基于机器学习的实验室断层摩擦状态预测

地震灾害给人类造成了巨大的生命和财产损失，因此可靠的地震预测理论与方法对于防震减灾具有重要意义。然而，天然地震的监测难度大、不可控因素多，所以研究实验室尺度的地震成为地震预测研究的重要手段。机器学习算法具有强大的数据挖掘和表征能力，建立实验室地震机器学习预测模型，有助于理解实验室地震的孕震机理，为探究实验室地震和天然地震的预测方法提供经验。

相比于室内物理实验，基于数值模拟的实验室地震可以布置大量监测点，获取丰富的断层摩擦响应信息。然而，众多监测点所获数据可能存在冗余信息，从而影响机器学习模型的性能。因此，如何筛选出关键监测点对于建立高效的实验室地震预测方法具有重要意义。本文基于数值模拟的实验室地震数据集，建立了断层宏观摩擦状态的机器学习预测模型，采用可解释方法分析了剪切板与断层泥颗粒处监测点对模型预测性能的贡献差异，结果表明，关键监测点主要位于剪切板中央区域，且剪切板上关键监测点数据反映的是断层全局摩擦响应信息，而断层泥颗粒的运动反映的更多是局部摩擦响应信息。本文还通过计算非仿射位移探究了微观摩擦响应与宏观摩擦状态的时空关联，发现贡献大的微观摩擦响应特征往往位于颗粒重排较剧烈的区域。

在实验室地震中发现了类似于在天然地震中观察到的缓慢的震前滑动阶段， 这一阶段很可能指示了地震成核过程，探索地震成核过程有助于理解地震孕震机理。本文分别基于滑移早期阶段的剪切板运动和系统动能，建立了滑移持续时间和摩擦降的预测模型。结果表明滑移早期剪切板在剪
切方向上位移的四阶矩对断层滑移持续时间的预测贡献最大，且四阶矩越高，持续时间越长。而完整滑移阶段剪切方向上速度信号的二阶矩是摩擦降预测的最佳特征。此外， 系统动能也包含有关于滑移持续时间和摩擦降的前兆信息，且对系统动能进行频率分解可以挖掘出更多的前兆信息。
 

[1] 陈运泰. 地震预测: 回顾与展望[J]. 中国科学: D 辑, 2009, (12): 1633-58.
[2] BRACE W, BYERLEE J. Stick-slip as a mechanism for earthquakes[J]. Science, 1966, 153(3739): 990-2.
[3] JOHNSON P A, FERDOWSI B, KAPROTH B M, et al. Acoustic emission and microslip precursors to stick-slip failure in sheared granular material[J]. Geophysical Research Letters, 2013, 40(21): 5627-31.
[4] RIVIERE J, LV Z, JOHNSON P A, et al. Evolution of b-value during the seismic cycle: Insights from laboratory experiments on simulated faults[J]. Earth and Planetary Science Letters, 2018, 482: 407-13.
[5] BERGEN K J, JOHNSON P A, DE HOOP M V, et al. Machine learning for data-driven discovery in solid Earth geoscience[J]. Science, 2019, 363(6433).
[6] 吕征. 含颗粒物模拟断层粘滑运动机制的实验研究[D]; 清华大学, 2019.
[7] ZHUO Y-Q, LIU P, CHEN S, et al. Laboratory Observations of Tremor-Like Events Generated During Preslip[J]. Geophys Res Lett, 2018, 45(14): 6926-34.
[8] DAHMEN K A, BEN-ZION Y, UHL J T. A simple analytic theory for the statistics of avalanches in sheared granular materials[J]. Nature Physics, 2011, 7(7): 554-7.
[9] RIVIèRE J, LV Z, JOHNSON P A, et al. Evolution of b-value during the seismic cycle: Insights from laboratory experiments on simulated faults[J]. Earth and Planetary Science Letters, 2018, 482: 407-13.
[10] DOROSTKAR O, GUYER R A, JOHNSON P A, et al. On the micromechanics of slip events in sheared, fluid-saturated fault gouge[J]. 2017, 44(12): 6101-8.
