6-magnitude quake collapsed entire apartment buildings on top of unsuspecting people, . 

Known as the Izmit quake, in 2011 scientists said they found of a precursor to the tragic temblor: A series of often small, unnoticeable quakes — known as foreshocks — occurred in the same area for around 44 minutes before the major quake struck. This, the scientists argued, was evidence of something unique happening before the massive earthquake.

And if this foreshock activity could somehow be reliably detected, perhaps we could predict future earthquakes with plenty of notice for people in any given area — an ability that has completely eluded scientists.

But seven years later, earthquake researchers have put a damper on this optimism. 

In a study Monday in Nature Geoscience, scientists found that these series of small quakes were not a unique sign of a looming earthquake, but instead a domino-effect of tremors that happened to have ultimately triggered a far bigger one.

“They’re knocking down the idea that something happens before earthquakes,” Duncan Agnew, a geophysicist at Scripps Institution of Oceanography at the University of California, San Diego, said in an interview.

Foreshocks are earthquakes, emphasizes Angew, but they only earn the title “foreshock” when they happen before a big one. 

A leveled apartment building in Istanbul in November 1999.

“They are something you can only see in retrospect,” Angew, who had no role in the study, said. “There’s absolutely no way to tell them apart otherwise. That’s like trying to predict the weather, except you can’t see anything.”

“For decades, people have been thinking about how we can distinguish foreshocks from ordinary earthquakes,” William Ellsworth, a Stanford University geologist and lead author of the study, said in an interview. “They [2011 researchers] made this rather bold suggestion that the foreshocks were different — we just came up with the opposite result.”

By analyzing additional measurements from the 1999 quake, Ellsworth and coauthor Fatih Bulut, a geologist at Turkey’s Kandilli Observatory and Earthquake Research Institute, found that a series of foreshocks before the quake did not all happen in one place, as previously suggested. 

Instead, “we found that none of them were at the same spot,” said Ellsworth, adding that more than two dozen imperceptible quakes “popped off all over the place.” 

From this, the study’s authors conclude that nothing special was happening, but a “cascade-like” effect of small earthquakes rippled around the area, until one happened to have triggered massive underground movement — and, ultimately, a tragedy. 

The fault comparison between the infamous San Andreas fault in California, and the fault in Turkey.

However, perhaps there’s something else still happening before earthquakes that geologists haven’t yet discovered, or been able to measure.

“It’s possible there are clues out there that we don’t know yet,” John Vidale, Director of the Southern California Earthquake Center who had no role in the study,” said in an interview. “We don’t know what we don’t know.” 

Or, there could be no clues at all — no true, precursor of a coming quake.

“It may not be there — that’s the thing we don’t know,” said Ellsworth. 

But even so, we do have a solid grasp of areas that are earthquake prone, and when a region is overdue for a large quake. 

Digging into the ground along fault lines — cracks in the Earths’ crust where pressure builds and eventually slips — gives scientists a good understanding of how often big quakes have happened in the past, and whether or not we’re due for another one.

“If it’s been a long time since the last one, the odds are somewhat higher now,” said Vidale. 

Though, “somewhat” is the key word here. Geologists know we can’t predict quakes, but are well aware they can hit at any time — sometimes with small foreshocks, and sometimes without.

“If it happens in the next 20 minutes I won’t be surprised,” said Angew, of the next big rupture of the San Andreas fault in California. “If it doesn’t happen in my lifetime, I won’t be surprised.”

A boy stands in front of his destroyed home in Adapazari, Turkey, in 1999.

In the end, while earthquake prediction is a much-desired ability, perhaps it doesn’t matter as much as we think it does. 

“It really doesn’t matter when they happen if we’re prepared,” said Ellsworth. “We’ll get through them.”

For those that live in quake country, this means heeding the advice of the U.S. Geologic Survey. 

And on a larger scale, it also means building structures that are safe during earthquakes. In the case of Turkey two decades ago, many buildings simply weren’t built to withstand a big quake. 

“Turkey has very good building codes,” said Ellsworth. “The problem is buildings were not being built to the code.” 

“Earthquakes don’t kill people, buildings kill people,” he added.

Scientists will still keep searching for the elusive clues that might foreshadow a looming quake, perhaps giving people 10 or 20 minutes to prepare. 

Today, the best we have are early warning systems, like that being tested in California and in use in Mexico City. These don’t predict earthquakes, but can give people perhaps 30 seconds notice that a big quake has happened, and that seismic waves will soon hit their area. 

For now, preparation isn’t just the smart option — it remains the only option. There’s much that isn’t known about what happens before earthquakes, and that isn’t likely to change soon.

“What we want to know about is buried under miles of solid rock,” said Ellsworth.