Machine Learning-Driven Earthquake Early Warning Using Optical Fiber Mesh Networks

Abstract:
We demonstrate interconnected meshed optical networks as sensing-localization grid for earthquake early detection. We integrate ground displacement data from seven local earthquakes, magnitudes four to six, into a Waveplate model to enhance a machine-learning algorithm that accurately detects primary waves and improves nearby urban safety.