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

When applied to the data sets taken from the Long Beach area, the algorithms detected more earthquakes and made it easier to figure out how and where they started. The team observed four times more seismic detections in the data compared to the official number.

It's not the only work that uses artificial intelligence to find earthquakes. Researchers from Penn State have been training deep-learning software to accurately predict earthquakes, a task that has perplexed experts for centuries. Members of the team trained models for phase picking, or measuring the arrival times of seismic waves within an earthquake signal, which can be used to estimate the quake's location.

Paula Koelemeijer, a seismologist at Royal Holloway University of London, who was not involved in the study, says that deep- learning is useful for earthquake monitoring because it can take the burden off human seismologists.

Seismologists used to look at graphs produced by sensors that record the motion of the ground during an earthquake and identify patterns by sight. Koelemeijer says that deep learning could help cut through large volumes of data and make that process quicker and more accurate.