We develop a procedure for substantially improving the quality of segmented 3D micro-Computed Tomography (micro-CT) images of rocks with a Machine Learning (ML) Generative Model. The proposed model enhances the resolution eightfold (8x) and addresses segmentation inaccuracies due to the overlapping X-ray attenuation in micro-CT measurement for different rock minerals and phases. The proposed generative model is a 3D Deep Convolutional Wasserstein Generative Adversarial Network with Gradient Penalty (3D DC WGAN-GP). The algorithm is trained on segmented 3D low-resolution micro-CT images and segmented unpaired complementary 2D high-resolution Laser Scanning Microscope (LSM) images. The algorithm was demonstrated on multiple samples of Berea sandstones. We achieved high-quality super-resolved 3D images with a resolution of 0.44 μm/voxel and accurate segmentation for constituting minerals and pore space. The proposed procedure can significantly expand the modern capabilities of digital rock physics.
我们开发了一种程序，用于通过机器学习（ML）生成模型大幅提高岩石分段三维微计算机断层扫描（micro-CT）图像的质量。该模型将分辨率提高了8倍（8倍），并解决了在不同岩石矿物和相的微ct测量中由于x射线衰减重叠而导致的分割不准确。提出的生成模型是三维深度卷积Wasserstein梯度惩罚生成对抗网络（3D DC WGAN-GP）。该算法分别对三维低分辨率微ct图像和二维高分辨率激光扫描显微镜（LSM）图像进行分割训练。该算法在Berea砂岩的多个样本上进行了验证。我们获得了分辨率为0.44 μm/体素的高质量超分辨率3D图像，并对构成矿物和孔隙空间进行了精确分割。所提出的程序可以显著扩展数字岩石物理的现代能力。
{"title":"Super-resolution of 3D Micro-CT images using generative adversarial Networks: Enhancing resolution and segmentation accuracy","authors":"Evgeny Ugolkov , Xupeng He , Hyung Kwak , Hussein Hoteit","doi":"10.1016/j.cageo.2025.106018","DOIUrl":"10.1016/j.cageo.2025.106018","url":null,"abstract":"<div><div>We develop a procedure for substantially improving the quality of segmented 3D micro-Computed Tomography (micro-CT) images of rocks with a Machine Learning (ML) Generative Model. The proposed model enhances the resolution eightfold (8x) and addresses segmentation inaccuracies due to the overlapping X-ray attenuation in micro-CT measurement for different rock minerals and phases. The proposed generative model is a 3D Deep Convolutional Wasserstein Generative Adversarial Network with Gradient Penalty (3D DC WGAN-GP). The algorithm is trained on segmented 3D low-resolution micro-CT images and segmented unpaired complementary 2D high-resolution Laser Scanning Microscope (LSM) images. The algorithm was demonstrated on multiple samples of Berea sandstones. We achieved high-quality super-resolved 3D images with a resolution of 0.44 μm/voxel and accurate segmentation for constituting minerals and pore space. The proposed procedure can significantly expand the modern capabilities of digital rock physics.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"205 ","pages":"Article 106018"},"PeriodicalIF":4.4,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144763906","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pore size distribution (PSD) is vital for characterizing microscopic information and fluid transport in geomaterials, but traditional methods struggle with irregular pore shapes and digital imaging errors, often leading to inaccurate results. This study presents an improved morphological transformation-based algorithm that iteratively fills voids with maximal circles or spheres and introduces an optimized scheme for small-pore representation, significantly reducing measurement errors. Validation on eight 2D scanning electron microscope and six 3D computer tomography images shows the proposed method achieves up to 67 % lower relative error for small pore sizes and produces permeability predictions with a mean deviation within 3 % of experimental values, outperforming established techniques. Statistical analysis confirms that, for most samples, predicted permeability values fall within or approaching the 95 % confidence interval of measured data, demonstrating robust consistency across imaging sources and magnifications. Furthermore, the quantitative evaluation of pore geometry and PSD curves using different methods reveals that complex and randomly distributed pore geometries strongly influence PSD curve morphology, underscoring the importance of geometric characterization. These advancements enable more reliable and repeatable pore structure quantification, offering practical value for geoscience and engineering applications.
