Stanford AI Detection System Could Predict Earthquakes

Stanford University's Earthquake Transformer AI detects seismic activities.
Chris Young
The photo credit line may appear like thisenot-poloskun/iStock

A group of researchers unveiled a new method for using artificial intelligence (AI) to enhance our ability to read seismic waves and, in doing so, improve our understanding of how they begin, and even how they come to a stop.

Published in Nature Communicationsthe paper details a method that automates earthquake detection at the same time as tuning out much of the noise inherent to seismic data.

RELATED: 11 OF THE DEADLIEST EARTHQUAKES OF ALL TIME

AI earthquake detection

Mostafa Mousavi and a team of researchers use artificial intelligence to focus on millions of tiny subtle shifts in the Earth's crust. They hope that these tiny movements might act as a Rosetta Stone of sorts for deciphering warning signs for big earthquakes.

"By improving our ability to detect and locate these very small earthquakes, we can get a clearer view of how earthquakes interact or spread out along the fault, how they get started, even how they stop,” Stanford geophysicist Gregory Beroza, one of the paper’s authors, explained in a Stanford University press release.

The team has developed several machine learning systems for earthquake detection. Amongst these is CRED, developed in 2019, which was inspired by voice-trigger algorithms in virtual assistant systems. 

The new paper details the team's most recent iteration, a model that detects very small earthquakes with weak signals that are usually overlooked by current methods. They call their new system Earthquake Transformer. The system uses an "attention mechanism" to pour over large amounts of data and hone in on the most important elements.

Anticipating earthquakes in the future

In order to test the Earthquake Transformer, the team trained their algorithm on data that included one million hand-labeled seismograms recorded over the past two decades globally, excluding Japan. For the test, they then selected five weeks of continuous data recorded in Japan at the time of the magnitude-6.6 Tottori earthquake and its aftershocks from 20 years ago.

During the test, the model detected and located 21,092 events – more than two and a half times the number of earthquakes that had been picked out by hand. What's more, Earthquake Transformer used data from only 18 of the 57 stations that Japanese scientists originally used to study the sequence.

According to Beroza, the system is ready to be used for real-time detection of earthquakes.

"Earthquake monitoring using machine learning in near real-time is coming very soon," Beroza explained. "The more information we can get on the deep, three-dimensional fault structure through improved monitoring of small earthquakes, the better we can anticipate earthquakes that lurk in the future." 

BE THE FIRST COMMENTER
Already have an account? Log in