Machine learning earthquake prediction method based on seismic activity images and its application in North China
-
摘要:
本文开展了基于地震活动图像预测华北地区地震的机器学习方法研究。根据华北地区M≥5.0中强震、强震及大震的地震平均时间间隔，采用1970年以来华北地区M≥5.0中强震、强震及大震震前大量不同时窗长的ML≥3.0地震活动图像作为输入数据集，提出了基于地震活动图像预测地震的机器学习方法，并进行了震例回溯。利用本文提出的方法选取“未拓展图像数据集”和“含拓展图像数据集”对华北地区发生中强地震进行预测对比，结果显示，数据集样本量的增加有利于提高地震预测模型的精度，其中“含拓展图像数据集”预测地震的准确率可达77%；对于华北地区无震区、少震区的M≥5.0地震，可采用非1年窗长的较大时间间隔（3年、7年以上）的ML≥3.0地震活动图像验证。
Abstract:This paper presents a study on machine learning methods for predicting earthquakes in North China based on seismic activity images. According to the average time interval between moderately strong earthquakes （magnitude 5.0 or above）, strong earthquakes, and major earthquakes in North China, a large number of ML≥3.0 seismic activity images at different time scales before the occurrence of such earthquakes since 1970 are used as input datasets. The machine learning method for predicting earthquakes based on seismic activity images is proposed and tested using case studies. Using the method proposed in this paper, we select “unexpanded image dataset” and “expanded image dataset” to predict moderate-strong earthquakes in North China, it is found that increasing the sample size of the input image dataset is beneficial to improve the accuracy of the earthquake prediction models. The accuracy rate of earthquake activity image prediction in North China using the expanded image dataset can reach 77%. For M≥5.0 earthquakes happened in aseismic zones or low-seismic regions in North China, seismic activity images with larger time intervals than one year （such as three years or more than seven years） with ML≥3.0 seismic activity images can be used to verify the target earthquakes in aseismic zones or low-seismic regions. This research can provide important guidance for earthquake prediction in North China.
-
Keywords:
- North China /
- seismic activity images /
- machine learning /
- earthquake prediction
-
-
![]()
图 1 华北地区活动地块及其边界带$ ［ $修改自张培震等（2003，2013）和韩竹军等（2003）$ ］ $
Ⅱ级地块：K1，K2和K3分别为鄂尔多斯、华北平原和鲁东—黄海活动地块；Ⅲ级地块： K2−1，K2−2和K2−3分别为太行山次级、冀鲁次级、豫淮次级活动地块。Ⅰ级活动地块边界带：D1−1为燕山—渤海带，D1−2为阴山带，D1−3为秦岭—大别造山带；Ⅱ级活动地块边界带：D2−1为汾渭断陷带，D2−2为郯庐断裂带；Ⅲ级活动地块边界带：D3−1为河北平原带，D3−2为安阳—菏泽—临沂带
Figure 1. Distribution of active- tectonic blocks and their boundaries in North China （revised from to Zhang et al，2003，2013；Han et al，2003）
The second-grade blocks：K1 is Ordos active block，K2 is North China Plain block，K3 is Ludong-Huanghai active block. The third-grade blocks：K2−1 is Taihangshan secondary active block，K2−2 is Hebei-Shandong secondary active block，K2−3 is Yuhuai secondary active block. The first-grade active boundaries：D1−1 is Yanshan-Bohai seismic zone，D1−2 is Yinshan seismic zone，D1-3 is Qinling-Dabie orogenic zone. The second-grade active boundaries：D2−1 is Fenhe-Weihe down-faulted zone，D2−2 is Tanlu fault zone. The third-grade active boundaries：D3−1 is Hebei plain seismic belt，D3−2 is Anyang-Heze-Linyi seismic zone