{"title":"An improved method for pore size distribution measurement of porous geomaterials based on microscopic images","authors":"Shijia Ma, Jiangfeng Liu, Zhipeng Wang, Ruinian Sun, Xinyue Zhang, Hongyang Ni","doi":"10.1016/j.cageo.2025.106017","DOIUrl":"10.1016/j.cageo.2025.106017","url":null,"abstract":"<div><div>Pore size distribution (PSD) is vital for characterizing microscopic information and fluid transport in geomaterials, but traditional methods struggle with irregular pore shapes and digital imaging errors, often leading to inaccurate results. This study presents an improved morphological transformation-based algorithm that iteratively fills voids with maximal circles or spheres and introduces an optimized scheme for small-pore representation, significantly reducing measurement errors. Validation on eight 2D scanning electron microscope and six 3D computer tomography images shows the proposed method achieves up to 67 % lower relative error for small pore sizes and produces permeability predictions with a mean deviation within 3 % of experimental values, outperforming established techniques. Statistical analysis confirms that, for most samples, predicted permeability values fall within or approaching the 95 % confidence interval of measured data, demonstrating robust consistency across imaging sources and magnifications. Furthermore, the quantitative evaluation of pore geometry and PSD curves using different methods reveals that complex and randomly distributed pore geometries strongly influence PSD curve morphology, underscoring the importance of geometric characterization. These advancements enable more reliable and repeatable pore structure quantification, offering practical value for geoscience and engineering applications.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"205 ","pages":"Article 106017"},"PeriodicalIF":4.4,"publicationDate":"2025-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144748729","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-16DOI: 10.1016/j.cageo.2025.106015
André V.S. Nascimento , Carlos A.M. Chaves , Susanne T.R. Maciel , George S. França , Giuliano S. Marotta
Seismology has made significant progress in high-resolution Earth imaging, largely driven by the increasing volume of freely available data. As a result, automated tools and machine learning algorithms are becoming essential for processing this vast amount of information. We present DisperPy, an open-source Python library developed to automatically extract group velocity dispersion curves from earthquake data. The analysis framework of DisperPy is structured around two primary tasks: (1) assessing the quality of waveforms to determine if dispersion extraction is feasible, and (2) measuring the group velocity dispersion curve for suitable waveforms. To address the first task, DisperPy uses a convolutional neural network trained on dispersion spectrograms to classify waveform quality. The model, based on the ResNet-34 architecture, is initialized with ImageNet-pretrained weights and fine-tuned using the fastai deep learning library. In the test set, the network achieves an accuracy of 92 % in distinguishing between high- and low-quality dispersion images. For the second task, DisperPy employs unsupervised learning techniques, starting with a Gaussian mixture model to separate dispersion energy from background noise, followed by k-means to separate the dispersion energy into clusters, making it easier to track amplitude maxima and then construct initial dispersion curves. Finally, a refinement of the initial dispersion is achieved using both the density-based spatial clustering of applications with noise algorithm and data quality criteria to remove possible outliers. To further test DisperPy, we conduct a surface wave tomography experiment across the contiguous United States using freely available vertical-component broadband waveforms. After processing the data with DisperPy and removing low-quality waveforms, the final dataset consisted of 194,325 unique dispersion curves. Consistent with previous studies, our maps reveal a prominent velocity dichotomy, with low velocities in the tectonically active western US and high velocities in the stable central and eastern US.
