Geophysical Prospecting最新文献

{"title":"A Self-Training U-Net Approach for First-Break Picking With Minimal Annotations","authors":"André Meneses,&nbsp;Ramon C. F. Araújo,&nbsp;Gilberto Corso,&nbsp;João M. de Araújo,&nbsp;Tiago Barros","doi":"10.1111/1365-2478.70233","DOIUrl":"https://doi.org/10.1111/1365-2478.70233","url":null,"abstract":"<div>\u0000 \u0000 <p>First-break picking is critical in seismic data processing. Traditional techniques such as short-term average/long-term average often fail in noisy environments, while deep learning methods like U-Net require large labelled datasets. We propose a site-specific self-training framework with three contributions: (1) a windowed labelling strategy expanding point annotations into 11-sample windows, reducing class imbalance by an order of magnitude; (2) a weighted loss function penalizing missed picks 100-fold more than false positives; and (3) an iterative self-training procedure augmenting training data with quality-controlled predictions. Using only 1% of gathers for manual labelling, our U-Net method achieves over 98% prediction coverage and above 97% accuracy (<span></span><math>\u0000 <semantics>\u0000 <mrow>\u0000 <mo>±</mo>\u0000 <mn>10</mn>\u0000 </mrow>\u0000 <annotation>$pm 10$</annotation>\u0000 </semantics></math> samples) across four mining exploration datasets. Compared to cross-site transfer approaches requiring fully annotated surveys, we achieve comparable first-break picking accuracy with two orders of magnitude less labelled data, particularly when acquisition parameters differ between sites.</p></div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 7","pages":""},"PeriodicalIF":1.7,"publicationDate":"2026-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148784164","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":"Seismic Wave Migration in Visco-Acoustic VTI Media: A One-Way Operator With De Wolf Renormalization Scattering Series Approximation","authors":"Huachao Sun,&nbsp;Bo Wang,&nbsp;Linfeng Zeng","doi":"10.1111/1365-2478.70240","DOIUrl":"https://doi.org/10.1111/1365-2478.70240","url":null,"abstract":"<div>\u0000 \u0000 <p>Real earth media cause absorption attenuation and anisotropy. Ignoring these effects leads to amplitude and imaging-position errors in migration profiles. The classical Born approximation generates large errors under the strong forward accumulative effect of visco-acoustic VTI (Transversely Isotropic media with a Vertical symmetry axis) media. Therefore, this article integrates the De Wolf approximation with visco-acoustic VTI media for the first time. We renormalize the Born series using the De Wolf approximation and establish its integral representation for visco-acoustic VTI media, thereby improving the accuracy of the Born approximation. Following the thin-slab approximation, the velocity model is divided into several thin slabs, each containing background parameters (velocity, quality factor and anisotropy) and perturbation parameters. Dual-domain screen approximation migration operators (frequency-wavenumber and frequency-space domains) are then constructed for visco-acoustic VTI media, leading to a prestack migration imaging method based on the De Wolf approximation. Algorithm tests on the Overthrust and Hess models show that the proposed migration operator effectively compensates for amplitude attenuation and corrects anisotropic effects, producing migration results with higher signal-to-noise ratio and resolution. The specific effects of viscosity and anisotropy on imaging are explicitly analysed based on comparative data of migration profiles, waveforms and amplitude spectra, further validating the algorithm's effectiveness.</p>\u0000 </div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 7","pages":""},"PeriodicalIF":1.7,"publicationDate":"2026-08-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148783493","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":"In Situ Stress Prediction Using Multiple Seismic Attributes Based on the Random Forest Algorithm","authors":"Shiqi Peng,&nbsp;Suping Peng,&nbsp;Chuangjian Li,&nbsp;Xiaoqin Cui,&nbsp;Jie Yang,&nbsp;Tijmen Jan Moser","doi":"10.1111/1365-2478.70235","DOIUrl":"https://doi.org/10.1111/1365-2478.70235","url":null,"abstract":"<div>\u0000 \u0000 <p>In situ stress characterisation is crucial for the exploration and development of coal resources. Conventional methods based on well-log data offer high-accuracy point measurements but are limited to wellbore locations, whereas traditional seismic-based predictions provide extensive spatial coverage but have lower accuracy and often fail to capture the complex, non-linear relationships between seismic attributes and the stress field. To bridge this gap, we introduce a data-driven approach that integrates the strengths of both data types using a random forest (RF) model. In our approach, geomechanical attributes (Young's modulus and Poisson's ratio) and geometric attributes (curvature), derived from pre-stack inversion, serve as the model's input features. High-resolution stress values calculated from well logs using the combined spring model serve as the training labels. The RF model, optimised via grid search and cross-validation, demonstrates high predictive accuracy. We apply the proposed RF-based method to a field dataset from the Daji area in Shanxi Province, North China, successfully generating a continuous three-dimensional (3D) in situ stress volume. The resulting stress field exhibits strong spatial consistency with regional tectonic features, validating the model's accuracy and geological applicability. This study demonstrates that a machine learning framework can effectively link seismic data with well-log-derived reference stress labels, extending sparse reference stress information into a continuous 3D volume that reliably characterises the inter-well stress heterogeneity. This provides a practical and effective framework for in situ stress analysis in complex geological settings.