QUAKESENSE: AI-DRIVEN RAPID SOURCE DETECTION FOR RELIABLE EARTHQUAKE EARLY WARNING SYSTEMS

Abstract

Accurate and timely estimation of an earthquake’s source location is essential for effective early warning and disaster mitigation. Traditional seismic analysis techniques, though scientifically sound, often suffer from latency and uncertainty in real-time applications due to the complexity of waveform interpretation and reliance on dense sensor networks. This study introduces QuakeSense, an AI-driven system that employs machine learning algorithms to rapidly and reliably estimate earthquake source locations using limited initial seismic data. The proposed approach integrates neural networks and ensemble models to analyze waveform features, predict epicentral coordinates, and continuously refine estimates as new signals arrive. By leveraging data from seismic stations and synthetic simulations, QuakeSense achieves a significant reduction in processing time while maintaining high localization accuracy. The model’s robustness and adaptability demonstrate that machine learning can revolutionize earthquake early warning systems by enhancing both speed and reliability, thereby contributing to reduced risk and improved public safety