It is estimated that for a tsunami to be caused by an earthquake, the latter must be of a strong magnitude, at least 6.5 on the Richter scale. A team of researchers from IRD, CNRS, Université Cotete d’Azur, Observatoire de la Cote d’Azur, Los Alamos National Laboratory and a team from the University of Kyoto have used artificial intelligence to make instantaneous predictions using artificial intelligence. PEGS). The title of their research Instant Tracking of Earthquake Growth with Elastogravity Signals. Published May 11 in the journal Nature.
Fortunately, tsunamis are rare natural disasters, but they can kill many people living in coastal areas or island states. Thus, the tsunami that followed the 9.1 magnitude earthquake in the Indian Ocean on December 26, 2004 affected 14 countries, including India, Indonesia, Sri Lanka and even Thailand, leaving an estimated 230,000 dead and missing.
Warning systems based on seismic waves have been put in place to limit the human and material damage caused by the disaster. The epicenter was reported below the ground, however; no tsunami alert was issued. The epicenter was reported below the ground, however; no tsunami alert was issued.
However, these precautionary measures fail to quickly estimate the magnitude of a very large earthquake. Thus, the Japanese precautionary system estimated 8 magnitudes instead of 9 during the 2011 Tokhoku (commonly known as Fukushima earthquake) earthquake off the Pacific coast, predicting 3-meter waves instead of 15, with an error (18,079) dead or missing with dramatic consequences. ). According to the research team, geodesy-based methods allow better predictions, but also the greater uncertainty and delay associated with slow seismic waves.
The team showed that it is possible to use gravitational signals (PEGS), which, although very weak, were discovered in 2017 with data from the 2011 earthquake, to instantly estimate the magnitude of a large earthquake.
PEGS (Prompt Elasto-Gravity Signal) is a gravitational wave created by the massive movement of rocks during a large earthquake. These signals propagate at the speed of light, much faster than seismic waves and thus give the population extra time to defend themselves.
The very low amplitude of PEGS has so far made it impossible to use them in alert systems. Researchers have solved this problem using an AI algorithm. They have created a model of PESGNet Acquiring deep knowledge, CNN, which uses data provided by PEGS recorded by a regional broadband seismometer in Japan before the arrival of seismic waves. After training on a database of 500,000 synthetic waveform data enhanced with empirical sound, the algorithm can instantly track a time function of the seismic source into real data. According to researchers, “ Our model uses a portion of the seismogram to unlock “real-time” access to the crater evolution of large earthquakes that are routinely treated as sound and can be immediate converters for pre-tsunami warning. A
Andrea Licciardi, geospatial geophysicist and first author of the study, says:
“ Tested in Japan, the algorithm proved to be able to predict the intensity of the Fukushima earthquake faster and more reliably than all existing systems without using seismic waves. A
Quentin Bletery, project initiative and adds to the same unit:
“ Operational warning systems have yet to be implemented, but our results indicate that PEGS can significantly improve tsunami warning systems. A
Source of the article: Liquierdi, A., Blattery, Pro., Ruet-Leduk, B. Etc. Instantaneous tracking of earthquake magnitude with elastogravity signal. Nature (2022). https://doi.org/10.1038/s41586-022-04672-7.