Article
Geochemistry & Geophysics
Lijing Wang, Frederic Joncour, Pierre-Emmanuel Barrallon, Thibault Harribey, Laurent Castanie, Sonia Yousfi, Sebastien Guillion
Summary: Seismic interpretation is crucial in locating subsurface horizons and understanding geologic formations. This study develops a semisupervised segmentation framework to utilize unlabeled seismic data for better subsurface horizon prediction. Data augmentation and deep ensembles are also employed to improve the prediction accuracy and reduce uncertainty. The findings suggest that this approach can help geophysicists achieve higher facies classification accuracy with reduced labeling efforts.
Article
Environmental Sciences
Chengliang Wu, Bo Feng, Xiaonan Song, Huazhong Wang, Rongwei Xu, Shen Sheng
Summary: Picking the reflection horizon is crucial in velocity inversion and seismic interpretation, but manual picking is inefficient for large-scale seismic data. This paper proposes a novel method using multiple seismic attributes and the Markov decision process (MDP) for automatic horizon picking. By considering cumulative reward, the lateral continuity of a seismic image can be effectively considered, and the impacts of abnormal waveform changes or bad traces in local areas for automatic horizon picking can be effectively avoided.
Article
Chemistry, Multidisciplinary
Wenliang Nie, Fei Xiang, Bo Li, Xiaotao Wen, Xiangfei Nie
Summary: In this study, a nonconvex L1-2 regularization method is innovatively applied to three-parameter prestack AVA inversion, addressing the uncertainty in seismic inversion by adding a priori constraints and regularization methods. The accuracy and stability of the method are validated through synthetic and observational data sets, demonstrating its potential as a useful tool in predicting reservoir locations.
APPLIED SCIENCES-BASEL
(2021)
Article
Geochemistry & Geophysics
Lele Pan, Jinghuai Gao, Yang Yang, Zhiguo Wang, Zhaoqi Gao
Summary: Seismic lithology interpretation based on seismic data is a challenging task due to its instability and multiple solutions. This study proposes a multiattribute integrated deep learning (MAIDL) workflow that combines wavelet scattering transform (WST) and seismic data to improve accuracy. The proposed model also incorporates residual blocks to address over-fitting and degradation issues. Testing on synthetic and field data demonstrates the effectiveness of the MAIDL model for automatic seismic lithology interpretation.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2023)
Article
Geochemistry & Geophysics
Zhengxiang He, Pingan Peng, Liguan Wang, Yuanjian Jiang
Summary: PickCapsNet is a scalable capsule network for P-wave arrival picking without feature extraction, employing recent advances in artificial intelligence. The method demonstrates high accuracy and stability in the identification of P-wave arrival times, outperforming other methods in practical applications.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2021)
Article
Geochemistry & Geophysics
Haoran Zhang, Tiansheng Chen, Yang Liu, Yuxi Zhang, Jiong Liu
Summary: Seismic facies interpretation using supervised deep learning algorithms, specifically enhanced encoder-decoder architecture, has been shown to be more accurate and efficient than traditional CNN methods, producing results that better match the geological background. Testing on the Netherlands F3 dataset demonstrated significantly higher mean intersection-over-union values for the encoder-decoders compared to conventional CNN, with faster prediction times.
Article
Geochemistry & Geophysics
Wenlong Wang, George A. McMechan, Jianwei Ma, Fei Xie
Summary: This study introduces the application of machine learning in root-mean-square (rms) velocity estimation, designing and evaluating classification and regression neural networks for extracting velocity trajectories from semblance data. Testing with synthetic data shows that the regression network achieves higher accuracy compared to the classification network.
Article
Mathematics
Luca Martino, Fernando Llorente, Ernesto Curbelo, Javier Lopez-Santiago, Joaquin Miguez
Summary: A novel adaptive importance sampling scheme is proposed for Bayesian inversion problems, where variables of interest and data noise power are inferred using different methods. The technique involves iterative steps of sampling and optimization, with the noise power acting as a tempered parameter for the posterior distribution of the variables of interest. Numerical experiments show the benefits of the proposed approach in Bayesian analysis.
