Earthquakes pose significant threats to human life and property, often triggering secondary disasters such as landslides, mudslides, and collapses. Although earthquake prediction remains a challenging global scientific problem, the study of seismic precursors plays a critical role. Earthquakes occur due to the instability and fracturing of underground rock layers when concentrated stress exceeds their strength limit. Borehole strainmeters provide direct observations of crustal strain, making pre-earthquake strain anomaly detection essential for precursor studies. In recent decades, variational autoencoders (VAEs) have been widely adopted for anomaly detection due to their powerful denoising capabilities. However, traditional VAE-based methods face difficulties in capturing both long-term heterogeneous patterns and fine-grained short-term trends simultaneously. To address this, we propose a new approach combining convolutional neural networks (CNN) with frequency-aware conditional VAE frameworks. The CNN extracts spatial dependencies among strain observation components, while frequency analysis improves temporal feature capture. By incorporating a target attention mechanism, our model selects the most relevant frequency-domain information to enhance reconstruction of both long-term and short-term trends. Experimental results on borehole strain data show that our model outperforms state-of-the-art methods. These findings confirm the practical value of our approach in overcoming current VAE-based detection limitations and emphasize the importance of integrating spatial and frequency representations in seismic precursor studies.

1. [1] D. K. Keefer, “Landslides caused by earthquakes,” Geological Society of America Bulletin, Vol.95, No.4, pp. 406-421, 1984. https://doi.org/10.1130/0016-7606(1984)95<406:LCBE>2.0.CO;2
2. [2] R. J. Geller, “Earthquake prediction: A critical review,” Geophysical J. Int., Vol.131, No.3, pp. 425-450, 1997. https://doi.org/10.1111/j.1365-246X.1997.tb06588.x
3. [3] S. Pulinets and V. M. V. Herrera, “Earthquake precursors: The physics, identification, and application,” Geosciences, Vol.14, No.8, Article No.209, 2024. https://doi.org/10.3390/geosciences14080209
4. [4] D. C. Agnew, “Strainmeters and tiltmeters,” Reviews of Geophysics, Vol.24, No.3, pp. 579-624, 1986. https://doi.org/10.1029/RG024i003p00579
5. [5] P. G. Silver and H. Wakita, “A search for earthquake precursors,” Science, Vol.273, No.5271, p. 77, 1996. https://doi.org/10.1126/science.273.5271.77
6. [6] G. Igarashi, S. Saeki, N. Takahata et al., “Ground-water radon anomaly before the Kobe earthquake in Japan,” Science, Vol.269, No.5220, pp. 60-61, 1995. https://doi.org/10.1126/science.269.5220.60
7. [7] E. Hauksson and J. G. Goddard, “Radon earthquake precursor studies in Iceland,” J. of Geophysical Research: Solid Earth, Vol.86, No.B8, pp. 7037-7054, 1981. https://doi.org/10.1029/JB086iB08p07037
8. [8] Z. Z. Darban, G. I. Webb, S. Pan, C. Aggarwal, and M. Salehi, “Deep learning for time series anomaly detection: A survey,” ACM Computing Surveys, Vol.57, No.1, pp. 1-42, 2024. https://doi.org/10.1145/3691338
9. [9] H. Kanamori, “Earthquake Prediction: An Overview,” Int. Geophysics, Vol.81, pp. 1205-1216, 2003. https://doi.org/10.1016/S0074-6142(03)80186-9
10. [10] C. Chi, K. Zhu, Z. Yu, M. Fan, K. Li, and H. Sun, “Detecting earthquake-related borehole strain data anomalies with variational mode decomposition and principal component analysis: A case study of the Wenchuan earthquake,” IEEE Access, Vol.7, pp. 157997-158006, 2019. https://doi.org/10.1109/ACCESS.2019.2950011
11. [11] D. P. Kingma, “Auto-encoding variational bayes,” arXiv:1312.6114, 2013.
12. [12] Z. Wang, C. Pei, M. Ma et al., “Revisiting VAE for Unsupervised Time Series Anomaly Detection: A Frequency Perspective,” Proc. of the ACM on Web Conf. 2024, pp. 3096-3105, 2024. https://doi.org/10.1145/3589334.3645710
13. [13] K. Sohn, H. Lee, and X. Yan, “Learning structured output representation using deep conditional generative models,” Advances in Neural Information Processing Systems, Vol.28, 2015.
14. [14] V. Fortuin, D. Baranchuk, G. Rätsch et al., “Gp-vae: Deep probabilistic time series imputation,” Int. Conf. on Artificial Intelligence and Statistics, pp. 1651-1661, 2020.
