{"title":"Convolutional neural network-based moment tensor inversion using domain adaptation for microseismicity monitoring","authors":"Jihun Choi, J. Byun, S. Seol, Seong Kon Lee","doi":"10.1080/08123985.2022.2086798","DOIUrl":null,"url":null,"abstract":"Microseismic monitoring is widely used to analyze the locations and growth directions of fractures formed at sites of hydraulic fracturing treatment and CO2 geologic sequestration. Because moment tensors can provide focal mechanisms, moment tensor inversion has received considerable attention in microseismic monitoring; the real-time processing of moment tensor inversion is important for rapid decision-making. Pre-trained machine learning (ML) models can make nearly instantaneous predictions in the application stage and thus present an attractive alternative to real-time processing. However, prior information regarding the velocity model at the target site is a prerequisite for generating the dataset used to train the ML model that is applied in moment tensor inversion. In addition, it is difficult to create the training dataset because it requires three-dimensional numerical modelling when the velocity model is complex; numerous simulations must be executed for sources with various locations and moment tensors. To overcome these limitations, we applied the domain adaptation technique to the convolutional neural network (CNN)-based moment tensor inversion method, which uses peak amplitudes and arrival times of P- and S-waves as input features. The CNN model was pre-trained with the dataset generated from a homogeneous velocity model. Then, in the domain adaptation stage, the pre-trained model was fine-tuned along with the target dataset. To validate the performance of the domain adaptation, moment tensors from both horizontal and tilted three-layer models were predicted. In each case, the domain-adapted model performance was similar to the performances of the CNN-based models that had been trained using the dataset generated with the exact target velocity models.","PeriodicalId":50460,"journal":{"name":"Exploration Geophysics","volume":"54 1","pages":"133 - 143"},"PeriodicalIF":0.6000,"publicationDate":"2022-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Exploration Geophysics","FirstCategoryId":"89","ListUrlMain":"https://doi.org/10.1080/08123985.2022.2086798","RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"GEOCHEMISTRY & GEOPHYSICS","Score":null,"Total":0}
引用次数: 1
Abstract
Microseismic monitoring is widely used to analyze the locations and growth directions of fractures formed at sites of hydraulic fracturing treatment and CO2 geologic sequestration. Because moment tensors can provide focal mechanisms, moment tensor inversion has received considerable attention in microseismic monitoring; the real-time processing of moment tensor inversion is important for rapid decision-making. Pre-trained machine learning (ML) models can make nearly instantaneous predictions in the application stage and thus present an attractive alternative to real-time processing. However, prior information regarding the velocity model at the target site is a prerequisite for generating the dataset used to train the ML model that is applied in moment tensor inversion. In addition, it is difficult to create the training dataset because it requires three-dimensional numerical modelling when the velocity model is complex; numerous simulations must be executed for sources with various locations and moment tensors. To overcome these limitations, we applied the domain adaptation technique to the convolutional neural network (CNN)-based moment tensor inversion method, which uses peak amplitudes and arrival times of P- and S-waves as input features. The CNN model was pre-trained with the dataset generated from a homogeneous velocity model. Then, in the domain adaptation stage, the pre-trained model was fine-tuned along with the target dataset. To validate the performance of the domain adaptation, moment tensors from both horizontal and tilted three-layer models were predicted. In each case, the domain-adapted model performance was similar to the performances of the CNN-based models that had been trained using the dataset generated with the exact target velocity models.
期刊介绍:
Exploration Geophysics is published on behalf of the Australian Society of Exploration Geophysicists (ASEG), Society of Exploration Geophysics of Japan (SEGJ), and Korean Society of Earth and Exploration Geophysicists (KSEG).
The journal presents significant case histories, advances in data interpretation, and theoretical developments resulting from original research in exploration and applied geophysics. Papers that may have implications for field practice in Australia, even if they report work from other continents, will be welcome. ´Exploration and applied geophysics´ will be interpreted broadly by the editors, so that geotechnical and environmental studies are by no means precluded.
Papers are expected to be of a high standard. Exploration Geophysics uses an international pool of reviewers drawn from industry and academic authorities as selected by the editorial panel.
The journal provides a common meeting ground for geophysicists active in either field studies or basic research.