[11] BRZINSKI T A, DANIELS K E. Sounds of Failure: Passive Acoustic Measurements of Excited Vibrational Modes[J]. Physical Review Letters, 2018, 120(21): 218003.
[12] GAO K, EUSER B J, ROUGIER E, et al. Modeling of Stick-Slip Behavior in Sheared Granular Fault Gouge Using the Combined Finite-Discrete Element Method[J]. J Geophys Res Solid Earth, 2018, 123: 5774– 92.
[13] ROUET-LEDUC B, HULBERT C, LUBBERS N, et al. Machine Learning Predicts Laboratory Earthquakes[J]. Geophys Res Lett, 2017, 44(18): 9276-82.
[14] JASPERSON H, BOLTON D C, JOHNSON P, et al. Attention Network Forecasts Time-to-Failure in Laboratory Shear Experiments[J]. J Geophys Res Solid Earth, 2021, 126(11): e2021JB022195.
[15] MCBECK J A, AIKEN J M, MATHIESEN J, et al. Deformation Precursors to Catastrophic Failure in Rocks[J]. Geophys Res Lett, 2020, 47(24): e2020GL090255.
[16] SHREEDHARAN S, BOLTON D C, RIVIèRE J, et al. Machine Learning Predicts the Timing and Shear Stress Evolution of Lab Earthquakes Using Active Seismic Monitoring of Fault Zone Processes[J]. J Geophys Res Solid Earth, 2021, 126(7): e2020JB021588.
[17] CHAIPORNKAEW L, ELSTON H, COOKE M, et al. Predicting Off-Fault Deformation From Experimental Strike-Slip Fault Images Using Convolutional Neural Networks[J]. Geophys Res Lett, 2022, 49(2): e2021GL096854.
[18] JOHNSON P A, ROUET-LEDUC B, PYRAK-NOLTE L J, et al. Laboratory earthquake forecasting: A machine learning competition[J]. Proceedings of the National Academy of Sciences, 2021, 118(5): e2011362118.
[19] MIGNAN A, BROCCARDO M. Neural Network Applications in Earthquake Prediction (1994-2019): Meta-Analytic and Statistical Insights on Their Limitations[J]. Seismological Research Letters, 2020, 91(4): 2330-42.
[20] BEROZA G C, SEGOU M, MOSTAFA MOUSAVI S. Machine learning and earthquake forecasting—next steps[J]. Nature Communications, 2021, 12(1): 4761.
[21] VAN KLAVEREN S, VASCONCELOS I, NIEMEIJER A. Predicting laboratory earthquakes with machine learning[J]. arXiv preprint arXiv:201106669, 2020.
[22] BERGEN K J, JOHNSON P A, DE HOOP M V, et al. Machine learning for data-driven discovery in solid Earth geoscience[J]. Science, 2019, 363(6433): eaau0323.
[23] ROSS Z E, MEIER M A, HAUKSSON E. P Wave Arrival Picking and First-Motion Polarity Determination With Deep Learning[J]. J Geophys Res-Sol Ea, 2018, 123(6): 5120-9.
[24] LI Z, MEIER M-A, HAUKSSON E, et al. Machine Learning Seismic Wave Discrimination: Application to Earthquake Early Warning[J]. Geophysical Research Letters, 2018, 45(10): 4773-9.
[25] BIANCO M J, GERSTOFT P, OLSEN K B, et al. High-resolution seismic tomography of Long Beach, CA using machine learning[J]. Sci Rep, 2019, 9(1): 14987.
[26] BAI T, TAHMASEBI P. Efficient and data-driven prediction of water breakthrough in subsurface systems using deep long short-term memory machine learning[J]. Computational Geosciences, 2020, 25(1): 285-97.
[27] DEVRIES P M R, VIEGAS F, WATTENBERG M, et al. Deep learning of aftershock patterns following large earthquakes[J]. Nature, 2018, 560(7720): 632-+.
[28] BREGMAN Y, RABIN N. Aftershock Identification Using Diffusion Maps[J]. Seismological Research Letters, 2018, 90(2A): 539-45.
[29] HAYES G P, EARLE P S, BENZ H M, et al. National earthquake information center strategic plan, 2019–23[R]: US Geological Survey, 2019.