</p>\u0000 </div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 7","pages":""},"PeriodicalIF":1.7,"publicationDate":"2026-08-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148753937","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":"Deep Learning Matching Filtering for Elastic Full Waveform Inversion","authors":"Chao Li,&nbsp;Yangkang Chen","doi":"10.1111/1365-2478.70236","DOIUrl":"https://doi.org/10.1111/1365-2478.70236","url":null,"abstract":"<div>\u0000 \u0000 <p>Elastic full waveform inversion (EFWI) aims to reconstruct high-resolution elastic properties by minimizing the waveform misfit between observed and simulated multicomponent seismic data, yet its practical performance is often limited by cycle skipping and modelling errors that conventional misfit-based methods cannot fully resolve. We propose a deep-learning-based matching-filtering framework, in which a lightweight neural network is trained to learn adaptive, data-driven matching filters that map synthetic elastic wavefields towards the observed data. Instead of directly enforcing waveform agreement, the learned filters absorb phase, amplitude and dispersion discrepancies arising from inaccurate initial models, elastic-parameter trade-offs and imperfect physics. The matching filters are optimized jointly with elastic parameters within an AD framework, enabling seamless gradient propagation through both the wave-equation solver and the neural components. Numerical experiments on synthetic elastic models and field data demonstrate that the proposed approach significantly reduces cycle skipping and improves the recovery of P- and S-wave velocities compared to conventional EFWI, particularly when the initial model is strongly biased. The results suggest that deep learning–based matching filtering provides a physically interpretable and computationally efficient pathway for robust elastic waveform inversion under realistic modelling uncertainties.</p></div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 7","pages":""},"PeriodicalIF":1.7,"publicationDate":"2026-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148753010","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":"Integrated Multi-Physics Characterisation of the Northern Kiruna Mining District With Machine-Learning-Assisted Interpretation","authors":"Oskar Rydman,&nbsp;Ervin Veress,&nbsp;Maxim Yu. Smirnov,&nbsp;Tobias E. Bauer,&nbsp;Thorkild M. Rasmussen,&nbsp;Niklas Juhojuntti","doi":"10.1111/1365-2478.70234","DOIUrl":"https://doi.org/10.1111/1365-2478.70234","url":null,"abstract":"<p>Integrated earth modelling aims to combine all available geoscientific information to derive one common earth model. However, combining multiple parameters in a meaningful way remains challenging. In this study, we aim to improve mineral exploration workflows through the integration of new magnetotelluric data with other geophysical and petrophysical data to enhance local geological understanding of the northern Kiruna mining district, Norrbotten, Sweden. The area is economically important and geoscientifically interesting. We derived three-dimensional (3D) geophysical models based on new magnetotelluric data and previously collected gravity and magnetic data. The magnetotelluric data and the resulting 3D electrical-resistivity model are described in detail. Additionally, a new petrophysical dataset containing surface and subsurface samples and their density, electrical resistivity and magnetic susceptibility is presented and linked to the 3D geophysical models. The geophysical 3D models are interpreted using information gained from the measured rock properties, allowing for correlation and cross-validation between models and geology. Integration of a clustering approach into the mineral exploration workflow allows for a holistic interpretation of all information available in the area. These interpretations and models improve understanding of the Per Geijer mineral system, its immediate surroundings and the relationships between local rock types and their geophysical and mineralogical signatures.