Microseismic event picking and classification for hot dry rock hydraulic fracturing monitoring using SeisFormer

Abstract

Accurate seismic monitoring is vital for the safe operation of enhanced geothermal systems in hot dry rock (HDR) reservoirs; however, robust P- and S-wave classification and precise first-arrival picking remain difficult under low signal-to-noise ratios and complex noise conditions. Hence, in this study, we present SeisFormer, a Transformer-based framework that couples adaptive multi-scale windowing with joint time-frequency analysis. It allocates time-frequency resolution on the fly to overcome the limitations of a fixed-window short-time Fourier transform and slowly extracts varying trends and dominant periodicities from waveform sequences. To stabilize the modeling of long-range dependencies, we introduce regularized pseudoinverse attention, which retains the speedups of low-rank approximations while damping amplification in directions associated with small singular values. We evaluated SeisFormer on a unified, multi-site dataset with data from HDR operations in the Qinghai Gonghe Basin and from an unconventional hydraulic-fracturing field in North China. Compared with baselines (EQTransformer, PhaseNet), it exhibited better performance across real-world data, noise-augmented data with non-stationary composite noise, and overlapping multi-event scenarios. On real-world data, it attained 98.30% classification accuracy, with mean arrival-time errors of 1.42 ms (P) and 2.29 ms (S). Ablations show that each component contributes substantially, indicating robustness for near-real-time monitoring and deployment.

Keywords

Microseismic monitoring / Hot dry rock hydraulic fracturing / Picking and classification / Transformer / Adaptive multi-scale windowing / Time-frequency domain

Funding

This work was supported by the National Key R&D Program of China (Grant Nos. 2018YFB1501803 and 2019YFC1804805-4) and the China Geological Survey Project (Grant No. DD2019135).

References

[1] Dong S, Jiao J, Zhou S, Lu P, Zeng Z. 3-D gravity data inversion based on enhanced dual U-Net framework. IEEE Trans Geosci Remote Sens. 2023; 61:1-11. doi:10.1109/TGRS.2023.3306980

[2] Cloetingh S, Sternai P, Koptev A, et al. Coupled surface to deep Earth processes: Perspectives from TOPO-EUROPE with an emphasis on climate- and energy-related societal challenges. Global Planet Change. 2023; 226:104140. doi:10.1016/j.gloplacha.2023.104140

[3] Ma W, Wang Y, Wu X, Liu G. Hot dry rock (HDR) hydraulic fracturing propagation and impact factors assessment via sensitivity indicator. Renew Energy. 2020; 146:2716-2723. doi:10.1016/j.renene.2019.08.097

[4] Xie J, Cao H, Wang D, Peng S, Fu G, Zhu Z. A comparative study on the hydraulic fracture propagation behaviors in hot dry rock and shale formation with different structural discontinuities. J Energy Eng. 2022; 148(6):04022040. doi:10.1061/(asce)ey.1943-7897.000.8.6

[5] Cheng Y, Zhang Y, Yu Z, Hu Z, Yang Y. An investigation on hydraulic fracturing characteristics in granite geothermal reservoir. Eng Fract Mech. 2020; 237:107252. doi:10.1016/j.engfracmech.2020.107252

[6] Cheng Y, Zhang Y, Yu Z, Hu Z. Investigation on reservoir stimulation characteristics in hot dry rock geothermal formations of China during hydraulic fracturing. Rock Mech Rock Eng. 2021; 54(8):3817-3845. doi:10.1007/s00603-021-02506-y

[7] Cheng Y, Zhang Y. Experimental study of fracture propagation: The application in energy mining. Energies. 2020; 13(6):1411. doi:10.3390/en13061411

[8] Wang H, Alkhalifah T, Waheed UB, Birnie C. Data-driven microseismic event localization: An application to the Oklahoma Arkoma Basin hydraulic fracturing data. IEEE Trans Geosci Remote Sens. 2022; 60:1-12. doi:10.1109/TGRS.2021.3120546

[9] Warpinski N. Microseismic monitoring: Inside and out. J Petrol Technol. 2009; 61(11):80-85. doi:10.2118/118537-JPT

[10] Allen RV. Automatic earthquake recognition and timing from single traces. Bull Seismol Soc Am. 1978; 68(5):1521-1532. doi:10.1785/bssa0680051521

[11] Akaike H. A new look at the statistical model identification. IEEE Trans Autom Control. 1974; 19(6):716-723. doi:10.1109/tac.1974.1100705

[12] Allen R. Automatic phase pickers: Their present use and future prospects. Bull Seismol Soc Am. 1982; 72(6B):S225-S242. doi:10.1785/BSSA07206B0225

[13] Álvarez I, Garcia L, Mota S, et al. An automatic P-phase picking algorithm based on adaptive multiband processing. IEEE Geosci Remote Sens Lett. 2013; 10(6):1488-1492. doi:10.1109/lgrs.2013.2260720

[14] Earle PS, Shearer PM. Characterization of global seismograms using an automatic-picking algorithm. Bull Seismol Soc Am. 1994; 84(2):366-376. doi:10.1785/bssa0840020366

[15] Akazawa TA. Technique for Automatic Detection of Onset Time of P- and S-Phases in Strong Motion Records. In: Proceedings 13th Conference on Earthquake Engineering. Vancouver, Canada. Paper No. 786; 2004.

