Geophysical Prospecting最新文献

{"title":"Borehole Logging Lithological Classification in Carbonatite-Hosted Niobium Deposit Integrating Geophysical and Geochemical Data With Self-Organizing Maps","authors":"Fabrizzio R. Costa,&nbsp;Lívia Namie S. Nishimura,&nbsp;Reginaldo Antonio R. Filho,&nbsp;Bruno P. Milanezi,&nbsp;Dong Yaosong,&nbsp;Cleyton de C. Carneiro,&nbsp;Carina Ulsen","doi":"10.1111/1365-2478.70159","DOIUrl":"https://doi.org/10.1111/1365-2478.70159","url":null,"abstract":"<p>Lithological classification from borehole data remains challenging in geologically complex alkaline–carbonatite systems, where compositional overlap and gradual transitions between lithotypes limit conventional interpretation methods. This study presents an integrated and automated workflow for drill-core lithological classification using self-organizing maps (SOMs), an unsupervised artificial neural network, applied to a carbonatite-hosted niobium deposit. The approach integrates 17 variables, including major-element geochemistry (e.g., Nb₂O₅, Fe₂O₃, P₂O₅ and SiO₂), physical properties (bulk density) and downhole geophysical logs (gamma radiation, magnetic susceptibility, resistivity and neutron porosity), enabling multivariate analysis of more than 40,000 drill-core samples. Principal component analysis (PCA) highlighted robust associations between mineralization-related variables, whereas inverse correlations of SiO₂, K₂O and resistivity enhanced host-rock discrimination. SOM modelling defined five clusters validated against reference lithologies, including nelsonites, carbonatite breccias, calcitic and barite-rich carbonatites and host rocks. The model accurately classified more than 95% of nelsonite and barite-rich carbonatite samples, with errors below ±5% for key variables such as Nb₂O₅ and Fe₂O₃, whereas partial overlap between breccias and host rocks reflected natural geological transitions. Overall, the results demonstrate that SOM-based unsupervised learning provides a robust, objective and reproducible framework for lithological domain definition, offering strong potential to enhance geological interpretation, exploration targeting and geometallurgical modelling in complex ore systems.</p>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 3","pages":""},"PeriodicalIF":1.8,"publicationDate":"2026-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/1365-2478.70159","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147566580","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Borehole Logging Lithological Classification in Carbonatite-Hosted Niobium Deposit Integrating Geophysical and Geochemical Data With Self-Organizing Maps","authors":"Fabrizzio R. Costa,&nbsp;Lívia Namie S. Nishimura,&nbsp;Reginaldo Antonio R. Filho,&nbsp;Bruno P. Milanezi,&nbsp;Dong Yaosong,&nbsp;Cleyton de C. Carneiro,&nbsp;Carina Ulsen","doi":"10.1111/1365-2478.70159","DOIUrl":"https://doi.org/10.1111/1365-2478.70159","url":null,"abstract":"<p>Lithological classification from borehole data remains challenging in geologically complex alkaline–carbonatite systems, where compositional overlap and gradual transitions between lithotypes limit conventional interpretation methods. This study presents an integrated and automated workflow for drill-core lithological classification using self-organizing maps (SOMs), an unsupervised artificial neural network, applied to a carbonatite-hosted niobium deposit. The approach integrates 17 variables, including major-element geochemistry (e.g., Nb₂O₅, Fe₂O₃, P₂O₅ and SiO₂), physical properties (bulk density) and downhole geophysical logs (gamma radiation, magnetic susceptibility, resistivity and neutron porosity), enabling multivariate analysis of more than 40,000 drill-core samples. Principal component analysis (PCA) highlighted robust associations between mineralization-related variables, whereas inverse correlations of SiO₂, K₂O and resistivity enhanced host-rock discrimination. SOM modelling defined five clusters validated against reference lithologies, including nelsonites, carbonatite breccias, calcitic and barite-rich carbonatites and host rocks. The model accurately classified more than 95% of nelsonite and barite-rich carbonatite samples, with errors below ±5% for key variables such as Nb₂O₅ and Fe₂O₃, whereas partial overlap between breccias and host rocks reflected natural geological transitions. Overall, the results demonstrate that SOM-based unsupervised learning provides a robust, objective and reproducible framework for lithological domain definition, offering strong potential to enhance geological interpretation, exploration targeting and geometallurgical modelling in complex ore systems.