Applied Computing and Geosciences最新文献

{"title":"CAGEY (CArbonate Grain Estimation with YOLO): Object detection for grain size, roundness, and dunham classification","authors":"Bingxuan Liu ,&nbsp;Haifa AlSalmi ,&nbsp;Cédric M. John","doi":"10.1016/j.acags.2026.100329","DOIUrl":"10.1016/j.acags.2026.100329","url":null,"abstract":"<div><div>Accurate grain size and roundness analysis is crucial for geological applications such as paleoenvironmental reconstructions or reservoir characterization, but traditional manual methods are time-consuming and prone to error. In this study, we introduce the CAGEY (CArbonate Grain Estimation with YOLO) framework to automate carbonate grain size and shape (roundness) estimates from core images with YOLOv5s. In addition, CAGEY integrates the predicted grain size and shape distributions in downstream classification of Dunham textures using Random Forest. We built a dataset of 114 high-resolution carbonate core images, yielding 46,974 labelled grains split into 82 training images (37,367 bounding box labels for grains), 8 validation images (2164 bounding box labels for grains), and 24 test images (7443 bounding box labels for grains). Our trained YOLOv5s model demonstrated strong performance in predicting grain sizes across various Dunham textures in our test set, yielding an R² of 0.95 on grain area ratios predictions with successful approximated grain size distributions. Performance was lower for finer-grained rocks, showing higher Wasserstein distances and mean squared errors. The grain roundness estimates derived from the prediction boxes also closely follow the labelled distribution. On lithology classification with Random Forest, CAGEY achieves an overall accuracy of 76% with an expected cost lower (better) than humans on the same task. Future improvements to the CAGEY framework could including the use of segmentation masks for better grain morphological accuracy, and improved annotation consistency to reduce training bias. But our research demonstrate that useable, low-cost grain-level characterization has strong potential for downstream carbonate lithology characterization and classification, offering an alternative to direct lithologic classification on images with deep learning.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"29 ","pages":"Article 100329"},"PeriodicalIF":3.2,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147396664","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Benchmarking lightweight deep networks for real-time flood segmentation in unmanned aerial vehicle imagery","authors":"Alireza Sharifi","doi":"10.1016/j.acags.2026.100319","DOIUrl":"10.1016/j.acags.2026.100319","url":null,"abstract":"<div><div>The ability to generate up-to-the-minute flood maps using remote sensing imagery is crucial for efficient environmental monitoring and catastrophe response. In order to quickly segregate floods from aerial and Unmanned Aerial Vehicle (UAV) images, this paper provides a comprehensive evaluation of two lightweight convolutional neural network architectures, namely Fast-SCNN and BiSeNetV2. Using a publicly accessible flood segmentation dataset, I establish an end-to-end pipeline and evaluate the two models for accuracy (Intersection over Union (IoU), Dice coefficient, F1-score) and speed (inference time per picture). With a validation set F1-score of 0.738 and a peak IoU of 0.608, extensive trials demonstrate that Fast-SCNN outperforms all other segmentation methods. Although BiSeNetV2's inference time is 0.27 ms per picture, it achieves comparable results. Although both algorithms are good at detecting flooded areas, qualitative research shows that they are not very good at defining complex borders or small flood patches. Findings highlight benefits and drawbacks of operational flood monitoring using state-of-the-art real-time semantic segmentation networks. This provides helpful details for implementing remote sensing-based automated flood detection systems.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"29 ","pages":"Article 100319"},"PeriodicalIF":3.2,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188294","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing soil moisture prediction for monitoring biological signals: A frequency-domain and spatial weighting approach","authors":"Xiaoning Li ,&nbsp;Mengyuan Zhang ,&nbsp;Zhichao Zhong ,&nbsp;Qingliang Li ,&nbsp;Xiaolin Li","doi":"10.1016/j.acags.2026.100328","DOIUrl":"10.1016/j.acags.2026.100328","url":null,"abstract":"<div><div>Biological signals—such as plant electrical activity and microbial byproducts—are increasingly valued as indicators of how ecosystems respond to environmental change. These signals are closely linked to soil moisture, a key environmental factor that shapes plant hydration, microbial function, and root-soil interactions. As such, accurately predicting soil moisture is essential for monitoring ecosystem health and understanding the dynamic interplay between living organisms and their environment. Despite recent advances, many existing prediction models struggle with two main limitations: sensitivity to high-frequency noise and poor spatial generalisation. To address these issues, we present a new modelling approach that combines frequency-domain filtering with spatial weighting. By applying a Fourier transform, our method