EARTHQUAKE MAGNITUDE PREDICTION FROM SINGLE-STATION STRONG-MOTION WAVEFORMS USING A MULTI-BRANCH DEEP NEURAL NETWORK
Keywords
DOI:
https://doi.org/10.34238/tnu-jst.14095Abstract
Rapid estimation of earthquake magnitude from strong motion waveforms at a single station is important for early warning and real-time seismic hazard assessment. This study assesses the ability to predict earthquake magnitude directly from single-station waveforms by jointly using waveform information, event parameters and station information within a unified deep learning model. The proposed MagWaveNet model is designed as a multi-branch deep neural network, in which three-component strong-motion waveforms, source parameters, source-to-station geometric parameters and station indices are combined to estimate magnitude. Experimental results with several data partition strategies show that MagWaveNet achieves a root mean square error of 0.359, a mean absolute error of 0.282 and a coefficient of determination of 0.923 on the split by station and year, while consistently yielding lower errors than the traditional machine learning models used for comparison.Downloads
References
A. Navarro-Rodríguez, O. A. Castro-Artola, E. E. García-Guerrero, O. A. Aguirre-Castro, U. J. Tamayo-Pérez, C. A. López-Mercado, and E. Inzunza-Gonzalez, “Recent advances in early earthquake magnitude estimation by using machine learning algorithms: A systematic review,” Applied Sciences, vol. 15, no. 7, 2025, Art. no. 3492.
Y.-M. Wu and L. Zhao, “Magnitude estimation using the first three seconds P-wave amplitude in earthquake early warning,” Geophysical Research Letters, vol. 33, 2006, Art. no. L16312.
Y. Wang, X. Li, Z. Wang, and J. Liu, “Deep learning for magnitude prediction in earthquake early warning,” Gondwana Research, vol. 123, pp. 164-173, 2023.
J. Zhu, S. Li, J. Song, and Y. Wang, “Magnitude estimation for earthquake early warning using a deep convolutional neural network,” Frontiers in Earth Science, vol. 9, 2021, Art. no. 653226.
J. Zhu, S. Li, and J. Song, “Magnitude estimation for earthquake early warning with multiple parameter inputs and a support vector machine,” Seismological Research Letters, vol 93, no. 1, pp. 126-136, 2022.
S. M. Mousavi and G. C. Beroza, “A machine-learning approach for earthquake magnitude estimation,” Geophysical Research Letters, vol. 47, no. 1, 2020, Art. no. e2019GL085976.
B. Hou, Y. Zhou, S. Li, Y. Wei, and J. Song, “Real-time earthquake magnitude estimation via a deep learning network based on waveform and text mixed modal,” Earth, Planets and Space, vol. 76, no. 1, 2024, Art. no. 58.
R. Sun, W. Qi, T. Zheng, and J. Qi, “Explainable machine-learning predictions for peak ground acceleration,” Applied Sciences, vol. 13, no. 7, 2023, Art. no. 4530.
Y. Liu, Q. Zhao, and Y. Wang, “Peak ground acceleration prediction for on-site earthquake early warning with a deep learning model,” Scientific Reports, vol. 14, 2024, Art. no. 5485.
B. Derras, P.-Y. Bard, F. Cotton, and A. Bekkouche, “Adapting the neural-network approach to PGA prediction: An example based on the KiK-net data,” Bulletin of the Seismological Society of America, vol. 102, no. 4, pp. 1446-1461, 2012.
Y. Okada et al., “Recent progress of seismic observation networks in Japan — Hi-net, F-net, K-NET and KiK-net,” Earth, Planets and Space, vol. 56, pp. xv–xxviii, 2004.
