A deep-learning algorithm could detect earthquakes by filtering out city noise

When applied to data sets taken from the Long Beach area, the algorithms detected significantly more earthquakes and made it easier to find out how and where they started. And when one of these applies to data In California, too, the 2014 earthquake in La Habra, the team found four times more seismic detections in the “denominated” data than the number officially recorded.

Implementing AI is not the only task in earthquake detection. Researchers at Penn State are training deep-learning algorithms to make accurate predictions of how upcoming earthquakes might be – a task that has puzzled experts for centuries. And members of the Stanford team had previously trained models to select a phase, or to measure the time of arrival of seismic waves within an earthquake signal, which can be used to estimate the location of an earthquake.

Paula Colmeizer, a seismologist at the Royal Holloway University in London, who was not involved in the study, says deep-learning algorithms are especially useful for seismic monitoring because they can carry the burden of human seismologists.

In the past, seismologists have looked at graphs produced by sensors that record the motion of the earth during an earthquake, and they will visually identify the pattern. Deep learning can make that process faster and more accurate, says Colemeizer.
“It simply came to our notice then [the algorithm] Working in a noisy urban environment is very useful, as noise in an urban environment can be a nightmare and can be very challenging, ”she says.

Similar Posts

Leave a Reply

Your email address will not be published.