</p>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 3","pages":""},"PeriodicalIF":1.8,"publicationDate":"2026-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/1365-2478.70159","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147566588","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Audio-Frequency Magnetotellurics Signal Denoising Based on Multilevel Wavelet Convolutional Neural Networks","authors":"Liang Zhang,&nbsp;Zhengguang Liu,&nbsp;Guangyin Lu,&nbsp;Jingtian Tang,&nbsp;Mingbiao Yu,&nbsp;Hao Zhang,&nbsp;Jing Sun","doi":"10.1111/1365-2478.70158","DOIUrl":"https://doi.org/10.1111/1365-2478.70158","url":null,"abstract":"<div>\u0000 \u0000 <p>Audio-frequency magnetotellurics (AMT) is one of the commonly used methods in geophysical exploration; however, its signal energy is relatively weak and easily submerged by various cultural noises, making denoising a critical step in AMT data processing. Currently, deep learning-based neural networks have achieved superior denoising performance compared to traditional methods in many fields, but in AMT denoising, the neglect of the sparsity characteristic of cultural noise results in degraded denoising performance. To enable the neural network to consider sparsity features during the denoising process and thereby enhance denoising accuracy, we adopt a convolutional neural network (CNN) as the network backbone and design a multilevel wavelet convolutional neural network (MWCNN) from the perspective of sparse representation. This network improves CNN blocks via shortcut connections and enhances feature transmission efficiency by replacing pooling layers and interpolation with wavelet transforms, thereby enabling the network to account for the sparsity of cultural noise, capture underlying noise spectral information and improve denoising performance. Furthermore, we discuss the influence of various network parameters on denoising performance. Finally, we validate the effectiveness of MWCNN in AMT denoising through comparative experiments on both synthetic and field AMT datasets against wavelet transform, bounded influence remote reference processing, data-driven tight frame, CNN and residual networks. Comprehensive evaluations based on signal-to-noise ratio, wavelet time-frequency spectra, denoised results and residuals, apparent resistivity and phase curves, error analysis, one-dimensional inversion results and Nyquist diagrams confirm the superiority of MWCNN for AMT denoising.</p>\u0000 </div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 3","pages":""},"PeriodicalIF":1.8,"publicationDate":"2026-03-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147566453","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Audio-Frequency Magnetotellurics Signal Denoising Based on Multilevel Wavelet Convolutional Neural Networks","authors":"Liang Zhang,&nbsp;Zhengguang Liu,&nbsp;Guangyin Lu,&nbsp;Jingtian Tang,&nbsp;Mingbiao Yu,&nbsp;Hao Zhang,&nbsp;Jing Sun","doi":"10.1111/1365-2478.70158","DOIUrl":"https://doi.org/10.1111/1365-2478.70158","url":null,"abstract":"<div>\u0000 \u0000 <p>Audio-frequency magnetotellurics (AMT) is one of the commonly used methods in geophysical exploration; however, its signal energy is relatively weak and easily submerged by various cultural noises, making denoising a critical step in AMT data processing. Currently, deep learning-based neural networks have achieved superior denoising performance compared to traditional methods in many fields, but in AMT denoising, the neglect of the sparsity characteristic of cultural noise results in degraded denoising performance. To enable the neural network to consider sparsity features during the denoising process and thereby enhance denoising accuracy, we adopt a convolutional neural network (CNN) as the network backbone and design a multilevel wavelet convolutional neural network (MWCNN) from the perspective of sparse representation. This network improves CNN blocks via shortcut connections and enhances feature transmission efficiency by replacing pooling layers and interpolation with wavelet transforms, thereby enabling the network to account for the sparsity of cultural noise, capture underlying noise spectral information and improve denoising performance. Furthermore, we discuss the influence of various network parameters on denoising performance. Finally, we validate the effectiveness of MWCNN in AMT denoising through comparative experiments on both synthetic and field AMT datasets against wavelet transform, bounded influence remote reference processing, data-driven tight frame, CNN and residual networks. Comprehensive evaluations based on signal-to-noise ratio, wavelet time-frequency spectra, denoised results and residuals, apparent resistivity and phase curves, error analysis, one-dimensional inversion results and Nyquist diagrams confirm the superiority of MWCNN for AMT denoising.</p>\u0000 </div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 3","pages":""},"PeriodicalIF":1.8,"publicationDate":"2026-03-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147566024","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Self-Supervised Seismic Data Reconstruction With Blind-Trace Mask Mapper and Attention","authors":"Xinyu Cai,&nbsp;Anxiang Di,&nbsp;Chunxia Zhang,&nbsp;Xiaoli Wei,&nbsp;Baisong Jiang,&nbsp;Jiangshe Zhang,&nbsp;Deng Xiong","doi":"10.1111/1365-2478.70155","DOIUrl":"https://doi.org/10.1111/1365-2478.70155","url":null,"abstract":"<div>\u0000 \u0000 <p>High-quality seismic data are crucial for interpreting subsurface structures. Due to economic costs, geological obstacles and other factors, the acquired seismic data are often sampled regularly or irregularly along spatial coordinates. Such incomplete data pose significant challenges to subsequent seismic data processing workflows, making seismic data reconstruction one of the most cost-effective and critical techniques to address this issue in seismic exploration. While deep learning has yielded promising results in data reconstruction, supervised methods rely heavily on large volumes of labelled seismic data – a resource that is scarce in practical applications. To tackle this limitation, this study proposes a self-supervised