![]()
图 2 华北地区机器学习中选取的四个地震活动集中区
HB1：阴山—燕山—渤海带及其附近地区；HB2：汾渭断陷带及其附近地区；HB3：河北平原带及其附近地区；HB4：郯庐断裂带以东及鲁东—黄海活动地块，下同
Figure 2. Four earthquake activity concentration zones in North China selected in machine learning
HB1：Yinshan-Yanshan-Bohai belt and its surrounding areas；HB2： Fenhe-Weihe down-faulted zone and its surrounding areas；HB3：Hebei plain belt and its surrounding areas；HB4：East of Tanlu fault zone and Ludong-Huanghai active-tectonic block，the same below
![]()
图 3 1976年唐山MS7.8地震发生前7年的华北地区ML≥3.0地震活动图像
Figure 3. ML≥3.0 seismic activity images in North China seven years before the Tangshan MS7.8 earthquake in 1976
![]()
图 4 基于华北地区地震活动图像的EQLP机器学习方法流程图
Figure 4. Flowchart of EQLP machine learning method based on seismic activity images in North China
![]()
图 5 1970—2022年华北地区M≥5.0地震的空间分布
Figure 5. Spatial distribution of earthquakes with M≥5.0 in North China from 1970 to 2022
![]()
图 7 预测模型2的损失值和精度
（a） 训练集损失值和精度曲线；（b） 训练集和验证集的损失值
Figure 7. Loss value and accuracy of the prediction model 2
（a） Loss value and accuracy curves of training set；（b） Loss values of the training set and validation set
![]()
图 8 研究选取的6个华北地区地震的震中分布图
Figure 8. Distributions of six earthquakes in North China selected in the study
表 1 华北地区不同时期的目标震级档的地震平均发震间隔统计
Table 1 The statistics on average time interval of earthquakes with different target magnitude intervals in North China at different periods
震级档 有历史记载
以来地震
时间间隔/a1900年以来
地震时间
间隔/a1970年以来
地震时间
间隔/aM5.0—5.9 2 1 1 M6.0—6.9 7 3 2 M≥7.0 28 7 1 
下载: 导出CSV
表 2 基于不同条件地震预测模型的华北地区M5左右地震的震例回溯情况表
Table 2 Retrospective analysis of earthquakes with M5.0 in North China based on different seismic prediction models
序号 震例 时窗长度 基本不同条件模型的地震预测结果 模型预测同实际
目标地震的
对应情况模型1 模型2 1 1 998年1月10日
河北张北MS6.2C 阴山—燕山—渤海带及
附近地区M6.0阴山—燕山—渤海带及
附近地区M6.0模型1 （2次） D 阴山—燕山—渤海带及
附近地区M6.0阴山—燕山—渤海带及
附近地区M6.0模型2 （2次） 2 2 006年7月4日
河北文安MS5.1A 郯庐断裂带以东及
鲁东—黄海活动地块M5.0河北平原带及其附近地区M5.0 模型1 （0次） B 郯庐断裂带以东及
鲁东—黄海活动地块M5.0河北平原带及其附近地区M5.0 模型2 （2次） 3 2 020年7月12日
河北古冶MS5.1A 河北平原带及其附近地区M5.0 河北平原带及其附近地区M5.0 模型1 （2次） B 河北平原带及其附近地区M5.0 河北平原带及其附近地区M5.0 模型2 （2次） 4 2 021年11月17日
江苏大丰海域MS5.0A 郯庐断裂带以东及
鲁东—黄海活动地块M5.0郯庐断裂带以东及
鲁东—黄海活动地块M5.0模型1 （2次） B 郯庐断裂带以东及
鲁东—黄海活动地块M5.0郯庐断裂带以东及
鲁东—黄海活动地块M5.0模型2 （2次） 5 2 023年4月25日
黄海海域ML4.8A 郯庐断裂带以东及
鲁东—黄海活动地块M5.0郯庐断裂带以东及
鲁东—黄海活动地块M5.0模型1 （2次） B 郯庐断裂带以东及
鲁东—黄海活动地块M5.0河北平原带及附近地区M5.0 模型2 （1次） 6 2 023年8月6日
山东德州MS5.5A 郯庐断裂带以东及
鲁东—黄海活动地块M5.0郯庐断裂带以东及
鲁东—黄海活动地块M5.0模型1 （0次） B 郯庐断裂带以东及
鲁东—黄海活动地块M5.0郯庐断裂带以东及
鲁东—黄海活动地块M5.0模型2 （0次） *7 2 023年8月6日
山东德州MS5.5震前3年 河北平原带及附近地区M5.0 郯庐断裂带以东及
鲁东—黄海活动地块M5.0模型1 （2次） 震前7年 河北平原带及附近地区M5.0 郯庐断裂带以东及
鲁东—黄海活动地块M5.0模型2 （0次） 注：模型1为基于未拓展图像数据集训练的模型；模型2为基于含拓展图像数据集的训练的模型。A为震前1个月的前1年的窗长，B为震前1年的窗长，C为震前1个月的前3年的窗长，D为震前3年的窗长。*为补充的相关震例，输入图像的地震时窗增加到大于1年窗长的预测情况。最后一列括号注释为模型1和模型2能够准确预测该目标地震的总次数。
下载: 导出CSV
-
陈会忠. 2020. 我国地震观测历程[J]. 城市与减灾,(6):10–20. Chen H Z. 2020. History of earthquake monitoring in China[J]. City and Disaster Reduction,(6):10–20 (in Chinese).