{"title":"DisperPy: A machine learning based tool to automatically pick group velocity dispersion curves from earthquakes","authors":"André V.S. Nascimento , Carlos A.M. Chaves , Susanne T.R. Maciel , George S. França , Giuliano S. Marotta","doi":"10.1016/j.cageo.2025.106015","DOIUrl":"10.1016/j.cageo.2025.106015","url":null,"abstract":"<div><div>Seismology has made significant progress in high-resolution Earth imaging, largely driven by the increasing volume of freely available data. As a result, automated tools and machine learning algorithms are becoming essential for processing this vast amount of information. We present <em>DisperPy</em>, an open-source Python library developed to automatically extract group velocity dispersion curves from earthquake data. The analysis framework of <em>DisperPy</em> is structured around two primary tasks: (1) assessing the quality of waveforms to determine if dispersion extraction is feasible, and (2) measuring the group velocity dispersion curve for suitable waveforms. To address the first task, <em>DisperPy</em> uses a convolutional neural network trained on dispersion spectrograms to classify waveform quality. The model, based on the ResNet-34 architecture, is initialized with ImageNet-pretrained weights and fine-tuned using the fastai deep learning library. In the test set, the network achieves an accuracy of 92 % in distinguishing between high- and low-quality dispersion images. For the second task, <em>DisperPy</em> employs unsupervised learning techniques, starting with a Gaussian mixture model to separate dispersion energy from background noise, followed by <em>k-means</em> to separate the dispersion energy into clusters, making it easier to track amplitude maxima and then construct initial dispersion curves. Finally, a refinement of the initial dispersion is achieved using both the density-based spatial clustering of applications with noise algorithm and data quality criteria to remove possible outliers. To further test <em>DisperPy</em>, we conduct a surface wave tomography experiment across the contiguous United States using freely available vertical-component broadband waveforms. After processing the data with <em>DisperPy</em> and removing low-quality waveforms, the final dataset consisted of 194,325 unique dispersion curves. Consistent with previous studies, our maps reveal a prominent velocity dichotomy, with low velocities in the tectonically active western US and high velocities in the stable central and eastern US.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"205 ","pages":"Article 106015"},"PeriodicalIF":4.2,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144662576","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-16DOI: 10.1016/j.cageo.2025.106014
Wenjin Chen, Xiaolong Tan
We introduce a novel focused gravity inversion algorithm and develop corresponding software, highlighting three key innovations. First, we propose the Adaptive Preconditioned Conjugate Gradient Regularization algorithm, which efficiently and adaptively determines the regularization parameter. Second, we incorporate the Sigmoid Function to stabilize the inversion process, significantly accelerating iterative convergence. Third, we have developed a user-friendly software with a graphical user interface for this new method, utilizing the popular high-level and interactive programming language MATLAB. To promote knowledge sharing and resource accessibility, we have made the software open-source. To validate our approach, we tested the algorithm on both synthetic and real gravity data, demonstrating its exceptional capability to accurately reconstruct the 3D density distribution of complex subsurface structures. Furthermore, we conducted a comparative analysis between the new algorithm, the conjugate gradient method constrained by SF, and the standard conjugate gradient method. The results indicate that the new method requires fewer iterations and exhibits higher computational efficiency.