learning method for seismic data reconstruction, incorporating a global-aware blind-trace mask mapper and squeeze-and-excitation (SE) attention blocks. First, we introduce a global-aware blind-trace mask mapper, which mitigates the identity mapping problem in self-supervised learning frameworks and leverages global information to enable efficient reconstruction. This mapper samples all traces at the blind-trace positions on the interpolated volume and maps them to the same channel, allowing the loss function to optimize all blind traces simultaneously. Second, we adopt a nested U-Net (UNet++) architecture integrated with SE attention blocks as the interpolation network, enhancing the model's capacity to capture key features. Third, we guide network training using a hybrid loss function combining mean absolute error and structural similarity index, which captures both fine-grained details and global structural information to ensure reconstruction quality. Experiments on synthetic and field datasets demonstrate that our method outperforms existing approaches in interpolation accuracy.</p></div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 3","pages":""},"PeriodicalIF":1.8,"publicationDate":"2026-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147564406","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Self-Supervised Seismic Data Reconstruction With Blind-Trace Mask Mapper and Attention","authors":"Xinyu Cai,&nbsp;Anxiang Di,&nbsp;Chunxia Zhang,&nbsp;Xiaoli Wei,&nbsp;Baisong Jiang,&nbsp;Jiangshe Zhang,&nbsp;Deng Xiong","doi":"10.1111/1365-2478.70155","DOIUrl":"https://doi.org/10.1111/1365-2478.70155","url":null,"abstract":"<div>\u0000 \u0000 <p>High-quality seismic data are crucial for interpreting subsurface structures. Due to economic costs, geological obstacles and other factors, the acquired seismic data are often sampled regularly or irregularly along spatial coordinates. Such incomplete data pose significant challenges to subsequent seismic data processing workflows, making seismic data reconstruction one of the most cost-effective and critical techniques to address this issue in seismic exploration. While deep learning has yielded promising results in data reconstruction, supervised methods rely heavily on large volumes of labelled seismic data – a resource that is scarce in practical applications. To tackle this limitation, this study proposes a self-supervised learning method for seismic data reconstruction, incorporating a global-aware blind-trace mask mapper and squeeze-and-excitation (SE) attention blocks. First, we introduce a global-aware blind-trace mask mapper, which mitigates the identity mapping problem in self-supervised learning frameworks and leverages global information to enable efficient reconstruction. This mapper samples all traces at the blind-trace positions on the interpolated volume and maps them to the same channel, allowing the loss function to optimize all blind traces simultaneously. Second, we adopt a nested U-Net (UNet++) architecture integrated with SE attention blocks as the interpolation network, enhancing the model's capacity to capture key features. Third, we guide network training using a hybrid loss function combining mean absolute error and structural similarity index, which captures both fine-grained details and global structural information to ensure reconstruction quality. Experiments on synthetic and field datasets demonstrate that our method outperforms existing approaches in interpolation accuracy.</p></div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 3","pages":""},"PeriodicalIF":1.8,"publicationDate":"2026-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147564631","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Sonic Attenuations of qP and qSV Waves in Fractured Rock","authors":"Guangquan Li","doi":"10.1111/1365-2478.70156","DOIUrl":"10.1111/1365-2478.70156","url":null,"abstract":"<div>\u0000 \u0000 <p>In situ rocks often contain fractures/joints aligned in certain directions. When sonic waves propagate in such anisotropic rocks, fluid can squirt between fractures and the contiguous matrix. In this article, matrix pores are referred to as the primary porosity, whereas fractures are referred to as the secondary porosity. A new model of quasi-P (qP) and quasi-SV (qSV) waves in the band of sonic frequency is constructed, with the full consideration of mechanical anisotropy arising from orientated fractures. Because the mathematical problem involves seven partial differential equations, the method of matrices yielding from plane waves is used for solving the problem. The model is capable of predicting not only sonic qP-wave attenuation but also sonic qSV-wave attenuation. The sonic log of a sandstone reservoir in Saudi Arabia is used as an application example. The modelling successfully reproduces the measured velocities of qP and qSV waves at 10 kHz, as well as the measured quality factors of qP wave (100) and qSV wave (21). Remarkably, the modelling yields a dip angle of 63° at depths of 2.4–2.5 km, which is consistent with the tensional stress, the deep basement and the sandstone nature of the reservoir. It is demonstrated that the new model is capable of well regenerating sonic attenuations of qP and qSV waves in porous rock containing tensional fractures.