陈运泰. 1993. 地震预测研究概况[J]. 地震学刊,(1):17–23. Chen Y T. 1993. A survey of research in earthquake prediction[J]. Journal of Seismology,(1):17–23 (in Chinese).
陈章立,刘蒲雄. 1981. 大震前的区域活动特征[C]//国际地震预报讨论会论文选. 北京:地震出版社:197−205. Chen Z L,Liu P X. 1981. Characteristics of regional activity before a major earthquake[C]//Selected Papers of the International Symposium on Earthquake Forecasting. Beijing:Seismological Press:197−205 (in Chinese).
杜菊民,张庆龙,徐士银,杜松金,解国爱. 2009. 阴山晚中生代板内造山特征及其动力机制:以内蒙古大青山为例[J]. 地质学报,83(7):910–922. Du J M,Zhang Q L,Xu S Y,Du S J,Xie A G. 2009. Characters of Later Jurassic Yinshan intraplate orogeny belt:Evidence from Daqingshan,Inner Mongolia[J]. Acta Geologica Sinica,83(7):910–922 (in Chinese).
冯德益,蒋淳,汪德馨,田山,郑熙铭. 1994. 神经网络方法在地震预报研究中的初步应用[J]. 地震,(4):23–29. Feng D Y,Jiang C,Wang D X,Tian S,Zheng X M. 1994. Preliminary application of neural network method in earthquake prediction research[J]. Earthquake,(4):23–29 (in Chinese).
冯锐. 2009. 中国地震科学史研究[J]. 地震学报,31(5):564–582. Feng R. 2009. Research on the history of Chinese seismology[J]. Acta Seismologica Sinica,31(5):564–582 (in Chinese).
冯锐. 2018. 中国近代地震学史纲要[J]. 中国地震,34(2):172–206. Feng R. 2018. A brief history of seismology in Republican China[J]. Earthquake Research in China,34(2):172–206 (in Chinese).
高战武. 2001. 张家口—蓬莱断裂带地震地质特征研究[D]. 北京:中国地震局地质研究所:1−130. Gao Z W. 2001. A Study on Characteristics of Seismo-Geology of the Zhangjiakou-Penglai Fault Zone[D]. Beijing:Institute of Geology,China Earthquake Administration:1−130 (in Chinese).
韩竹军,徐杰,冉勇康,陈立春,杨晓平. 2003. 华北地区活动地块与强震活动[J]. 中国科学(D辑),33(增刊):108–118. Han Z J,Xu J,Ran Y K,Chen L C,Yang X P. 2003. Active blocks and strong seismic activity in North China region[J]. Science in China:Series D,46(S2):153–167. doi: 10.1360/03dz0012
季同仁. 1986. 山东地区地震活动性特征[J]. 地震研究,9(3):289–298. Ji T R. 1986. Seismicity features in Shandong Province[J]. Journal of Seismological Research,9(3):289–298 (in Chinese).
蒋海昆,王锦红. 2023. 适用于机器学习的地震序列类型判定特征重要性讨论[J]. 地震研究,46(2):155–172. Jiang H K,Wang J H. 2023. Discussion on the importance of the features for the judgement of earthquake sequence types applicable to machine learning[J]. Journal of Seismological Research,46(2):155–172 (in Chinese).
蒋海昆,侯海峰,王锜. 2000. 华北地区大范围内中等地震活动平静的统计检验及其预测意义[J]. 内陆地震,14(2):97–104. Jiang H K,Hou H F,Wang Q. 2000. Quiescence of moderate earthquake activity in North China and its significance for earthquake prediction[J]. Inland Earthquake,14(2):97–104 (in Chinese).