Article
Computer Science, Interdisciplinary Applications
Yamei Cao, Hui Zhou, Bo Yu
Summary: A decorated linearized seismic-petrophysics inversion (DLSPI) method is proposed based on a linearized seismic-petrophysics model and principal component analysis (PCA). This method can improve the accuracy of petrophysical properties estimation by avoiding the statistical correlation between different parameters.
COMPUTERS & GEOSCIENCES
(2023)
Article
Energy & Fuels
Daiane Rossi Rosa, Denis Jose Schiozer, Alessandra Davolio
Summary: In this study, Bayesian time-lapse seismic inversion was applied to a deep-water heavy oil field in the Campos Basin, Brazil. The results showed a significant improvement in vertical resolution and revealed important geological features and fluid movement. This information is crucial for updating geological models and calibrating dynamic models in the future.
JOURNAL OF PETROLEUM SCIENCE AND ENGINEERING
(2022)
Article
Geochemistry & Geophysics
Wei Xiang, Xingyao Yin, Zhengqian Ma, Kun Li, Song Pei
Summary: This article investigates the monoclinic medium with vertical fractures using numerical simulation and inversion methods. The fracture weaknesses and Thomsen parameters are successfully estimated by the linearized approximate formula and Bayesian inference. The results demonstrate the stability and feasibility of the proposed method in analyzing the problems of vertical fractures and background anisotropy in seismic data.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Environmental Sciences
Wuxin Xiao, Katy Louise Sheen, Qunshu Tang, Jamie Shutler, Richard Hobbs, Tobias Ehmen
Summary: The study uses seismic oceanography to invert submesoscale dynamics in the ocean, extracting temperature and salinity information. By adapting an iteratively updated prior model based on hydrographic data, the study overcomes the challenge of scarce continuous reflection horizons, improving the accuracy of inversion techniques. Uncertainties in the results are mainly influenced by the resolution of hydrographic data and regions such as the edge of the Mediterranean eddy.
FRONTIERS IN MARINE SCIENCE
(2021)
Article
Geochemistry & Geophysics
Eduardo Valero Cano, Jubran Akram, Daniel B. Peter
Summary: A new fully automated four-step workflow has been developed for efficient classification and picking of P-and S wave arrival times on microseismic data sets, yielding accurate arrival picks for high and low signal-to-noise ratio waveforms.
Article
Geosciences, Multidisciplinary
Ahmed Zidan, Yunyue Li, Arthur Cheng
Summary: This paper proposes a method to improve the amplitude inversion problem by using gradient descent and regularization functions, which enhances the consistency between the inversion results and the a priori geological information. The method is successfully applied to estimate the elastic and seismic anisotropy parameters of shale formations.
JOURNAL OF APPLIED GEOPHYSICS
(2023)
Article
Chemistry, Multidisciplinary
Wenqi Gao, Youxue Wang, Yang Yang, Sanxi Peng, Songping Yu, Lu Liu, Lei Yan
Summary: We used the ray tracing technique based on the IASP91 Earth model to calculate travel times and identify seismic phases. This technique accurately calculates travel times for seismic phases found in conventional travel time tables. We analyzed and discussed waveform data received from stations in the Guangxi area. Our numerical modeling results show good agreement with the theoretical travel times from the ISAP91 tables. The validity of using correlation analysis to pick seismic phases and determine travel times is demonstrated.
