There is no way to spot big earthquakes ahead of time

All earthquakes look the same when they start, making it unlikely we will be able to predict which will cause the most devastation from early observations.

Early warning systems rely on seismometers picking up tremors and sounding the alarm for nearby cities before major shaking starts. Even a few seconds’ warning can make a lot of difference, both for individual people and for organisations like hospitals, says Daniel Trugman at the Scripps Institution of Oceanography in La Jolla, California.

For example, Mexico’s early warning system gave everyone a 10 to 15 second heads-up before Tuesday’s magnitude 7.1 earthquake, says Men-Andrin Meier at the California Institute of Technology in Pasadena.

The trouble is, such warning systems must estimate the size of an earthquake from the earliest seismometer readings. By looking at past earthquakes, Meier had hoped to show that large and small earthquakes started differently, so that the systems could spot the really big ones.

“Unfortunately, our results show that such hopes are not realistic,” he says.

Tremors spreading

While quakes occur suddenly, they don’t reach their maximum strength and spread all at once. Instead, they start in a small area and then spread along a plate boundary at several kilometres per second. These movements determine how wide an area is affected, and how badly.

Earthquakes grow until they hit a peak and then decay, in a fairly symmetrical fashion, says Meier. Most models assume that this growth in strength accelerates after the start of the earthquake, he says. But when he and his colleagues analysed 116 of the largest earthquakes from the past three decades, they found the quakes all grew at steady rates, regardless of their eventual size. Different-sized earthquakes were statistically indistinguishable until they reached their peak intensity.

But Meier did find a way to estimate a minimum final size for a quake. The final size of a growing earthquake will be statistically at least twice its current size, he says. So an earthquake that reaches magnitude 7.2 during its growing phase will at least double to a magnitude 7.4 before it peaks.

Despite such progress in detecting seismic events and assessing the risk to particular areas, long-term earthquake prediction hours or days ahead is still a long way off. Seismologists know roughly where earthquakes are likely to occur, but cannot predict when and how they will go off.

“Last Tuesday, nobody knew a 7.1 earthquake would hit Mexico in one week’s time,” says Meier. “Anyone who says they knew that is a charlatan.”

Read more: Italy to test earthquake predictor to reduce deaths and damage