[30] PANAKKAT A, ADELI H. Recurrent neural network for approximate earthquake time and location prediction using multiple seismicity indicators[J]. Computer‐Aided Civil and Infrastructure Engineering, 2009, 24(4): 280-92.
[31] ALLEN R M, MELGAR D. Earthquake Early Warning: Advances, Scientific Challenges, and Societal Needs[J]. Annual Review of Earth and Planetary Sciences, 2019, 47(1): 361-88.
[32] BRYKOV M N, PETRYSHYNETS I, PRUNCU C I, et al. Machine Learning Modelling and Feature Engineering in Seismology Experiment[J]. Sensors (Basel), 2020, 20(15).
[33] ASIM K M, MARTíNEZ-ÁLVAREZ F, BASIT A, et al. Earthquake magnitude prediction in Hindukush region using machine learning techniques[J]. Natural Hazards, 2017, 85: 471-86.
[34] ROUET-LEDUC B, HULBERT C, JOHNSON P A. Continuous chatter of the Cascadia subduction zone revealed by machine learning[J]. Nature Geoscience, 2019, 12(1): 75-9.
[35] MAIR K, FRYE K M, MARONE C. Influence of grain characteristics on the friction of granular shear zones[J]. Journal of Geophysical Research: Solid Earth, 2002, 107(B10): ECV 4-1-ECV 4-9.
[36] IKARI M J, MARONE C, SAFFER D M. On the relation between fault strength and frictional stability[J]. Geology, 2011, 39(1): 83-6.
[37] XU S, FUKUYAMA E, YAMASHITA F, et al. Strain rate effect on fault slip and rupture evolution: Insight from meter-scale rock friction experiments[J]. Tectonophysics, 2018, 733: 209-31.
[38] RUBINO V, LAPUSTA N, ROSAKIS A J. Intermittent lab earthquakes in dynamically weakening fault gouge[J]. Nature, 2022, 606(7916): 922-9.
[39] XU S, FUKUYAMA E, YAMASHITA F. Robust Estimation of Rupture Properties at Propagating Front of Laboratory Earthquakes[J]. Journal of Geophysical Research: Solid Earth, 2019, 124(1): 766-87.
[40] AHARONOV E, SPARKS D. Stick-slip motion in simulated granular layers[J]. Journal of Geophysical Research: Solid Earth, 2004, 109(B9).
[41] FERDOWSI B, GRIFFA M, GUYER R A, et al. Three-dimensional discrete element modeling of triggered slip in sheared granular media[J]. Phys Rev E Stat Nonlin Soft Matter Phys, 2014, 89(4): 042204.
[42] MA G, ZOU Y, GAO K, et al. Size polydispersity tunes slip avalanches of granular gouge[J]. Geophysical Research Letters, 2020, 47(23): e2020GL090458.
[43] GAO K, GUYER R A, ROUGIER E, et al. Plate motion in sheared granular fault system[J]. Earth Planet Sc Lett, 2020, 548: 116481.
[44] GAO K, GUYER R, ROUGIER E, et al. From Stress Chains to Acoustic Emission[J]. Phys Rev Lett, 2019, 123(4): 048003.
[45] ZHANG Y, GAO K, LI C. Two slip regimes in sheared granular fault[J]. Earth and Planetary Science Letters, 2023, 608: 118086.
[46] HULBERT C, ROUET-LEDUC B, JOHNSON P A, et al. Similarity of fast and slow earthquakes illuminated by machine learning[J]. Nature Geoscience, 2018, 12(1): 69-74.
[47] CORBI F, SANDRI L, BEDFORD J, et al. Machine learning can predict the timing and size of analog earthquakes[J]. Geophysical Research Letters, 2019, 46(3): 1303-11.
[48] LUBBERS N, BOLTON D C, MOHD‐YUSOF J, et al. Earthquake Catalog‐Based Machine Learning Identification of Laboratory Fault States and the Effects of Magnitude of Completeness[J]. Geophysical Research Letters, 2018, 45(24).
