Earthquake prediction is a difficult problem in contemporary science, and applications of machine learning methods in the prediction have drawn intensive attention. Large earthquakes can cause huge casualties and economic losses, and are the main goals of earthquake prediction. We studies intermediate-term (one-year) earthquake prediction in Sichuan and Yunnan provinces using the earthquake catalogue since 1970 by the sliding time-space window technique and LSTM(long short-term memory) neural networks. Sixteen earthquake prediction indexes that reflect the temporal and spatial features of earthquake sequences were used in the neural network. The neural network was trained using data sets from 1970 to 2004 (70% of all earthquake catalogues). Retrospective prediction tests were conducted on earthquakes after 2005, the accuracy rate (actual magnitude fell within ±0.5 of the predicted magnitude) was 70.2%, over-prediction rate was 18.7%, and under-prediction rate was 11.1%. The 2008 Wenchuan MS8.0 earthquake was retrospectively predicted. In order to understand the robustness of the model, we have done some tests, such as to expand the study area, change the weights of large earthquakes in calculation of the mean square error, etc. The LSTM neural network model still performed well in the tests.
[1] Gilbert G K. Earthquake forecasts introduction[J]. Science, 1909, 29(734):121-138.
[2] 傅承义. 有关地震預告的几个問題[J]. 科学通报, 1963, 8(3):30-36.
[3] Walsh J. Earthquake prediction:OST panel recommends 10-year program[J]. Science, 1965, 150(3694):321-323.
[4] Hagiwara T, Rikitake T. Japanese program on earthquake prediction:a prediction program now under way in Japan succeeds in long-range forecast of the Matsushiro earthquake[J/OL]. Science, 1967, 157(3790):761-768[2021-03-20]. https://science.sciencemag.org/content/157/3790/761.
[5] Savarensky E F. On the prediction of earthquakes[J]. Tectonophysics, 1968, 6(1):17-27.
[6] American Seismology Delegation. Earthquake research in China[EB/OL]. EOS Trans AGU 56, 1975:838-881. (2011-06-03)[2021-03-20]. https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/EO056i011p00838.
[7] Kossobokov V G. Earthquake prediction:20 years of global experiment[J]. Natural Hazards, 2013, 69(2):1155-1177.
[8] Wang K. Predicting the 1975 Haicheng earthquake[J]. Bulletin of the Seismological Society of America, 2006, 96(3):757-795.
[9] Bakun W H, Lindh A G. The Parkfield, California, earthquake prediction experiment[J]. Science, 1985, 229(4714):619-624.
[10] Fitzpartrick C. The first ‘A’ alert of the Parkfield earthquake prediction experiment:a description of organizational response[J/OL]. International Journal of Mass Emergencies and Disasters, 1994, 12(2):183-197[2021-03-20]. http://ijmed.org/articles/422/https://science.sciencemag.org/content/157/3790/761.
[11] Geller R J, Jackson D D, Kagan Y Y, et al. Earthquakes cannot be predicted[J/OL]. Science, 1997, 275(5306):1616-1618[2021-03-20]. https://science.sciencemag.org/content/275/5306/1616.full.
[12] Vere-Jones D. On the variance properties of stress release models[J/OL]. Australian Journal of Statistics, 1988, 30A(1):123-135[2021-03-20]. https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1467-842X.1988.tb00469.x.
[13] Ogata Y. Space-time point-process models for earthquake occurrences[J]. Annals of the Institute of Statistical Mathematics, 1998, 50(2):379-402.
[14] Gelfand I M, Guberman S A, Keilis-Borok V I, et al. Pattern recognition applied to earthquake epicenters in California[J]. Physics of the Earth and Planetary Interiors, 1976, 11(3):227-283.
[15] Keilis-Borok V I, Kossobokov V G. Premonitory activation of earthquake flow:algorithm M8[J]. Physics of the Earth and Planetary Interiors, 1990, 61(1/2):73-83.
[16] Lichtman A J, Keilis-Borok V I. Pattern recognition applied to presidential elections in the United States, 1860-1980:role of integral social, economic, and political traits[J]. PNAS, 1981, 78(11):7230-7234.
[17] Panakkat A, Adeli H. Neural network models for earthquake magnitude prediction using multiple seismicity indicators[J]. International Journal of Neural Systems, 2007, 17(1):13-33.
[18] Adeli H, Panakkat A. A probabilistic neural network for earthquake magnitude prediction[J]. Neural Networks, 2009, 22(7):1018-1024.
[19] 聂红林, 袁孝, 胡伍生, 等. 基于BP神经网络技术的区域短期地震预测模型研究[J]. 现代测绘, 2012, 35(2):3-5, 9.
[20] Wang Q L, Guo Y F, Yu L X, et al. Earthquake prediction based on spatio-temporal data mining:an LSTM network approach[J]. IEEE Transactions on Emerging Topics in Computing, 2020, 8(1):148-158.
[21] Asim K M, Idris A, Iqbal T, et al. Earthquake prediction model using support vector regressor and hybrid neural networks[J]. PLoS One, 2018, 13(7):e0199004.
[22] Huang J P, Wang X A, Zhao Y, et al. Large earthquake magnitude prediction in Taiwan based on deep learning neural network[J]. Neural Network World, 2018, 28(2):149-160.
[23] Li R, Lu X B, Li S W, et al. DLEP:a deep learning model for earthquake prediction[C]//2020 International Joint Conference on Neural Networks (IJCNN). July 19-24, 2020, Glasgow, UK. IEEE, 2020:1-8.
[24] 马宗晋. 关于1979年巴黎国际地震预报讨论会情况介绍及渐进式地震预报的三个理论问题:"多阶段"、"多点场"、"多类型"的讨论(摘要)[J]. 国际地震动态, 1980, 10(3):1-3.
[25] 张国民, 李宣瑚. 我国的地震预报研究亟待加强[J]. 中国科技论坛, 1998(3):31-33.
[26] Li Y J, Liu S F, Chen L W, et al. Mechanism of crustal deformation in the Sichuan-Yunnan region, southeastern Tibetan Plateau:insights from numerical modeling[J]. Journal of Asian Earth Sciences, 2017, 146:142-151.
[27] 张国民, 傅征祥, 桂燮泰, 等. 地震预报引论[M]. 北京:科学出版社, 2001.
[28] 陈运泰. 地震预测:回顾与展望[J]. 中国科学(D辑:地球科学), 2009, 39(12):1633-1658.
[29] Shi Y L, Liu J, Zhang G M. An evaluation of Chinese annual earthquake predictions, 1990-1998[J]. Journal of Applied Probability, 2001, 38(A):222-231.
[30] Field E H. Overview of the working group for the development of regional earthquake likelihood models (RELM)[J]. Seismological Research Letters, 2007, 78(1):7-16.
[31] Eyben F, Weninger F, Squartini S, et al. Real-life voice activity detection with LSTM recurrent neural networks and an application to Hollywood movies[C]//2013 IEEE International Conference on Acoustics, Speech and Signal Processing. May 26-31, 2013, Vancouver, BC, Canada. IEEE, 2013:483-487.
[32] 吴忠良. 关于b值应用于地震趋势预测的讨论[J]. 地震学报, 2001, 23(5):548-551.
[33] 庄建仓, 刘杰, 薛艳, 等. 论统计相关性和因果性:以中国油价上涨与全球大地震关系为例[J]. 地震, 2017, 37(1):1-9.
[34] Vere-Jones D, Robinson R, Yang W Z. Remarks on the accelerated moment release model:problems of model formulation, simulation and estimation[J]. Geophysical Journal International, 2001, 144(3):517-531.