[16] Kurz JH, Grosse CU, Reinhardt HW. Strategies for reliable automatic onset time picking of acoustic emissions and of ultrasound signals in concrete. Ultrasonics. 2005; 43(7):538-546. doi:10.1016/j.ultras.2004.12.005

[17] Anikiev D, Birnie C, Waheed UB, et al. Machine learning in microseismic monitoring. Earth Sci Rev. 2023; 239:104371. doi:10.1016/j.earscirev.2023.104371

[18] LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015; 521:436-444. doi:10.1038/nature14539

[19] Mousavi SM, Beroza GC. Deep-learning seismology. Science. 2022; 377:725-729. doi:10.1126/science.abm4470

[20] Zhu W, Beroza GC. PhaseNet: A deep-neural-network-based seismic arrival-time picking method. Geophys J Int. 2019; 216(1):261-273. doi:10.1093/gji/ggy423

[21] He Z, Peng P, Wang L, Jiang Y. PickCapsNet: Capsule network for automatic P-wave arrival picking. IEEE Geosci Remote Sens Lett. 2021; 18(4):617-621. doi:10.1109/lgrs.2020.2983196

[22] Zhao Y, Xu H, Yang T, Wang D, Sun D. A hybrid recognition model of microseismic signals for underground mining based on CNN and LSTM networks. Geomat Nat Hazard Risk. 2021; 12(1):2803-2834. doi:10.1080/19475705.2021.1968043

[23] Mousavi SM, Ellsworth WL, Zhu W, Chuang LY, Beroza GC. Earthquake transformer: An attentive deep-learning model for simultaneous earthquake detection and phase picking. Nat Commun. 2020; 11:3952. doi:10.1038/s41467-020-17591-w

[24] Saad OM, Chen Y, Siervo D, et al. EQCCT: A production-ready earthquake detection and phase-picking method using the compact convolutional transformer. IEEE Trans Geosci Remote Sens. 2023; 61:1-15. doi:10.1109/TGRS.2023.3319440

[25] Li S, Yang X, Cao A, et al. SeisT: A foundational deep-learning model for earthquake monitoring tasks. IEEE Trans Geosci Remote Sens. 2024; 62:1-15. doi:10.1109/tgrs.2024.3371503

[26] Li XN, Chen FJ, Lai YP, Tang P, Liang X. ICAT-net: A lightweight neural network with optimized coordinate attention and transformer mechanisms for earthquake detection and phase picking. J Supercomput. 2025; 81:191. doi:10.1007/s11227-024-06664-y

[27] Peng P, Lei R, Wang JM. Enhancing microseismic signal classification in metal mines using transformer-based deep learning. Sustainability. 2023; 15(20):14959.doi:10.3390/su152014959

[28] Zhang X, Wang X, Zhang Z, Wang Z. CNN-transformer for microseismic signal classification. Electronics. 2023; 12(11):2468. doi:10.3390/electronics12112468

[29] Zhu L, Chuang L, McClellan JH, Liu E, Peng Z. A Multi- Channel Approach for Automatic Microseismic Event Association using RANSAC-Based Arrival Time Event Clustering (RATEC). [arXiv Preprint]; 2017. doi:10.48550/arXiv.1702.01856

[30] Li Z, Gou X, Jin W, Qin N. Frequency features of microseismic signals. Chin J Geotech Eng. 2008; 30(6):830-834.

[31] Aki K, Richards PG. Quantitative Seismology. 2nd ed. Sausalito, CA: University Science Books; 2002.

[32] Ma C, Ran X, Xu W, et al. Fine classification method for massive microseismic signals based on short-time Fourier transform and deep learning. Remote Sens. 2023; 15(2):502. doi:10.3390/rs15020502

[33] Xiong Y, Zeng Z, Chakraborty R, et al. Nyströmformer: A Nyström-based algorithm for approximating self-attention. Proc AAAI Conf Artif Intell. 2021; 35(16):14138-14148. doi:10.48550/arXiv.2102.03902

[34] Piao X, Chen Z, Murayama T, et al. Fredformer: Frequency Debiased Transformer for Time Series Forecasting. In: Proceedings 30th ACM SIGKDD Conference Knowledge Discovery and Data Mining (KDD 2024). Long Beach, CA, USA; 2024. p. 1234-1245. doi:10.1145/3637528.3671928

[35] Soto-Quiros P. A fast method to estimate the Moore- Penrose inverse for well-determined numerical rank matrices based on Tikhonov regularization. J Math Comput Sci. 2024; 37(1):59-81. doi:10.22436/jmcs.037.01.05

[36] Longhas PRA, Abdul AM. Application of Tikhonov regularization in generalized inverse of adjacency matrix of undirected graph. Int J Math Trends Technol. 2022; 68(2):1-6. doi:10.14445/22315373/ijmtt-v68i2p501

[37] Gulliksson ME, Wedin PÅ, Wei Y. Perturbation identities for regularized Tikhonov inverses and weighted pseudoinverses. BIT Numer Math. 2000; 40:513-523. doi:10.1023/A:1022319830134

[38] Ross ZE, Meier MA, Hauksson E, Heaton TH. Generalized seismic phase detection with deep learning. Bull Seismol Soc Am. 2018; 108(5A):2894-2901. doi:10.1785/0120180080