A Real-Time Seismic Intensity Prediction Framework Based on Machine Learning for Earthquake Early Warning

Abstract

Real-time estimation of seismic intensity provides a critical basis for earthquake early warning (EEW) and emergency response. Most current methods estimate seismic intensity by relying only on one or a few parameters extracted from the P-wave without systematic feature selection, potentially neglecting essential characteristics from the ground motion (GM) records, such as amplitude and frequency-domain features. To address this issue, a real-time seismic intensity prediction framework based on machine learning is proposed to predict seismic intensity based on GM records obtained from the Kyoshine Network (K-Net) from 2010 to 2018. A sorting importance technique is proposed to assess 24 candidate parameters to eliminate redundancy and identify an optimal parameter combination, enhancing both model accuracy and computational efficiency. To validate the accuracy and generalization ability of the proposed method, earthquakes occurring between 2019 and 2021 are assumed to be new events and used as a verification case dataset. The results demonstrate that the optimized models achieved an accuracy above 90% for time windows of 3 s and longer, while lower accuracy and larger dispersion were observed in the first 1–2 s after P-wave arrival. Compared with models trained using all candidate parameters, the optimized models improve the efficiency of feature calculation, model training, and prediction by approximately factors of two, ten, and two, respectively, while maintaining comparable prediction performance. Finally, an actual seismic event is selected, and three earthquake early warning (EEW) decision thresholds are defined to further evaluate the model performance. These results highlight the promising potential of the proposed method for predicting seismic intensity during emergency response in real time.