</p>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 7","pages":""},"PeriodicalIF":1.7,"publicationDate":"2026-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/1365-2478.70234","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148752608","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":"Time-Reversal Full Wavefield Inversion: From Waveform Misfit to Wavefield Focus","authors":"Evgeny Landa,&nbsp;Sergey Fomel","doi":"10.1111/1365-2478.70237","DOIUrl":"https://doi.org/10.1111/1365-2478.70237","url":null,"abstract":"<p>We introduce time-reversal full wavefield inversion (TRFWI), a velocity-model-building method that replaces waveform-misfit minimization with a physically motivated wavefield-focusing criterion. The recorded data are reversed in time and backpropagated through a trial velocity model; the inversion objective is the degree to which the wavefield focuses on the known source position and onset time. This Research Note presents a proof-of-concept demonstration using a simple acoustic example. The focusing functional exhibits a clear maximum at the true velocity, confirming that time-reversal focusing provides a viable alternative to classical full waveform inversion (FWI). Unlike waveform-misfit objectives, the focusing criterion naturally incorporates the full recorded wavefield, including multiples and scattered energy as constructive information. Optimization strategies and field-data applications will be addressed in a subsequent publication.</p>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 7","pages":""},"PeriodicalIF":1.7,"publicationDate":"2026-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/1365-2478.70237","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148617207","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":"Motion Sensitivity Analysis and Terrain Correction for UAV-Borne Gamma-Ray Spectrometry","authors":"Neeraj Nainwal,&nbsp;Alexander Braun,&nbsp;Marshall MacNabb","doi":"10.1111/1365-2478.70232","DOIUrl":"https://doi.org/10.1111/1365-2478.70232","url":null,"abstract":"<p>Uncrewed aerial vehicles (UAVs) equipped with gamma-ray spectrometry sensors enable high-resolution surveys but face challenges in rugged terrain. Conventional airborne gamma-ray processing commonly approximates the source–detector geometry using a single height-above-ground parameter. This approximation is often unsuitable for low-altitude UAV surveys, where rapid topographic changes, variable flight altitude, and platform motion can influence the measured response. This study first evaluates platform motion using a moving point-spread function sensitivity analysis to quantify footprint elongation associated with survey speed, integration time, and flight altitude. A three-dimensional, grid-based terrain-response model is then developed to account for variations in source–detector distance, terrain geometry, and line-of-sight visibility. Controlled synthetic simulations are used to evaluate the model under known terrain and source configurations. The motion analysis shows that footprint elongation is small under typical survey conditions but increases with higher survey speed, longer integration time, and lower flight altitude. The synthetic simulations further show how terrain geometry changes the measured response and demonstrate that the terrain-response model recovers the known source distribution. When applied to UAV radiometric data collected over a rugged porphyry copper–molybdenum deposit in Arizona, USA, terrain-correction magnitudes reached approximately 15% in areas of pronounced topographic relief. The largest adjustments occurred on slopes, ridges, and in topographic depressions, indicating that terrain effects were the primary source of geometric variation under the conditions examined. The proposed workflow provides a practical approach for identifying and correcting geometric effects in low-altitude UAV gamma-ray surveys over complex terrain.</p>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 7","pages":""},"PeriodicalIF":1.7,"publicationDate":"2026-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/1365-2478.70232","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148617210","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":"Adaptive Ground-Roll Suppression Based on Adaptive Chirp Mode Decomposition in the Fx Domain","authors":"Zhiwei Li,&nbsp;Ying Shi,&nbsp;Ning Wang,&nbsp;Siyuan Chen,&nbsp;Zhaojun Song","doi":"10.1111/1365-2478.70222","DOIUrl":"https://doi.org/10.1111/1365-2478.70222","url":null,"abstract":"<div>\u0000 \u0000 <p>Ground roll severely affects the signal-to-noise ratio of seismic data. Characterized by low frequency, strong energy and low-velocity, it significantly masks the primary reflections. Ground roll manifests as a non-stationary, frequency-modulated signal in the <i>fx</i> domain. To separate the intrinsic mode functions (IMFs) containing ground roll, we propose a pre-stack ground-roll suppression method based on adaptive chirp mode decomposition in the <i>fx</i> domain (<i>fx</i>-ACMD). The <i>fx</i>-ACMD algorithm iteratively optimizes through three steps: demodulation, adaptive filtering and reconstruction. This framework incorporates greedy recursive decomposition and adaptive bandwidth updating. The application of time-varying demodulation operators makes it particularly suitable for adaptive mode decomposition of non-stationary signals. By further exploiting the localized high-wavenumber characteristic of ground roll in the <i>fx</i> domain, we propose <i>fx</i>-adaptive ACMD (<i>fx</i>-AACMD). The <i>fx</i>-AACMD operates without manual intervention and applies the data-driven IMFs selection strategy to automatically identify ground-roll-containing IMFs. Through localized filtering, it effectively suppresses ground roll while maintaining stability and preserving amplitudes. Tests on synthetic and field data confirm that the proposed methods achieve effective suppression of ground roll.