filters out high-frequency noise while preserving low-frequency patterns that are ecologically and biologically relevant. In parallel, a cosine Latitude-weighted loss function is introduced to balance spatial representation during model training. We evaluated this strategy across four deep learning models. We found consistent gains across all four deep-learning models, with the Lat-FT-LSTM (i.e., the LSTM model combining Fourier transform and Latitude-weighted loss) performing best. Relative to the baseline (LSTM), the coefficient of determination (R²) rose by 6.45% and the Kling-Gupta Efficiency (KGE) by 6.31%, while unbiased RMSE and Bias fell by 10.22% and 13.33%, respectively. Improvements were most substantial in biologically active regions—North Africa, the Middle East, Central Asia, and North America—suggesting that combining frequency-domain filtering with latitude-based weighting improves both accuracy and transferability. Together, these results suggest that integrating basic signal processing and spatial weighting into the modelling workflow can enhance ecosystem monitoring and the interpretation of biological responses.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"29 ","pages":"Article 100328"},"PeriodicalIF":3.2,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The arctic knowledge-based system: Science gateway integration for petascale arctic data processing and geospatial feature prediction","authors":"Andrew Wilcox ,&nbsp;Meisam Shayeghmoradi ,&nbsp;Stephen Miller ,&nbsp;Ian Nesbitt ,&nbsp;Saisri Pogalla ,&nbsp;Aymane Ahajjam ,&nbsp;Walker McKee ,&nbsp;Sheridan Parker ,&nbsp;Matthew Johnson ,&nbsp;Aaron Bergstrom ,&nbsp;Naima Kaabouch ,&nbsp;Timothy Pasch","doi":"10.1016/j.acags.2026.100322","DOIUrl":"10.1016/j.acags.2026.100322","url":null,"abstract":"<div><div>Science gateways have become essential platforms that integrate computational resources, data services, and workflows for domain researchers, enabling artificial intelligence-driven (AI) analyses at scale. Building on this paradigm, we introduce the Science Gateway component of the Arctic Knowledge-Based System (A-KBS), designed to advance AI-assisted modeling of permafrost dynamics and other Arctic geospatial processes. The A-KBS provides researchers with a unified portal to configure and execute multi-horizon prediction tools for active layer thickness, ground deformation, wildfire occurrence, freeze/thaw states, soil and air temperature analyses, and to run global scale geospatial HPC workflows leveraging data from across the circumpolar Arctic. This system orchestrates workloads through Kubernetes-based (K8s) containerized environments, Globus Data Transfer/Compute services, and distributed computing tools such as Slurm, ParSL, and Ray.io. Its web portal is deployed on the University of North Dakota’s (UND) virtualized, load-balanced K8s cluster with cloud migration enabled by Rancher, while Python-based AI functions authored in JupyterHub are executed on remote systems through Globus Compute Endpoints. Current development has integrated the A-KBS with the UND high-performance computing (HPC) Talon cluster and Amazon Web Services-managed K8s resources, with a roadmap in place to extend this integration to other HPC environments including the San Diego Supercomputer Center’s Expanse System. By coupling scalable infrastructure with a suite of existing AI-driven workflows for environmental prediction tasks, the A-KBS accelerates Arctic science, strengthens cryospheric research, and supports decision-making through its integration with cyberinfrastructure surrounding the DRP (Defense Resiliency Platform Against Extreme Cold Weather) initiative.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"29 ","pages":"Article 100322"},"PeriodicalIF":3.2,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188347","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Synergizing Radon transform and DINOv2 for artifact-resilient digital rock segmentation","authors":"Shuai Hou ,&nbsp;Danping Cao ,&nbsp;Zhiyu Hou ,&nbsp;Yali Zhu ,&nbsp;Ziqiang Wang","doi":"10.1016/j.acags.2025.100315","DOIUrl":"10.1016/j.acags.2025.100315","url":null,"abstract":"<div><div>Precise segmentation of core images acquired by X-ray computed tomography (X-CT) is fundamental for subsequent digital rock physics analysis. However, in cores with abundant high-density minerals, insufficient X-ray penetration often leads to distorted pore structures and severe artifacts, which compromise segmentation accuracy. To address this challenge, we propose a hybrid method that synergizes Radon transform with the DINOv2 vision foundation model. Our approach employs two stage strategy: first, a ‘hard correction’ via Radon transform to leverage the directional features of artifacts and suppress directional artifact patterns in the sinogram domain; second, a “soft learning” mechanism, where the fine-tuned DINOv2 model leverages its global semantic priors to discern and rectify residual artifact interference in the feature space. This hybrid approach not only suppresses artifact interference but also preserves the continuity and visual plausibility of pore structures. Experimental results demonstrate that, on synthetic artifact-contaminated samples, the model achieves incremental three-phase segmentation performance (mIoU = 0.755, F1 = 92.15 %). While Appling to real core images, with heavy mineral artifacts, the model demonstrates robust error-correcting capabilities, yielding segmentations that align more closely with the expected rock microstructure than the original ground truth annotations. This study provides an effective and validated framework for artifact resilient digital rock segmentation, offering substantial improvements for quantitative DRP studies.