</p>\u0000 </div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 3","pages":""},"PeriodicalIF":1.8,"publicationDate":"2026-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147562779","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"WAHNet: A Wave-Attention Hybrid Network for Surface Wave Suppression in Seismic Data","authors":"Lei Gao,&nbsp;Shijun Zheng,&nbsp;Fan Min","doi":"10.1111/1365-2478.70151","DOIUrl":"10.1111/1365-2478.70151","url":null,"abstract":"<div>\u0000 \u0000 <p>Noise in seismic data includes random noise and coherent noise. Surface waves, as typical coherent noise, are characterized by low-frequency, low apparent velocity and high amplitude. Its physical propagation characteristics are similar to some effective signals, which leads to the difficulty of distinguishing them in the time-frequency domain. Convolutional neural networks have been increasingly used for surface wave suppression due to their adaptive feature extraction. However, they struggle in separating surface waves from effective signals with their overlapping frequency range. In this paper, we propose a wave-attention hybrid network (WAHNet) to handle this issue. WAHNet has two types of new modules: (1) wavelet-gated enhancement module extracts and enhances the time-frequency domain features of seismic signals by combining wavelet convolution with a gating mechanism and (2) adaptive wavelet attention module introduces a joint attention mechanism in the time-frequency domain to alleviate the time-frequency aliasing problem of surface waves and effective signals. The experimental results show that, compared with the five popular methods, WAHNet can suppress surface waves more effectively while retaining richer effective signals.</p></div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 3","pages":""},"PeriodicalIF":1.8,"publicationDate":"2026-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147562780","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"WAHNet: A Wave-Attention Hybrid Network for Surface Wave Suppression in Seismic Data","authors":"Lei Gao,&nbsp;Shijun Zheng,&nbsp;Fan Min","doi":"10.1111/1365-2478.70151","DOIUrl":"https://doi.org/10.1111/1365-2478.70151","url":null,"abstract":"<div>\u0000 \u0000 <p>Noise in seismic data includes random noise and coherent noise. Surface waves, as typical coherent noise, are characterized by low-frequency, low apparent velocity and high amplitude. Its physical propagation characteristics are similar to some effective signals, which leads to the difficulty of distinguishing them in the time-frequency domain. Convolutional neural networks have been increasingly used for surface wave suppression due to their adaptive feature extraction. However, they struggle in separating surface waves from effective signals with their overlapping frequency range. In this paper, we propose a wave-attention hybrid network (WAHNet) to handle this issue. WAHNet has two types of new modules: (1) wavelet-gated enhancement module extracts and enhances the time-frequency domain features of seismic signals by combining wavelet convolution with a gating mechanism and (2) adaptive wavelet attention module introduces a joint attention mechanism in the time-frequency domain to alleviate the time-frequency aliasing problem of surface waves and effective signals. The experimental results show that, compared with the five popular methods, WAHNet can suppress surface waves more effectively while retaining richer effective signals.</p></div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 3","pages":""},"PeriodicalIF":1.8,"publicationDate":"2026-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147562982","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Sonic Attenuations of qP and qSV Waves in Fractured Rock","authors":"Guangquan Li","doi":"10.1111/1365-2478.70156","DOIUrl":"https://doi.org/10.1111/1365-2478.70156","url":null,"abstract":"<div>\u0000 \u0000 <p>In situ rocks often contain fractures/joints aligned in certain directions. When sonic waves propagate in such anisotropic rocks, fluid can squirt between fractures and the contiguous matrix. In this article, matrix pores are referred to as the primary porosity, whereas fractures are referred to as the secondary porosity. A new model of quasi-P (qP) and quasi-SV (qSV) waves in the band of sonic frequency is constructed, with the full consideration of mechanical anisotropy arising from orientated fractures. Because the mathematical problem involves seven partial differential equations, the method of matrices yielding from plane waves is used for solving the problem. The model is capable of predicting not only sonic qP-wave attenuation but also sonic qSV-wave attenuation. The sonic log of a sandstone reservoir in Saudi Arabia is used as an application example. The modelling successfully reproduces the measured velocities of qP and qSV waves at 10 kHz, as well as the measured quality factors of qP wave (100) and qSV wave (21). Remarkably, the modelling yields a dip angle of 63° at depths of 2.4–2.5 km, which is consistent with the tensional stress, the deep basement and the sandstone nature of the reservoir. It is demonstrated that the new model is capable of well regenerating sonic attenuations of qP and qSV waves in porous rock containing tensional fractures.</p>\u0000 </div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 3","pages":""},"PeriodicalIF":1.8,"publicationDate":"2026-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147562983","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}