[49] SHREEDHARAN S, BOLTON D C, RIVIèRE J, et al. Machine Learning Predicts the Timing and Shear Stress Evolution of Lab Earthquakes Using Active Seismic Monitoring of Fault Zone Processes[J]. Journal of Geophysical Research: Solid Earth, 2021, 126(7).
[50] BOLTON D C, SHOKOUHI P, ROUET‐LEDUC B, et al. Characterizing Acoustic Signals and Searching for Precursors during the Laboratory Seismic Cycle Using Unsupervised Machine Learning[J]. Seismological Research Letters, 2019, 90(3): 1088-98.
[51] SHOKOUHI P, GIRKAR V, RIVIèRE J, et al. Deep learning can predict laboratory quakes from active source seismic data[J]. Geophysical Research Letters, 2021, 48(12): e2021GL093187.
[52] REN C X, DOROSTKAR O, ROUET‐LEDUC B, et al. Machine Learning Reveals the State of Intermittent Frictional Dynamics in a Sheared Granular Fault[J]. Geophysical Research Letters, 2019, 46(13): 7395-403.
[53] MA G, MEI J, GAO K, et al. Machine learning bridges microslips and slip avalanches of sheared granular gouges[J]. Earth and Planetary Science Letters, 2022, 579.
[54] MEI J, MA G, WANG Q, et al. Micro-and macroscopic aspects of the intermittent behaviors of granular materials related by graph neural network[J]. International Journal of Solids and Structures, 2022, 251: 111763.
[55] WANG K, JOHNSON C W, BENNETT K C, et al. Predicting fault slip via transfer learning[J]. Nat Commun, 2021, 12(1): 7319.
[56] MUNJIZA A, SMOLJANOVIĆ H, ŽIVALJIĆ N, et al. Structural applications of the combined finite–discrete element method[J]. Computational particle mechanics, 2020, 7: 1029-46.
[57] YAN C, ZHENG H, SUN G, et al. Combined finite-discrete element method for simulation of hydraulic fracturing[J]. Rock mechanics and rock engineering, 2016, 49: 1389-410.
[58] KNIGHT E E, ROUGIER E, LEI Z, et al. HOSS: an implementation of the combined finite-discrete element method[J]. Computational Particle Mechanics, 2020, 7: 765-87.
[59] GAO K, EUSER B J, ROUGIER E, et al. Modeling of Stick-Slip Behavior in Sheared Granular Fault Gouge Using the Combined Finite-Discrete Element Method[J]. Journal of Geophysical Research: Solid Earth, 2018, 123(7): 5774-92.
[60] CHEN T, HE T, BENESTY M, et al. Xgboost: extreme gradient boosting[J]. R package version 04-2, 2015, 1(4): 1-4.
[61] KE G, MENG Q, FINLEY T, et al. LightGBM: a highly efficient gradient boosting decision tree[Z]. Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach, California, USA; Curran Associates Inc. 2017: 3149–57
[62] MUNJIZA A. Discrete elements in transient dynamics of fractured media[D]; Swansea University, 1992.
[63] GELLER D A, ECKE R E, DAHMEN K A, et al. Stick-slip behavior in a continuum-granular experiment[J]. Phys Rev E Stat Nonlin Soft Matter Phys, 2015, 92(6): 060201.
[64] TSAI J C, VOTH G A, GOLLUB J P. Internal granular dynamics, shear-induced crystallization, and compaction steps[J]. Physical Review Letters, 2003, 91(6): 064301.
[65] WANG K, JOHNSON C W, BENNETT K C, et al. Predicting Future Laboratory Fault Friction Through Deep Learning Transformer Models[J]. Geophysical Research Letters, 2022, 49(19).
[66] LAURENTI L, TINTI E, GALASSO F, et al. Deep learning for laboratory earthquake prediction and autoregressive forecasting of fault zone stress[J]. Earth and Planetary Science Letters, 2022, 598.
[67] JOHNSON P A, ROUET-LEDUC B, PYRAK-NOLTE L J, et al. Laboratory earthquake forecasting: A machine learning competition[J]. Proc Natl Acad Sci U S A, 2021, 118(5).
