Earthquake and tsunami prediction enhanced by deep learning model


A deep learning model developed at Los Alamos National Laboratory could help estimate earthquake magnitude in real time and provide earlier warnings of tsunamis.

A team of scientists from Los Alamos National Laboratory is working on a solution to predict natural disasters, starting with earthquakes and tsunamis. To achieve this, they developed a deep learning model capable of capturing the gravity waves generated by an earthquake and thus predicting the risk of an ensuing tsunami.

“Our model unlocks real-time estimation of earthquake magnitude, using data regularly treated as noise, and can immediately transform early warning of tsunamis,” said group scientist Bertrand Rouet-Leduc. of Los Alamos geophysics.

Rapid and reliable estimation of the magnitude of large earthquakes is crucial to mitigating the risks associated with strong shaking and tsunamis.

Currently, the systems used to detect earthquakes are based on seismic waves, which makes them unable to quickly estimate the size of large earthquakes. Since the estimates are based on the shaking produced by an earthquake at the time it occurs, they cannot prevent the consequences of the disaster until it is already too late.

Moreover, standard systems cannot distinguish between magnitude 8 and magnitude 9 earthquakes, although the latter are 30 times more energetic and destructive. Even approaches that rely on GPS to estimate earthquake magnitude are often subject to large uncertainties and latency issues.

Unlike seismic-based early warning, the Prompt Elasto-Gravity Signals (PEGS) speed-of-light approach designed by the Rouet-Leduc team relies on gravity waves. This new method saturate in magnitude and can immediately distinguish between magnitude 8 and magnitude 9 earthquakes. The PEGS approach, in fact, becomes more accurate the larger the earthquake. This method had never been tested before for earthquake early warning.

The Los Alamos research team has shown that PEGS can be used in real time to track the growth and magnitude of an earthquake immediately after it reaches a certain size. The team developed a deep learning model that exploits the information carried by PEGS, which is recorded by regional broadband seismometers in Japan.

After training the deep learning model on a database of synthetic waveforms augmented with empirical noise measured on the seismic network, the team was able to show the first example of instant tracking of a seismic source on data. real.

This model, combined with real-time data, can alert communities much earlier if a subduction mega-earthquake is large enough to create a tsunami that will breach levees, potentially saving many thousands of lives.

Sign up for the E&T News email to get great stories like this delivered to your inbox every day.

Source link


Comments are closed.