The existing optical fiber terrestrial network can be leveraged to serve as a wide distributed network of sensors, especially to detect mechanical stresses as the optical signal polarization is significantly influenced by external disturbances. Exploiting this trend, paves the way for employing the optical fiber network in environmental sensing, like detecting earthquakes or tracking anthropic activities. The purpose is to examine the changes in the state of light polarization caused by birefringence induced by seismic events. Consequently, we have developed a Python-based Waveplate Model to track state of light polarization changes in buried optical fiber cables. This model integrates real ground motion data from a 4.9 magnitude earthquake that occurred southwest Marradi city in Italy, and converts it into strain values along the fiber cable. To further investigate the effects of this particular seismic activity, we propose a centralized smart grid fiber network approach based on a neural network model with an attention mechanism for earthquake early warnings. Along with the aforementioned Waveplate Model, numerous sets of polarization evolution were produced on two distinct sensing points with different distances from the epicenter in two different cities, enabling earthquake early detection upon P-wave arrivals that precede the earthquake’s destructive surface waves and allowing for a swift initiation of emergency plans including early warning alerts and earthquake countermeasures.
|