{"title":"An Adaptive Preconditioned Conjugate Gradient Regularization (APCGR) algorithm with Sigmoid Function (SF) constraint for efficient three-dimensional (3D) gravity focusing inversion","authors":"Wenjin Chen, Xiaolong Tan","doi":"10.1016/j.cageo.2025.106014","DOIUrl":"10.1016/j.cageo.2025.106014","url":null,"abstract":"<div><div>We introduce a novel focused gravity inversion algorithm and develop corresponding software, highlighting three key innovations. First, we propose the Adaptive Preconditioned Conjugate Gradient Regularization algorithm, which efficiently and adaptively determines the regularization parameter. Second, we incorporate the Sigmoid Function to stabilize the inversion process, significantly accelerating iterative convergence. Third, we have developed a user-friendly software with a graphical user interface for this new method, utilizing the popular high-level and interactive programming language MATLAB. To promote knowledge sharing and resource accessibility, we have made the software open-source. To validate our approach, we tested the algorithm on both synthetic and real gravity data, demonstrating its exceptional capability to accurately reconstruct the 3D density distribution of complex subsurface structures. Furthermore, we conducted a comparative analysis between the new algorithm, the conjugate gradient method constrained by SF, and the standard conjugate gradient method. The results indicate that the new method requires fewer iterations and exhibits higher computational efficiency.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"205 ","pages":"Article 106014"},"PeriodicalIF":4.2,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144654450","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-14DOI: 10.1016/j.cageo.2025.105993
Jun Yi , ZhongLi Qi , XiangChengZhen Li , Fuqiang Lai , Wei Zhou
The mathematical model-based methods used for conventional oil and gas resources often perform poorly in fluid recognition of tight-sand reservoir, due to the mutual interference of various factors such as reservoir lithology and pore structure. Booming artificial intelligence technologies and accumulating logging data provide a solid foundation for the application of machine learning methods as new tools for fluid identification. However, there is often a serious class imbalance, which can easily lead to the inability to achieve ideal classification results, in the proportion of categories of the collected well logging data. Consequently, this issue has become a huge challenge for the academic and industrial communities. To address this, a novel class-imbalance learning framework for fluid recognition (CILF) is proposed to tight-sand gas reservoirs of Qingshimao-Gaoshawo area of Ordos Basin, in China. Specifically, an improved label propagation algorithm based on semi-supervised learning (SS-LPA) is designed at the data level, which can reduce the imbalance rate of raw data to some extent after assigning high-confidence labels to unlabeled samples. At the model level, -network, as an effective reinforcement learning approach, is introduced into ensemble learning framework (QNEL), which can enhance the multi-classification accuracy of fluid identification by training multiple baseline models that are given different weights for feedback on imbalanced data. The experimental results from 35 tight-sand wells in Qingshimao-Gaoshawo area of Ordos Basin validate the effectiveness of the proposed framework. Specifically, the performance of CILF is the best on all three typical evaluation metrics, and it outperforms others in 12 out of a total of 18 categories. In terms of the average scores for six categories, the precision, recall rate, and F1 score of the proposed framework reach 0.988, 0.984, and 0.985, respectively.
{"title":"A novel class-imbalance learning framework for fluid recognition: Application to Qingshimao-Gaoshawo tight-sand gas reservoirs in the Ordos Basin, China","authors":"Jun Yi , ZhongLi Qi , XiangChengZhen Li , Fuqiang Lai , Wei Zhou","doi":"10.1016/j.cageo.2025.105993","DOIUrl":"10.1016/j.cageo.2025.105993","url":null,"abstract":"<div><div>The mathematical model-based methods used for conventional oil and gas resources often perform poorly in fluid recognition of tight-sand reservoir, due to the mutual interference of various factors such as reservoir lithology and pore structure. Booming artificial intelligence technologies and accumulating logging data provide a solid foundation for the application of machine learning methods as new tools for fluid identification. However, there is often a serious class imbalance, which can easily lead to the inability to achieve ideal classification results, in the proportion of categories of the collected well logging data. Consequently, this issue has become a huge challenge for the academic and industrial communities. To address this, a novel class-imbalance learning framework for fluid recognition (CILF) is proposed to tight-sand gas reservoirs of Qingshimao-Gaoshawo area of Ordos Basin, in China. Specifically, an improved label propagation algorithm based on semi-supervised learning (SS-LPA) is designed at the data level, which can reduce the imbalance rate of raw data to some extent after assigning high-confidence labels to unlabeled samples. At the model level, <span><math><mi>Q</mi></math></span>-network, as an effective reinforcement learning approach, is introduced into ensemble learning framework (QNEL), which can enhance the multi-classification accuracy of fluid identification by training multiple baseline models that are given different weights for feedback on imbalanced data. The experimental results from 35 tight-sand wells in Qingshimao-Gaoshawo area of Ordos Basin validate the effectiveness of the proposed framework. Specifically, the performance of CILF is the best on all three typical evaluation metrics, and it outperforms others in 12 out of a total of 18 categories. In terms of the average scores for six categories, the precision, recall rate, and F1 score of the proposed framework reach 0.988, 0.984, and 0.985, respectively.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"205 ","pages":"Article 105993"},"PeriodicalIF":4.2,"publicationDate":"2025-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144654451","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-14DOI: 10.1016/j.cageo.2025.105999