</p>\u0000 </div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 7","pages":""},"PeriodicalIF":1.7,"publicationDate":"2026-07-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148616038","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":"SARAS-PINN: Enhancing Traveltime Modelling Through Structure-Aware Initialization and Residual-and-Gradient-Based Adaptive Sampling","authors":"Yanjin Xiang,&nbsp;Zhiliang Wang,&nbsp;Ziang Song,&nbsp;Rong Huang,&nbsp;Guojie Song,&nbsp;Fan Min,&nbsp;Shulin Pan,&nbsp;Lei Zhao","doi":"10.1111/1365-2478.70230","DOIUrl":"https://doi.org/10.1111/1365-2478.70230","url":null,"abstract":"<div>\u0000 \u0000 <p>Accurate seismic traveltime modelling in geologically complex media is fundamental to seismic imaging and inversion. While physics-informed neural networks (PINNs) have emerged as a promising mesh-free alternative to traditional eikonal solvers, they often suffer from low accuracy near sharp velocity contrasts and slow convergence in complex models. In this study, we develop SARAS-PINN, a structure-aware and residual-gradient-adaptive sampling framework designed to enhance PINN-based traveltime modelling. Our method integrates two key innovations: (1) structure-aware initialization using elliptical exclusion zones, where axes are dynamically scaled by local velocity and directional gradients to prioritize structural features; and (2) residual-and-gradient-based adaptive sampling, which leverages both the local partial differential equation residual and its spatial gradient to refine the training set iteratively. Numerical experiments on layered, Marmousi and BP models demonstrate significant improvements. In the layered model, SARAS-PINN reduces the mean absolute error (MAE) from 0.1274 to 0.0086 ms, a 15-fold improvement in precision. For the Marmousi model, our approach achieves a 3.6-fold computational speedup in reaching the target MAE (<span></span><math>\u0000 <semantics>\u0000 <mrow>\u0000 <mo>≤</mo>\u0000 <mn>2.5</mn>\u0000 </mrow>\u0000 <annotation>$le 2.5$</annotation>\u0000 </semantics></math> ms) compared to standard PINN. Furthermore, the inclusion of the residual gradient allows for the precise capture of sharp interfaces and salt bodies that are typically undersampled by traditional residual-only methods (e.g., residual-based adaptive refinement with distribution). These results confirm that SARAS-PINN provides a robust and efficient tool for high-precision seismic simulations.</p></div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 6","pages":""},"PeriodicalIF":1.7,"publicationDate":"2026-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148615906","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":"Acoustic Impedance Inversion Method Based on Learnable Mismatch Function","authors":"Zixiang Zhou,&nbsp;Guochang Liu,&nbsp;Zhichao Li,&nbsp;Qibin Wu,&nbsp;Jiahe Lu","doi":"10.1111/1365-2478.70224","DOIUrl":"https://doi.org/10.1111/1365-2478.70224","url":null,"abstract":"<div>\u0000 \u0000 <p>Acoustic impedance inversion is a core technique for reservoir prediction in seismic exploration, which is fundamentally a nonlinear, ill-posed inverse problem. In traditional inversion methods, the choice of the mismatch function between observed and simulated data directly affects the accuracy of the inversion. The pointwise comparison based on the L2 norm fails to fully utilize the structural information embedded in the data. To address this issue, we propose an acoustic impedance inversion method based on a learnable mismatch function. This method abandons the traditional paradigm of directly comparing seismic data and introduces a dual-branch Siamese network that maps observed seismic records and forward-modelled convolutional records into a shared latent feature space. The Euclidean distance between extracted features is used as a mismatch metric, enabling data-driven learning of the mismatch function. Unlike existing inversion methods based on fixed norms, the proposed method employs a self-supervised learning framework that does not require labelled impedance data for training, significantly reducing dependence on well-log data. In the design of the Siamese network, we introduce the ConvKAN module to replace some traditional convolutional layers, leveraging its learnable nonlinear activation function based on B-spline basis functions to enhance the network's feature-expression capability for weak impedance contrasts. Similarly, a total variation regularization is introduced to constrain the lateral continuity of the inversion results. Finally, the performance of the proposed method is validated through acoustic impedance inversion experiments on synthetic and field data.</p></div>","PeriodicalId":12793,"journal":{"name":"Geophysical Prospecting","volume":"74 6","pages":""},"PeriodicalIF":1.7,"publicationDate":"2026-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148615942","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}