AI Model Predicts Earthquakes Before They Occur

Researchers have developed a new artificial intelligence algorithm that correctly predicted 70 percent of earthquakes a week before they happened in China during a seven-month trial. This raises hopes that the technology could one day be used to limit the impact of earthquakes on humans. and economies.

An artificial intelligence developed by researchers at the University of Texas at Austin has been trained to detect statistical bumps in real-time seismic data that researchers had associated with previous earthquakes.

The result was a weekly forecast in which the AI successfully predicted 14 earthquakes within about 200 miles of their estimated location and with almost exactly the calculated magnitude. It missed one earthquake and gave eight false warnings.

It is not yet known whether the same approach will work in other places, but the effort is a milestone in the research of AI-guided earthquake prediction.

“Earthquake prediction is the holy grail,” said Sergey Fomel, a professor in UT’s Office of Economic Geology.

“We are not yet close to making predictions anywhere in the world, but our achievement tells us that what we thought was an impossible problem is, in principle, solvable.”

The results of the study have been published in the Bulletin of the Seismological Society of America.

“You don’t see earthquakes coming,” said Alexandros Savvaidis, a senior scientist who directs the bureau’s Texas Seismological Network Program (TexNet) — the state’s seismic network.

“It’s a matter of milliseconds, and the only thing you can control is how prepared you are. Even at 70 percent, that’s a huge result and can help minimize financial and human losses, and has the potential to significantly improve earthquake preparedness worldwide.

The researchers said their method was successful by following a relatively simple machine learning approach. The AI was given a set of statistical properties based on the team’s knowledge of earthquake physics and then told to train on a database of five years of seismic records.

Once trained, the AI made its predictions by listening for signs of future earthquakes among the background sounds of the earth.

The researchers are convinced that in places with robust seismic monitoring networks, such as California, Italy, Japan, Greece, Turkey and Texas, AI could improve its success rate and narrow down its predictions to a few tens of kilometers.