[68] JASPERSON H, BOLTON D C, JOHNSON P, et al. Attention Network Forecasts Time‐to‐Failure in Laboratory Shear Experiments[J]. Journal of Geophysical Research: Solid Earth, 2021, 126(11).
[69] TRUGMAN D T, MCBREARTY I W, BOLTON D C, et al. The Spatiotemporal Evolution of Granular Microslip Precursors to Laboratory Earthquakes[J]. Geophysical Research Letters, 2020, 47(16).
[70] SHREEDHARAN S, BOLTON D C, RIVIèRE J, et al. Preseismic Fault Creep and Elastic Wave Amplitude Precursors Scale With Lab Earthquake Magnitude for the Continuum of Tectonic Failure Modes[J]. Geophysical Research Letters, 2020, 47(8).
[71] BREIMAN L, FRIEDMAN J, OLSHEN R, et al. Cart[J]. Classification and regression trees, 1984.
[72] GéRON A. Hands-on machine learning with Scikit-Learn, Keras, and TensorFlow[M]. " O'Reilly Media, Inc.", 2022.
[73] BREIMAN L. Random forests[J]. Machine learning, 2001, 45: 5-32.
[74] FRIEDMAN J H. Greedy function approximation: a gradient boosting machine[J]. Annals of statistics, 2001: 1189-232.
[75] BREIMAN L. Bagging predictors[J]. Machine learning, 1996, 24: 123-40.
[76] SCHAPIRE R E. The boosting approach to machine learning: An overview[J]. Nonlinear estimation and classification, 2003: 149-71.
[77] WOLPERT D H. Stacked generalization[J]. Neural networks, 1992, 5(2): 241-59.
[78] USTUN T S, HUSSAIN S M S, ULUTAS A, et al. Machine learning-based intrusion detection for achieving cybersecurity in smart grids using IEC 61850 GOOSE messages[J]. Symmetry, 2021, 13(5): 826.
[79] LUNDBERG S M, ERION G G, LEE S-I. Consistent individualized feature attribution for tree ensembles[J]. arXiv preprint arXiv:180203888, 2018.
[80] RIBEIRO M T, SINGH S, GUESTRIN C. " Why should i trust you?" Explaining the predictions of any classifier, F 2016].
[81] FALK M L, LANGER J S. Dynamics of viscoplastic deformation in amorphous solids[J]. Phys Rev E, 1998, 57(6): 7192.
[82] MA G, ZOU Y, CHEN Y, et al. Spatial correlation and temporal evolution of plastic heterogeneity in sheared granular materials[J]. Powder technology, 2021, 378: 263-73.
[83] MOUSAVI S M, BEROZA G C. A Machine‐Learning Approach for Earthquake Magnitude Estimation[J]. Geophysical Research Letters, 2020, 47(1).
[84] MARTINEZ-CANTIN R. BayesOpt: a Bayesian optimization library for nonlinear optimization, experimental design and bandits[J]. J Mach Learn Res, 2014, 15(1): 3735-9.
[85] VICTORIA A H, MARAGATHAM G. Automatic tuning of hyperparameters using Bayesian optimization[J]. Evolving Systems, 2021, 12(1): 217-23.
[86] REN C X, PELTIER A, FERRAZZINI V, et al. Machine Learning Reveals the Seismic Signature of Eruptive Behavior at Piton de la Fournaise Volcano[J]. Geophys Res Lett, 2020, 47(3): e2019GL085523.
[87] KHOSRAVIKIA F, CLAYTON P. Machine learning in ground motion prediction[J]. Computers & Geosciences, 2021, 148.
[88] FERDOWSI B, GRIFFA M, GUYER R A, et al. Microslips as precursors of large slip events in the stick‐slip dynamics of sheared granular layers: A discrete element model analysis[J]. Geophysical Research Letters, 2013, 40(16): 4194-8.
[89] RUIZ S, ADEN-ANTONIOW F, BAEZ J C, et al. Nucleation Phase and Dynamic Inversion of theMw6.9 Valparaíso 2017 Earthquake in Central Chile[J]. Geophysical Research Letters, 2017, 44(20): 10,290-10,7.