Kai Gao, Ting Chen
Geological fault detection and characterization are crucial for understanding subsurface dynamics across scales. While methods for fault delineation based on either seismicity location analysis or seismic image reflector discontinuity are well-established, a systematic approach that integrates both data types remains absent. We develop a novel machine learning model that unifies seismic reflector images and seismicity location information to automatically identify geological faults and characterize their geometrical properties. The model encodes a seismic image and a seismicity location image separately, and fuses the encoded features with a spatial-channel attention fusion module to improve the learning of important features in both inputs. We design an automated strategy to generate high-quality synthetic training data and labels. To improve the realism of the seismicity location image, we include random seismicity noise and missing seismicity location associated with some of the faults. We validate the model’s efficacy and accuracy using synthetic data examples and two field data examples. Moreover, we show that fine-tuning the trained model with a small, domain-specific dataset enhances its fidelity for field data applications. The results demonstrate that integrating seismicity location and seismic images into a unified framework allows the end-to-end neural network to achieve higher fidelity and accuracy in delineating subsurface faults and their geometrical properties compared with image-only fault detection methods. Our approach offers an adaptive data-driven tool for geological fault characterization and seismic hazard mitigation, bridging the gap between seismicity location and image-based fault detection methods.
{"title":"Seismicity-constrained fault detection and characterization with a multitask machine learning model","authors":"Kai Gao, Ting Chen","doi":"10.1016/j.cageo.2025.105999","DOIUrl":"10.1016/j.cageo.2025.105999","url":null,"abstract":"<div><div>Geological fault detection and characterization are crucial for understanding subsurface dynamics across scales. While methods for fault delineation based on either seismicity location analysis or seismic image reflector discontinuity are well-established, a systematic approach that integrates both data types remains absent. We develop a novel machine learning model that unifies seismic reflector images and seismicity location information to automatically identify geological faults and characterize their geometrical properties. The model encodes a seismic image and a seismicity location image separately, and fuses the encoded features with a spatial-channel attention fusion module to improve the learning of important features in both inputs. We design an automated strategy to generate high-quality synthetic training data and labels. To improve the realism of the seismicity location image, we include random seismicity noise and missing seismicity location associated with some of the faults. We validate the model’s efficacy and accuracy using synthetic data examples and two field data examples. Moreover, we show that fine-tuning the trained model with a small, domain-specific dataset enhances its fidelity for field data applications. The results demonstrate that integrating seismicity location and seismic images into a unified framework allows the end-to-end neural network to achieve higher fidelity and accuracy in delineating subsurface faults and their geometrical properties compared with image-only fault detection methods. Our approach offers an adaptive data-driven tool for geological fault characterization and seismic hazard mitigation, bridging the gap between seismicity location and image-based fault detection methods.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"205 ","pages":"Article 105999"},"PeriodicalIF":4.2,"publicationDate":"2025-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144672224","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-11DOI: 10.1016/j.cageo.2025.105998
Victor Lamas, Alejandro Cortiñas, Miguel R. Luaces
Background:
The Internet of Things has revolutionized data collection in geosciences through extensive sensor networks. However, developing web-based data warehousing systems for IoT data remains costly and complex. While studies address sensor variability and data ingestion architectures, they often overlook the critical data warehouse component needed to manage IoT data volume and variability. Additionally, Model-Driven Engineering techniques have been used to create dashboards for urban activities but lack advanced map-based visualizations, which are essential for geospatial data.
Objectives:
This study aims to address the challenges of creating IoT data warehouses for geosciences, encouraging scientists to share sensor data analysis results using a simple, user-friendly, and cost-effective approach.