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"29 ","pages":"Article 100315"},"PeriodicalIF":3.2,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145977261","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Interactive 3D simulation of Taal Volcano eruption plumes","authors":"Neil Patrick Del Gallego,&nbsp;Onesimus Dela Torre,&nbsp;Carlos Miguel Arquillo,&nbsp;Vaughn Vincent Cordero,&nbsp;Joaquin Lorenzo Sales,&nbsp;Stephen Lance Yarte","doi":"10.1016/j.acags.2026.100331","DOIUrl":"10.1016/j.acags.2026.100331","url":null,"abstract":"<div><div>The Taal Volcano in the Philippines is one of the most active volcanoes due to its recurrent, sometimes highly destructive eruptions, especially the 12 January 2020 phreatomagmatic eruption classified as VEI=4. Even until 2025, minor phreatic eruptions were recorded in Taal. Thus, there is a need to communicate its hazards and disaster preparedness effectively, and one possible method would be 3D computer graphics simulations. In this work, we present the first interactive, real-time 3D graphics-based simulation tailored for Taal’s eruption plumes, titled AnitoPlume. We quantitatively validate AnitoPlume through real-world eruption footages and user experience metrics, achieving a Structural Similarity Index (SSIM) of 0.871, a Root Mean Square Error (RMSE) of 61.764/255.0, and a System Usability Scale (SUS) of 72, indicating the above-average visual similarity and potential usability of our proposed simulator. We demonstrate applications of AnitoPlume by recreating observable characteristics of the 2020 Taal eruption. Through our simulator, we aim to provide a valuable, lightweight 3D visualization software for disaster communication and educational awareness to one of the Philippines’ most active volcanoes.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"29 ","pages":"Article 100331"},"PeriodicalIF":3.2,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147396661","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhanced anomaly detection in well log data through the application of ensemble GANs","authors":"Abdulrahman Al-Fakih ,&nbsp;Ardiansyah Koeshidayatullah ,&nbsp;Tapan Mukerji ,&nbsp;SanLinn I. Kaka","doi":"10.1016/j.acags.2025.100316","DOIUrl":"10.1016/j.acags.2025.100316","url":null,"abstract":"<div><div>Detecting subtle anomalies in well log data can significantly enhance subsurface characterization and inform reservoir decision-making. Well logs provide detailed records of rock and fluid properties. However, identifying anomalous patterns in these data remains a persistent challenge due to the complexity of geological systems and limitations in traditional statistical models. Classical anomaly detection approaches, such as Gaussian mixture models (GMMs), tend to oversimplify the intricacies of well log responses and often misinterpret geological heterogeneities as tool noise. This misclassification leads to high false positive rates and unreliable anomaly identification. In this context, generative models, particularly generative adversarial networks (GANs), have shown promise in structured data domains; however, they remain underutilized in geoscience applications. This study aims to benchmark the performance of ensemble GANs (EGANs) against GMMs for anomaly detection in well log data. The proposed EGANs framework aggregates multiple independently trained GANs to enhance model stability and robustness. Anomalies are detected based on discriminator scoring, while performance is evaluated using precision, recall, and F1-score across four key logs: gamma ray, travel time, bulk density, and neutron porosity. The results demonstrate that EGANs consistently outperform GMMs across all logs, achieving higher precision (up to 0.70) and F1-scores (up to 0.79), with statistically significant improvements confirmed via paired t-tests. These findings highlight the ability of EGANs to model complex subsurface patterns and detect subtle deviations more effectively than conventional probabilistic methods. This study introduces the first application of EGANs to petrophysical anomaly detection, bridging deep learning with geoscience workflows. It offers a scalable framework for integrating data-driven anomaly detection into reservoir modeling, quality control, and near-real-time decision-making in drilling operations. Future work will focus on multivariate analysis, cross-basin validation, and real-time deployment, advancing toward a more intelligent, adaptive reservoir monitoring system.