[90] SOCQUET A, VALDES J P, JARA J, et al. An 8 month slow slip event triggers progressive nucleation of the 2014 Chile megathrust[J]. Geophysical Research Letters, 2017, 44(9): 4046-53.
[91] KATO A, OBARA K, IGARASHI T, et al. Propagation of slow slip leading up to the 2011 M w 9.0 Tohoku-Oki earthquake[J]. Science, 2012, 335(6069): 705-8.
[92] RUIZ S, METOIS M, FUENZALIDA A, et al. Intense foreshocks and a slow slip event preceded the 2014 Iquique M w 8.1 earthquake[J]. Science, 2014, 345(6201): 1165-9.
[93] GUERIN-MARTHE S, NIELSEN S, BIRD R, et al. Earthquake Nucleation Size: Evidence of Loading Rate Dependence in Laboratory Faults[J]. J Geophys Res Solid Earth, 2019, 124(1): 689-708.
[94] KANEKO Y, NIELSEN S B, CARPENTER B M. The onset of laboratory earthquakes explained by nucleating rupture on a rate-and-state fault[J]. 2016, 121(8): 6071-91.
[95] OHNAKA M. Nonuniformity of the constitutive law parameters for shear rupture and quasistatic nucleation to dynamic rupture: a physical model of earthquake generation processes[J]. Proceedings of the National Academy of Sciences, 1996, 93(9): 3795-802.
[96] DIETERICH J H. Earthquake nucleation on faults with rate-and state-dependent strength[J]. Tectonophysics, 1992, 211(1): 115-34.
[97] WECH A G, CREAGER K C, HOUSTON H, et al. An earthquake‐like magnitude‐frequency distribution of slow slip in northern Cascadia[J]. Geophysical Research Letters, 2010, 37(22).
[98] BOLTON D C, SHOKOUHI P, ROUET‐LEDUC B, et al. Characterizing acoustic signals and searching for precursors during the laboratory seismic cycle using unsupervised machine learning[J]. 2019, 90(3): 1088-98.
[99] KAO C S, CARVALHO F C S, LABUZ J F. Micromechanisms of fracture from acoustic emission[J]. International Journal of Rock Mechanics and Mining Sciences, 2011, 48(4): 666-73.
[100] RICE J R. Elastic wave emission from damage processes[J]. Journal of Nondestructive Evaluation, 1980, 1: 215-24.
[101] TRUGMAN D T, MCBREARTY I W, BOLTON D C, et al. The spatiotemporal evolution of granular microslip precursors to laboratory earthquakes[J]. Geophysical Research Letters, 2020, 47(16): e2020GL088404.
[102] WANG K, JOHNSON C W, BENNETT K C, et al. Predicting fault slip via transfer learning[J]. Nature Communications, 2021, 12(1): 7319.
[103] DOROSTKAR O, GUYER R A, JOHNSON P A, et al. On the role of fluids in stick‐slip dynamics of saturated granular fault gouge using a coupled computational fluid dynamics‐discrete element approach[J]. Journal of Geophysical Research: Solid Earth, 2017, 122(5): 3689-700.
[104] JOHNSON P A, FERDOWSI B, KAPROTH B M, et al. Acoustic emission and microslip precursors to stick‐slip failure in sheared granular material[J]. Geophysical Research Letters, 2013, 40(21): 5627-31.
[105] MEROZ Y, MEADE B J. Intermittent granular dynamics at a seismogenic plate boundary[J]. Physical Review Letters, 2017, 119(13): 138501.
[106] ANDERSON D L. New Theory of the Earth[M]. 2 ed. Cambridge: Cambridge University Press, 2007.
[107] HULBERT C, ROUET-LEDUC B, JOHNSON P A, et al. Similarity of fast and slow earthquakes illuminated by machine learning[J]. Nature Geoscience, 2019, 12(1): 69-74.
[108] OBARA K, KATO A. Connecting slow earthquakes to huge earthquakes[J]. Science, 2016, 353(6296): 253-7.
[109] LEI X, MA S. Laboratory acoustic emission study for earthquake generation process[J]. Earthquake Science, 2014, 27: 627-46.