15. [15] H. Hua, F. Qian, G. Zhang et al., “Unsupervised Seismic Facies Deep Clustering Via Lognormal Mixture-Based Variational Autoencoder,” IEEE J. of Selected Topics in Applied Earth Observations and Remote Sensing, Vol.16, pp. 9831-9842, 2023. https://doi.org/10.1109/JSTARS.2023.3325969
16. [16] S. Yu and J. Ma, “Deep learning for geophysics: Current and future trends,” Reviews of Geophysics, Vol.59, No.3, Article No.e2021RG000742, 2021. https://doi.org/10.1029/2021RG000742
17. [17] A. He and X. Jin, “Deep variational autoencoder classifier for intelligent fault diagnosis adaptive to unseen fault categories,” IEEE Trans. on Reliability, Vol.70, No.4, pp. 1581-1595, 2021. https://doi.org/10.1109/TR.2021.3090310
18. [18] S. Hochreiter, “Long Short-term Memory,” Neural Computation MIT-Press, 1997. https://doi.org/10.1162/neco.1997.9.8.1735
19. [19] C. Chi, C. Li, Y. Han et al., “Pre-earthquake anomaly extraction from borehole strain data based on machine learning,” Scientific Reports, Vol.13, No.1, Article No.20095, 2023. https://doi.org/10.1038/s41598-023-47387-z
20. [20] N. S. M. Ridzwan and S. H. M. Yusoff, “Machine learning for earthquake prediction: A review (2017–2021),” Earth Science Informatics, Vol.16, No.2, pp. 1133-1149, 2023. https://doi.org/10.1007/s12145-023-00991-z
21. [21] Z. Li, W. Chen, and D. Pei, “Robust and unsupervised KPI anomaly detection based on conditional variational autoencoder,” 2018 IEEE 37th Int. Performance Computing and Communications Conf. (IPCCC), 2018. https://doi.org/10.1109/PCCC.2018.8710885
22. [22] H. Xu, W. Chen, N. Zhao et al., “Unsupervised anomaly detection via variational auto-encoder for seasonal KPIs in web applications,” Proc. of the 2018 World Wide Web Conf., pp. 187-196, 2018. https://doi.org/10.1145/3178876.3185996
23. [23] A. Le Guennec, S. Malinowski, and R. Tavenard, “Data augmentation for time series classification using convolutional neural networks,” ECML/PKDD Workshop on Advanced Analytics and Learning on Temporal Data, 2016.
24. [24] Q. Wen, L. Sun, F. Yang et al., “Time series data augmentation for deep learning: A survey,” arXiv:2002.12478, 2020. https://doi.org/10.24963/ijcai.2021/631
25. [25] Z. Li, Y. Zhao, J. Han et al., “Multivariate time series anomaly detection and interpretation using hierarchical inter-metric and temporal embedding,” Proc. of the 27th ACM SIGKDD Conf. on Knowledge Discovery & Data Mining, pp. 3220-3230, 2021. https://doi.org/10.1145/3447548.3467075
26. [26] C. Zhang, T. Zhou, Q. Wen, and L. Sun, “TFAD: A decomposition time series anomaly detection architecture with time-frequency analysis,” Proc. of the 31st ACM Int. Conf. on Information & Knowledge Management, pp. 2497-2507, 2022. https://doi.org/10.1145/3511808.3557470
27. [27] S. Zhang, H. Tong, J. Xu, and R. Maciejewski, “Graph convolutional networks: a comprehensive review,” Computational Social Networks, Vol.6, No.1, pp. 1-23, 2019. https://doi.org/10.1186/s40649-019-0069-y
28. [28] A. Siffer, P.-A. Fouque, A. Termier, and C. Largouet, “Anomaly detection in streams with extreme value theory,” Proc. of the 23rd ACM SIGKDD Int. Conf. on Knowledge Discovery and Data Mining, pp. 1067-1075, 2017. https://doi.org/10.1145/3097983.3098144
29. [29] H. Ren, B. Xu, Y. Wang et al., “Time-series anomaly detection service at Microsoft,” Proc. of the 25th ACM SIGKDD Int. Conf. on Knowledge Discovery & Data Mining, pp. 3009-3017, 2019. https://doi.org/10.1145/3292500.3330680
30. [30] C. Zhang, T. Zhou, Q. Wen, and L. Sun, “TFAD: A decomposition time series anomaly detection architecture with time-frequency analysis,” Proc. of the 31st ACM Int. Conf. on Information & Knowledge Management, pp. 2497-2507, 2022. https://doi.org/10.1145/3511808.3557470
31. [31] H. Xu, W. Chen, N. Zhao et al., “Unsupervised anomaly detection via variational auto-encoder for seasonal KPIs in web applications,” Proc. of the 2018 World Wide Web Conf., pp. 187-196, 2018. https://doi.org/10.1145/3178876.3185996
32. [32] H. Zhou, S. Zhang, J. Peng et al., “Informer: Beyond efficient transformer for long sequence time-series forecasting,” Proc. of the AAAI Conf. on Artificial Intelligence, Vol.35, No.12, pp. 11106-11115, 2021. https://doi.org/10.1609/aaai.v35i12.17325
33. [33] J. Xu, H. Wu, J. Wang, and M. Long, “Anomaly Transformer: Time series anomaly detection with association discrepancy,” Int. Conf. on Learning Representations, 2022.
34. [34] S. Zhang, Z. Zhong, D. Li et al., “Efficient KPI anomaly detection through transfer learning for large-scale web services,” IEEE J. on Selected Areas in Communications, Vol.40, No.8, pp. 2440-2455, 2022. https://doi.org/10.1109/JSAC.2022.3180785
35. [35] T. Kieu, B. Yang, C. Guo et al., “Anomaly detection in time series with robust variational quasi-recurrent autoencoders,” 2022 IEEE 38th Int. Conf. on Data Engineering (ICDE), pp. 1342-1354, 2022. https://doi.org/10.1109/ICDE53745.2022.00105