李京锦,关晓明,王亮,纪延辉,李哲. 2016. 华北地区震群与强震的空间分布关系研究[J]. 防灾减灾学报,32(2):52–57. Li J J,Guan X M,Wang L,Ji Y H,Li Z. 2016. Swarm in North China and the study on the relationship between the spatial distribution of earthquake[J]. Journal of Disaster Prevention and Reduction,32(2):52–57 (in Chinese).
李林芳,石耀霖,程术. 2022. 长短时记忆神经网络在中期地震预报中的探索:以川滇地区为例[J]. 地球物理学报,65(1):12–25. Li L F,Shi Y L,Cheng S. 2022. Exploration of long short-term memory neural network in intermediate earthquake forecast:A case study in Sichuan-Yunnan region[J]. Chinese Journal of Geophysics,65(1):12–25 (in Chinese).
刘蒲雄,陈章立. 1989. 地震条带及其在地震预报中的作用[J]. 中国地震,5(1):23–32. Liu P X,Chen Z L. 1989. Earthquake belt and its role in earthquake prediction[J]. Earthquake Research in China,5(1):23–32 (in Chinese).
刘蒲雄,陈兆恩,高伟,吕晓健,韩丹. 1997. 大震前地震活动图像演变及中期向短期过渡的地震活动性标志[J]. 地震,17(2):113–125. Liu P X,Chen Z E,Gao W,Lü X J,Han D. 1997. Evolution of seismicity patterns before strong earthquakes and short-term indicator of seismic precursors[J]. Earthquake,17(2):113–125 (in Chinese).
陆远忠,叶金铎,蒋淳,刘杰. 2007. 中国强震前兆地震活动图像机理的三维数值模拟研究[J]. 地球物理学报,50(2):499–508. Lu Y Z,Ye J D,Jiang C,Liu J. 2007. 3D numerical simulation on the mechanism of precursory seismicity pattern before strong earthquake in China[J]. Chinese Journal of Geophysics,50(2):499–508 (in Chinese).
吕伟. 2016. 基于稀疏表示和卷积神经网络的水果图像分类与实现[D]. 广州:华南农业大学:1–67. Lü W. 2016. Fruit Classification and Implementation Based on Sparse Representation and Convolutional Neural Network[D]. Guangzhou:South China Agricultural University:1–67 (in Chinese).
邵志刚,王武星,刘琦,潘正洋,刘晓霞,王芃,魏文薪,冯蔚,尹晓菲. 2022. 活动地块理论框架下的地震物理预报展望[J]. 科学通报,67(13):1362–1377. Shao Z G,Wang W X,Liu Q,Pan Z Y,Liu X X,Wang P,Wei W X,Feng W,Yin X F. 2022. Prospects of earthquake physical forecasting under the framework of active-tectonic block theory[J]. Chinese Science Bulletin,67(13):1362–1377 (in Chinese).
孙其政. 1997. 测震学分析预报方法[M]. 北京:地震出版社:1−130. Sun Q Z. 1997. Seismological Analysis and Forecasting Methods[M]. Beijing:Seismological Press:1−130 (in Chinese).
王霞,宋美琴,陈慧. 2019. 华北地区地震空区的统计分析[J]. 地震,39(3):187–195. Wang X,Song M Q,Chen H. 2019. Statistical analysis of seismic gap in North China[J]. Earthquake,39(3):187–195 (in Chinese).
汪一鹏. 1979. 我国板内地震和中新生代应力场[J]. 地震地质,1(3):1–11. Wang Y P. 1979. Intraplate earthquake and Meso-Cenozoic stress field in China[J]. Seismology and Geology,1(3):1–11 (in Chinese).
魏光兴,周翠英. 1989. 以菏泽5.9级地震为例试论中等强度地震预报问题[J]. 地震,(1):67–69. Wei G X,Zhou C Y. 1989. Discussion on the prediction of moderately intense earthquake as an example[J]. Earthquake,(1):67–69 (in Chinese).