APPLIED SCIENCES-BASEL
(2023)
Article
Geochemistry & Geophysics
Yongchae Cho, Richard L. Gibson
Article
Geosciences, Multidisciplinary
Hyunggu Jun
JOURNAL OF APPLIED GEOPHYSICS
(2019)
Article
Geosciences, Multidisciplinary
Hyunggu Jun, Yongchae Cho, Joocheul Noh
JOURNAL OF MARINE SYSTEMS
(2019)
Article
Geosciences, Multidisciplinary
Han-Joon Kim, Chung-Ho Kim, Tianyao Hao, Lihua Liu, Kwang-Hee Kim, Hyunggu Jun, Hyeong-Tae Jou, Sunghoon Moon, Ya Xue, Zhiqiang Wu, Chuanchuan Lu, Sang Hoon Lee
JOURNAL OF ASIAN EARTH SCIENCES
(2019)
Article
Geochemistry & Geophysics
Yongchae Cho, Richard L. Gibson, Hyunggu Jun, Changsoo Shin
Article
Energy & Fuels
Yongchae Cho, Hyunggu Jun
Summary: Many scientists have developed technology to store CO2 in the subsurface and monitor the storage conditions for zero detectable leakage and greenhouse gas control. This study proposes a novel workflow to estimate stored CO2 volume and quantify uncertainty of injected volume using geophysical stochastic inversion with time-lapse 3D seismic volumes as inputs. Validation was performed using the Sleipner project, which showed that the predicted subsurface CO2 volume linearly increases in phase with injection rate and volume estimation error is less than 17%.
JOURNAL OF PETROLEUM SCIENCE AND ENGINEERING
(2021)
Article
Geochemistry & Geophysics
Yongchae Cho
Summary: The researcher developed an integrated and stochastic method that combines seismic and well log data to predict natural fracture networks, successfully constructing a reliable natural fracture model that can be used for subsequent geomechanical analysis.
INTERPRETATION-A JOURNAL OF SUBSURFACE CHARACTERIZATION
(2021)
Article
Geochemistry & Geophysics
Hyunggu Jun, Yongchae Cho
Summary: A machine learning-based time-lapse seismic data processing method is proposed, utilizing convolutional autoencoder for feature analysis and data augmentation to enhance data repeatability and accuracy in target change analysis.
GEOPHYSICAL JOURNAL INTERNATIONAL
(2022)
Article
Geochemistry & Geophysics
Yongchae Cho, Sean S. Dolan, Nishank Saxena, Vishal Das
Summary: This study investigates the limitation of conventional logging tools in estimating water saturation in formations with low conductivity water and tight porosity. A dielectric permittivity is proposed as an alternative that can differentiate water from other fluids in pore structures. By using a multichannel frequency dielectric logging tool and two analytic models, a workflow is proposed to evaluate an equivalent Archie's parameter that can match the responses of the dielectric logging tool with core measurement results.
Article
Geochemistry & Geophysics
Yongchae Cho, Carlos Perez Solano, John Kimbro, Yi Yang, Rene-Edouard Plessix, Kenneth Matson
Summary: Elastic full-waveform inversion is crucial in building velocity models, especially in areas with large salt bodies. It helps reduce artifacts caused by using acoustic approximation, but defining the initial shear velocity model remains challenging. By utilizing reflected converted waves from the top of salt interface, we can improve the estimation of background shear velocity, and combining wave-equation traveltime inversion with FWI yields higher-resolution results.
Article
Geochemistry & Geophysics
Honggeun Jo, Yongchae Cho, Michael Pyrcz, Hewei Tang, Pengcheng Fu
Summary: Estimating porosity models from seismic data is challenging due to low signal-to-noise ratio and insufficient resolution. In this paper, a machine learning-based workflow is proposed to convert seismic data into porosity models. The workflow uses a residual U-Net++ architecture to estimate porosity models from multiple poststack seismic volumes. Experimental results show that the method achieves high accuracy and robustness in porosity estimation, but further research and improvements are needed.
Article
Meteorology & Atmospheric Sciences
Hyunggu Jun, Hyeong-Tae Jou, Chung-Ho Kim, Sang Hoon Lee, Han-Joon Kim
Article
Geochemistry & Geophysics
Hyunggu Jun, Yongchae Cho
Summary: In this study, a machine learning-based method for processing time-lapse seismic data is proposed to enhance repeatability and capture accurate seismic information. The method involves steps such as training data construction, uniform manifold approximation, and data augmentation to improve the repeatability of time-lapse seismic data and accurately analyze seismic information.
GEOPHYSICAL JOURNAL INTERNATIONAL
(2021)