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"29 ","pages":"Article 100316"},"PeriodicalIF":3.2,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145977264","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Geochemical anomaly delineation utilizing copula-based outlier detection method","authors":"Shahed Shahrestani ,&nbsp;Emmanuel John M. Carranza ,&nbsp;Ioan Sanislav","doi":"10.1016/j.acags.2026.100325","DOIUrl":"10.1016/j.acags.2026.100325","url":null,"abstract":"<div><div>This study evaluates the effectiveness of Copula-Based Outlier Detection (COPOD) in identifying geochemical anomalies within the Toroud–Chah Shirin belt (TCSB) in Iran. The TCSB is a significant mineralized zone containing epithermal precious and base metal veins, skarn, gold placer, and Pb–Zn sedimentary-hosted deposits. Unlike proximity-based or learning-based models, COPOD is a fully deterministic and unsupervised statistical approach. It requires no hyperparameter tuning or assumptions regarding data distribution, making it ideal for the skewed, non-Gaussian nature of stream sediment datasets. By modeling multivariate dependencies through empirical cumulative distribution functions (ECDFs), COPOD captured complex element relationships, such as Ag–Pb and Bi–Au, which relate to sedimentary-hosted and epithermal gold deposits in the region. Comparative analysis using Receiver Operating Characteristic (ROC) curves demonstrates that COPOD outperforms both traditional uni-element mapping and the state-of-the-art Isolation Forest (IF) method. Using a 10% contamination threshold, the COPOD method identified 23 out of 32 known mineral occurrences, whereas the IF method captured 19. Furthermore, this study uses dimensional outlier graphs to provide transparent results, highlighting the influence of Co, Zn, Sb, and Pb on anomaly scores. Results from Lasso regression and random forest analysis further confirmed these elemental impacts. Comparison with the regional geological map shows that most anomalies occur within Paleogene volcanic units and the Cretaceous sedimentary unit that hosts Pb–Zn mineralization. However, some extend into surficial areas due to geochemical dispersion. Overall, COPOD offers a robust, efficient, and explainable alternative for multivariate geochemical anomaly delineation.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"29 ","pages":"Article 100325"},"PeriodicalIF":3.2,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188293","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Generation and validation of a high-resolution fossil diatom record from Saliña Bartol, Bonaire, based on convolutional neural networks","authors":"Romée van der Kuil ,&nbsp;Kees Nooren ,&nbsp;Madlene Nussbaum ,&nbsp;Edwin Bennink ,&nbsp;Francesca Sangiorgi ,&nbsp;Timme Donders","doi":"10.1016/j.acags.2026.100326","DOIUrl":"10.1016/j.acags.2026.100326","url":null,"abstract":"<div><div>Diatom microfossil remains preserved in lake and marine sediments are widely used as proxies for reconstructing past environmental conditions. However, generating high-resolution diatom records is time-consuming and relies on the potentially subjective view of expert taxonomists. Recent developments in deep learning have led to an increase in proof-of-concept studies using object detection models for diatom classification and segmentation, but their practical use in creating reliable, continuous paleoenvironmental records has yet to be demonstrated.</div><div>Here, we present the first application of an object detection model (YOLO11) trained on fossil diatom images to create a high-resolution diatom record for Saliña Bartol, a hypersaline lake in Bonaire. The model was trained on 3242 annotated objects representing 34 diatom taxa, and achieved a mean F1-score of 0.946.</div><div>To evaluate performance in a real-world paleoenvironmental context, the model was first applied to create a continuous record of 22 virtual microscope slides that were also manually counted. Comparison of automated and manual counts showed highly consistent patterns in downcore relative species abundance, with differences mostly caused by larger or morphologically variable taxa. The model was applied to the full dataset of 399 virtual slides, producing a continuous diatom record spanning <span><math><mo>∼</mo></math></span>905 years, with a resolution of <span><math><mo>∼</mo></math></span>2 years per sample.</div><div>The complete record reveals decadal-scale oscillations in aerophilous and brackish-water indicator species, suggesting variability in the hydrology of the catchment that in a traditional, low-resolution analysis would not have been detected. These results demonstrate that deep learning can be applied to automate fossil diatom quantification on a scale that manual analysis could not realistically achieve, marking the beginning of a new era in applying deep learning to create diatom-based paleoenvironmental reconstructions.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"29 ","pages":"Article 100326"},"PeriodicalIF":3.2,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147396665","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Corrigendum to “Assessing future flood hazards in the major rivers of Bangladesh using CMIP6 projections and integrated hydrologic-hydraulic modeling” [Appl. Comput. Geosci., Volume 29 (2026), 100327]","authors":"Aishia Fyruz Aishi ,&nbsp;Md Kamruzzaman Tusar ,&nbsp;Md Ariful Islam","doi":"10.1016/j.acags.2026.100330","DOIUrl":"10.1016/j.acags.2026.100330","url":null,"abstract":"","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"29 ","pages":"Article 100330"},"PeriodicalIF":3.2,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147396677","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}