温玉婷,李宁,刘雪琴,吴吉东,张鹏,解伟. 2010. 汶川地震与唐山地震损失与救助之对比[J]. 灾害学,25(2):68–72. Wen Y T,Li N,Liu X Q,Wu J D,Zhang P,Xie W. 2010. Contrast of disaster losses resulted from the Wenchuan and Tangshan earthquakes and rescue actions in these two events[J]. Journal of Catastrophology,25(2):68–72 (in Chinese).
吴刚. 1992. 汾渭断陷带地壳磁性结构研究[J]. 中国地震,8(3):69–73. Wu G. 1992. Research of the crustal magnetic structure in Fen-Wei downfaulted belt[J]. Earthquake Research in China,8(3):69–73 (in Chinese).
徐杰,牛娈芳,王春华,韩竹君. 1996. 唐山—河间—磁县新生地震构造带[J]. 地震地质,18(3):193–198. Xu J,Niu L F,Wang C H,Han Z J. 1996. Tangshan-Hejian-Cixian newly-generated seismotectonic zone[J]. Seismology and Geology,18(3):193–198 (in Chinese).
许可. 2012. 卷积神经网络在图像识别上的应用的研究[D]. 杭州:浙江大学:1−68. Xu K. 2012. Study of Convolutional Neural Network Applied on Image Recognition[D]. Hangzhou:Zhejiang University:1−68 (in Chinese).
薛艳,姜祥华,刘桂萍. 2020. 华北地区强震活动状态研究[J]. 地震,40(2):1–17. Xue Y,Jiang X H,Liu G P. 2020. Active state and trend on strong earthquakes in North China[J]. Earthquake,40(2):1–17 (in Chinese).
杨云,霍祝青,王维,李鸿宇,2016. 鲁东—黄海活动地块背景地震活动及未来强震危险性[J]. 地震工程学报, 38 (增刊):22−29. Yang Y,Huo Z Q,Wang W,Li H Y,2016. Background seismicity and application in seismic hazard assessment in Ludong-Huanghai active block[J]. China Earthquake Engineering Journal, 38 (S1):22−29 (in Chinese).
尹晓菲,张国民,邵志刚,王芃,孙鑫喆. 2020. 华北地区强震活动特点研究[J]. 地震,40(1):11–33. Yin X F,Zhang G M,Shao Z G,Wang P,Sun X Z. 2020. Research on activity characteristics of strong earthquakes in North China[J]. Earthquake,40(1):11–33 (in Chinese).
于书媛,陈靓,方良好,赵朋,张洁. 2015. 梅山—龙河口断裂中西段遥感解译及第四纪活动特征[J]. 防灾科技学院学报,17(2):13–21. Yu S Y,Chen L,Fang L H,Zhao P,Zhang J. 2015. Research on the remote sensing interpretation and active characteristics of middle and west part in Meishan-Longhekou fault in Pleistocene[J]. Journal of Institute of Disaster Prevention,17(2):13–21 (in Chinese).
张国民,张培震. 1999. 近年来大陆强震机理与预测研究的主要进展[J]. 中国基础科学,(增刊):49–60. Zhang G M,Zhang P Z. 1999. Recent research progress on the mechanism and forecast for continental strong earthquakes[J]. Chinese Basic Science,(S1):49–60 (in Chinese).
张国民,马宗晋,蒋铭. 1988. 华北强震规律的研究[J]. 中国地震,(3):72–76. Zhang G M,Ma Z J,Jiang M. 1988. Study on the law of strong earthquake in North China[J]. Earthquake Research in China,(3):72–76 (in Chinese).
张国民,张培震. 2000. “大陆强震机理与预测”中期学术进展[J]. 中国基础科学,(10):4–10. Zhang G M,Zhang P Z. 2000. Academic progress on “the mechanism and forecast for continental strong earthquake in the first two years”[J]. China Basic Science,(10):4–10 (in Chinese).
张国民,李丽,黎凯武,马宏生. 2001. 强震成组活动与潮汐力调制触发[J]. 中国地震,17(2):110–120. Zhang G M,Li L,Li K W,Ma H S. 2001. Group strong earthquakes and triggering by tidal stress[J]. Earthquake Research in China,17(2):110–120 (in Chinese).
