Early Warning for Seismic Events?

Caltech scientists use deep learning techniques to detect earthquakes.

Scientists at the Caltech Seismology lab are using deep learning technology to develop an early warning system for earthquakes. The research is based on the idea that electronic communication happens faster than seismic waves can pass through the ground. If you detect an earthquake at the point where it occurs, you have a few seconds to notify surrounding areas that the quake is coming. Those few seconds aren’t enough for a major evacuation or human intervention, but they could be enough to stop trains or shut down power to power lines, which could help to reduce the damage caused by the quake.

According to the announcement, the deep learning models “...use convolutional neural networks to look at a single sensor at a time to identify seismic waves, narrowing down the sensor’s datastream to a handful of discrete times with seismic activity. A second model, a recurrent neural network, recognizes wave patterns from several sensors over the course of a seismic event. The system unscrambles events that include multiple earthquakes in quick succession, and can reduce false triggers by a factor of 100 — greatly improving the reliability for early warning systems.”