中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 武汉岩土力学研究所
  3. 中科院武汉岩土力学所
A CNN-Based Method to Construct a Laterally Heterogeneous S-Wave Velocity Model From Spatially Windowed Surface Wave Data

文献类型：期刊论文

作者Shen, Jian1,2,3; Shi, Zhenming1,3; Liu, Liu2; Peng, Ming1,3; Wang, Dengyi1,2,3; Li, Shaojun2; Tsoflias, Georgios P.
刊名IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
出版日期2024
卷号62页码:15
关键词Training Image resolution Surface waves Geology Imaging Receivers Data models Numerical models Convolutional neural networks Dispersion Convolutional neural network data-driven method lateral heterogeneity local dispersion curve (DC) inversion surface wave
ISSN号0196-2892
DOI10.1109/TGRS.2024.3496855
英文摘要The multichannel analysis of surface waves (MASWs) method has been broadly used to investigate subsurface structures. The inversion of a single dispersion curve (DC) from an individual surface wave shot record can only generate a 1-D velocity model that fails to represent accurately the real geological condition in the presence of lateral velocity heterogeneity. To improve the imaging of sharp lateral variations along 2-D profiles, we propose a convolutional neural network-based-windowed MASW (CNN-WMASW) method. The CNN-WMASW method uses multichannel surface wave data that are spatially windowed by Gaussian functions to extract local DCs. The DCs are inverted simultaneously by the CNN-based data-driven method, which maps the phase velocity to the S-wave velocity from end to end without any initial model. The relationship between adjacent DCs is captured and considered by the CNN model, which is trained on a dataset created by waveform simulation rather than the theoretical solutions of the surface wave propagating mode in the layered model. We tested the effectiveness and reliability of the proposed CNN-WMASW method using numerical examples and two field data examples. Both results show improved lateral resolution in the S-wave velocity maps retrieved by the CNN-WMASW method compared to the traditional multichannel method, in which strong lateral heterogeneity is observed.
资助项目National Natural Science Foundation of China[42207211] ; National Natural Science Foundation of China[42172296] ; Key Laboratory of Geotechnical and Underground Engineering of Ministry of Education, Tongji University[KLE-TJGE-G2304]
WOS研究方向Geochemistry & Geophysics ; Engineering ; Remote Sensing ; Imaging Science & Photographic Technology
语种英语
WOS记录号WOS:001370201100002
出版者IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
源URL[http://119.78.100.198/handle/2S6PX9GI/43366]  
专题中科院武汉岩土力学所
通讯作者Liu, Liu
作者单位1.Tongji Univ, Key Lab Geotech & Underground Engn, Minist Educ, Shanghai 200092, Peoples R China
2.Chinese Acad Sci, Inst Rock & Soil Mech, State Key Lab Geomech & Geotech Engn, Wuhan 430071, Peoples R China
3.Tongji Univ, Coll Civil Engn, Dept Geotech Engn, Shanghai 200092, Peoples R China
推荐引用方式
GB/T 7714		 Shen, Jian,Shi, Zhenming,Liu, Liu,et al. A CNN-Based Method to Construct a Laterally Heterogeneous S-Wave Velocity Model From Spatially Windowed Surface Wave Data[J]. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING,2024,62:15.
APA Shen, Jian.,Shi, Zhenming.,Liu, Liu.,Peng, Ming.,Wang, Dengyi.,...&Tsoflias, Georgios P..(2024).A CNN-Based Method to Construct a Laterally Heterogeneous S-Wave Velocity Model From Spatially Windowed Surface Wave Data.IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING,62,15.
MLA Shen, Jian,et al."A CNN-Based Method to Construct a Laterally Heterogeneous S-Wave Velocity Model From Spatially Windowed Surface Wave Data".IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING 62(2024):15.

入库方式： OAI收割

来源：武汉岩土力学研究所

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。