张国民,马宏生,王辉,李丽. 2004. 中国大陆活动地块与强震活动关系[J]. 中国科学(D辑),34(7):591–599. Zhang G M, Ma H S, Wang H, Li L. 2004. Relationship between active land mass and strong earthquake activity in mainland China[J]. Science in China: Series D, 34（7）: 591−599 （in Chinese）. Zhang G M,Ma H S,Wang H,Li L. 2004. Relationship between active land mass and strong earthquake activity in mainland China[J]. Science in China:Series D, 34 (7):591−599 (in Chinese).
张国民,马宏生,王辉,王新岭. 2005. 中国大陆活动地块边界带与强震活动[J]. 地球物理学报,48(3):602–610. Zhang G M,Ma H S,Wang H,Wang X L. 2005. Boundaries between active-tectonic blocks and strong earthquakes in the China mainland[J]. Chinese Journal of Geophysics,48(3):602–610 (in Chinese). doi: 10.1002/cjg2.693
张培震. 1999. 中国大陆岩石圈最新构造变动与地震灾害[J]. 第四纪研究,19(5):404–413. Zhang P Z. 1999. Late Quaternary tectonic deformation and earthquake hazard in continental China[J]. Quaternary Sciences,19(5):404–413 (in Chinese).
张培震,邓起东,张国民,马瑾,甘卫军,闵伟,毛凤英,王琪. 2003. 中国大陆的强震活动与活动地块[J]. 中国科学(D辑),33(S1):12–20. Zhang P Z,Deng Q D,Zhang G M,Ma J,Gan W J,Min W,Mao F Y,Wang Q. 2003. Active tectonic blocks and strong earthquakes in the continent of China[J]. Science in China:Series D,46(S2):13–24. doi: 10.1360/03dz0002
张培震,邓起东,张竹琪,李海兵. 2013. 中国大陆的活动断裂、地震灾害及其动力过程[J]. 中国科学(地球科学),43(10):1607–1620. Zhang P Z,Deng Q D,Zhang Z Q,Li H B. 2013. Active faults,earthquake hazards and associated geodynamic processes in continental China[J]. Scientia Sinica Terrae,43(10):1607–1620 (in Chinese). doi: 10.1360/zd-2013-43-10-1607
张晓东. 2004. 中国大陆强震的成组活动特征及发生机制研究[D]. 北京:中国地震局地球物理研究所:1−290. Zhang X D. 2004. Study on Activity and Mechanism of Group Strong Earthquake in China Mainland[D]. Beijing:Institute of Geophysics,China Earthquake Administration:1−290 (in Chinese).
中国地震局监测预报司. 2020. 测震分析预测技术方法工作手册[M]. 北京:地震出版社:1−214. Department of Monitoring and Forecasting, China Earthquake Administration. 2020. Seismic Analysis and Prediction Technical Methods Workbook[M]. Beijing: Seismological Press: 1−214 （in Chinese）. Department of Monitoring and Forecasting,China Earthquake Administration. 2020. Seismic Analysis and Prediction Technical Methods Workbook[M]. Beijing:Seismological Press:1−214 (in Chinese).
朱传镇,傅昌洪,罗胜利. 1981. 震群与大地震关系的研究(华北地区)[J]. 地震学报,3(2):105–117. Zhu C Z,Fu C H,Luo S L. 1981. Study of earthquake swarms in relation to large earthquakes (North China area)[J]. Acta Seismologica Sinica,3(2):105–117 (in Chinese).
朱光,牛漫兰,刘国生,王勇生,谢成龙,李长城. 2005. 郯庐断裂带肥东段走滑运动的40Ar/39Ar法定年[J]. 地质学报,79(3):303–316. Zhu G,Niu M L,Liu G S,Wang Y S,Xie C L,Li C C. 2005. 40Ar/39Ar dating for the strike-slip movement on the Feidong part of the Tanlu fault belt[J]. Acta Geologica Sinica,79(3):303–316 (in Chinese).
朱日祥,陈凌,吴福元,刘俊来. 2011. 华北克拉通破坏的时间、范围与机制[J]. 中国科学(地球科学),41(5):583–592. Zhu R X,Chen L,Wu F Y,Liu J L. 2011. Timing,scale and mechanism of the destruction of the North China Craton[J]. Science China Earth Sciences,54(6):789–797. doi: 10.1007/s11430-011-4203-4
Adeli H,Panakkat A. 2009. A probabilistic neural network for earthquake magnitude prediction[J]. Neural Netw,22(7):1018–1024. doi: 10.1016/j.neunet.2009.05.003
Asim K M,Martínez-Álvarez F,Basit A,Iqbal T. 2017. Earthquake magnitude prediction in Hindukush region using machine learning techniques[J]. Nat Hazards,85(1):471–486. doi: 10.1007/s11069-016-2579-3
Goodfellow I,Bengio Y,Courville A. 2016. Deep Learning[M]. Cambridge:The MIT Press:1−66.
