Generalized Neural Networks for Real-Time Earthquake Early Warning

Abstract: Deep learning enhances earthquake monitoring capabilities by mining seismic waveforms directly. However, current neural networks, trained within specific areas, face challenges in generalizing to diverse regions. Here, we employ a data recombination method to create generalized earthquakes occurring at any location with arbitrary station distributions for neural network training. The trained models can then be applied to various regions with different monitoring setups for earthquake detection and parameter evaluation from continuous seismic waveform streams. This allows real-time Earthquake Early Warning (EEW) to be initiated at the very early stages of an occurring earthquake. When applied to substantial earthquake sequences across Japan and California (US), our models reliably report earthquake locations and magnitudes within 4 seconds after the first triggered station, with mean errors of 2.6-6.3 km and 0.05-0.17, These generalized neural networks facilitate global applications of realtime EEW, eliminating complex empirical configurations typically required by traditional methods. However, as the workflow involving many steps, the error in either of the steps would potentiality affect the final location and magnitude results, and they require more time to receive and analyze the earthquake signals. Unlike the automatic construction of earthquake catalogs, EEW systems require the rapid reporting of earthquake parameters at the very early stages of occurrence, often with limited information available from network stations.