Jordan M I,Mitchell T M. 2015. Machine learning:Trends,perspectives,and prospects[J]. Science,349(6245):255–260. doi: 10.1126/science.aaa8415
Keilis-Borok V I,Kossobokov V G. 1990. Times of increased probability of strong earthquakes (M≥7.5) diagnosed by Algorithm M8 in Japan and adjacent territories[J]. J Geophys Res:Solid Earth,95(B8):12413–12422. doi: 10.1029/JB095iB08p12413
Mignan A,Broccardo M. 2020. Neural network applications in earthquake prediction (1994−2019):Meta-analytic and statistical insights on their limitations[J]. Seismol Res Lett,91(4):2330–2342.
Mogi K. 1979. Two kinds of seismic gaps[J]. Pure Appl Geophys,117(6):1172–1186. doi: 10.1007/BF00876213
Molnar P,Tapponnier P. 1975. Cenozoic tectonics of Asia:Effects of a continental collision,features of recent continental tectonics in Asia can be interpreted as results of the India-Eurasia collision[J]. Science,189(4201):419–426. doi: 10.1126/science.189.4201.419
Panakkat A,Adeli H. 2007. Neural network models for earthquake magnitude prediction using multiple seismicity indicators[J]. Int J Neural Syst,17(1):13–33. doi: 10.1142/S0129065707000890
Reyes J,Morales-Esteban A,Martínez-Álvarez F. 2013. Neural networks to predict earthquakes in Chile[J]. Appl Soft Comput,13(2):1314–1328. doi: 10.1016/j.asoc.2012.10.014
Ross Z E,Meier M A,Hauksson E. 2018. P-wave arrival picking and first-motion polarity determination with deep learning[J]. J Geophys Res:Solid Earth, 123 (6):5120−5129.
Rundle J B,Donnellan A. 2020. Nowcasting earthquakes in southern California with machine learning:Bursts,swarms,and aftershocks may be related to levels of regional tectonic stress[J]. Earth Space Sci,7(9):e2020EA001097. doi: 10.1029/2020EA001097
Seeliger K,Fritsche M,Güçlü U,Schoenmakers S,Schoffelen J M,Bosch S E,Van Gerven M A J. 2018. Convolutional neural network-based encoding and decoding of visual object recognition in space and time[J]. NeuroImage,180:253–266. doi: 10.1016/j.neuroimage.2017.07.018
Shao Z G,Wu Y Q,Ji L Y,Diao F Q,Shi F Q,Li Y J,Long F,Zhang H,Wang W X,Wei W X,Wang P,Liu X X,Liu Q,Pan Z Y,Yun X F,Liu Y,Feng W,Zou Z Y,Cheng J,Lu R Q,Xu Y R,Li X. 2023. Assessment of strong earthquake risk in the Chinese mainland from 2021 to 2030[J]. Earthq Res Adv,3(1):100177. doi: 10.1016/j.eqrea.2022.100177
Teng C T,Chang Y M,Hsu K L,Fan F T. 1979. On the tectonic stress field in China and its relation to plate movement[J]. Phys Earth Planet Inter,18(4):257–273. doi: 10.1016/0031-9201(79)90063-3
Wang Q L,Guo Y F,Yu L X,Li P. 2020. Earthquake prediction based on spatio-temporal data mining:An LSTM network approach[J]. IEEE Trans Emerg Top Comput,8(1):148–158. doi: 10.1109/TETC.2017.2699169
Wyss M,Habermann R E. 1988. Precursory seismic quiescence[J]. Pure Appl Geophys,126(2/3/4):319–332.
Yang L X,Sun S Z. 2020. Seismic horizon tracking using a deep convolutional neural network[J]. J Petrol Sci Eng,187:106709. doi: 10.1016/j.petrol.2019.106709
计量
- 文章访问数: 476
- HTML全文浏览量: 97
- PDF下载量: 115