Methods:
The proposed framework integrates (i) a Domain-Specific Language metamodel to define sensors, dimensions, and measurement parameters, (ii) a Software Product Line for IoT data warehouse creation, and (iii) a low-code platform with command-line and web interfaces. The approach was validated through four case studies: meteorological, traffic and air quality, coastal, and oceanic monitoring systems.
Results:
The framework enables efficient IoT data warehouse creation with customized spatial, temporal, and attribute aggregation. Case studies demonstrate adaptability across domains, supporting real-time data ingestion, sensor mobility, and advanced visualization.
Conclusion:
The study presents a scalable, user-friendly framework for IoT data warehousing in geosciences using SPL and DSL technologies, addressing domain-specific challenges and empowering non-expert users. Future work includes usability assessments and expansion to other domains.
{"title":"Low-code framework for IoT data warehousing and visualization","authors":"Victor Lamas, Alejandro Cortiñas, Miguel R. Luaces","doi":"10.1016/j.cageo.2025.105998","DOIUrl":"10.1016/j.cageo.2025.105998","url":null,"abstract":"<div><h3>Background:</h3><div>The Internet of Things has revolutionized data collection in geosciences through extensive sensor networks. However, developing web-based data warehousing systems for IoT data remains costly and complex. While studies address sensor variability and data ingestion architectures, they often overlook the critical data warehouse component needed to manage IoT data volume and variability. Additionally, Model-Driven Engineering techniques have been used to create dashboards for urban activities but lack advanced map-based visualizations, which are essential for geospatial data.</div></div><div><h3>Objectives:</h3><div>This study aims to address the challenges of creating IoT data warehouses for geosciences, encouraging scientists to share sensor data analysis results using a simple, user-friendly, and cost-effective approach.</div></div><div><h3>Methods:</h3><div>The proposed framework integrates (i) a Domain-Specific Language metamodel to define sensors, dimensions, and measurement parameters, (ii) a Software Product Line for IoT data warehouse creation, and (iii) a low-code platform with command-line and web interfaces. The approach was validated through four case studies: meteorological, traffic and air quality, coastal, and oceanic monitoring systems.</div></div><div><h3>Results:</h3><div>The framework enables efficient IoT data warehouse creation with customized spatial, temporal, and attribute aggregation. Case studies demonstrate adaptability across domains, supporting real-time data ingestion, sensor mobility, and advanced visualization.</div></div><div><h3>Conclusion:</h3><div>The study presents a scalable, user-friendly framework for IoT data warehousing in geosciences using SPL and DSL technologies, addressing domain-specific challenges and empowering non-expert users. Future work includes usability assessments and expansion to other domains.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"205 ","pages":"Article 105998"},"PeriodicalIF":4.2,"publicationDate":"2025-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144632362","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-11DOI: 10.1016/j.cageo.2025.106013
Yijian Zeng , Fakhereh Alidoost , Bart Schilperoort , Yang Liu , Stefan Verhoeven , Meiert Willem Grootes , Yunfei Wang , Zengjing Song , Danyang Yu , Enting Tang , Qianqian Han , Lianyu Yu , Mostafa Gomaa Daoud , Prajwal Khanal , Yunfei Chen , Christiaan van der Tol , Raúl Zurita-Milla , Serkan Girgin , Bas Retsios , Niels Drost , Zhongbo Su
Droughts and heatwaves jeopardize terrestrial ecosystem services. The development of an open digital twin of the soil-plant system can help monitor and predict the impact of these extreme events on ecosystem functioning. We illustrate how our recently developed STEMMUS-SCOPE model—STEMMUS, Simultaneous Transfer of Energy, Mass and Momentum in Unsaturated Soil; SCOPE, Soil Canopy Observation of Photosynthesis and Energy fluxes—links soil-plant processes to novel satellite observables (e.g. solar-induced chlorophyll fluorescence), contributing to such a digital twin. This soil-plant digital twin allows a mechanistic window for tracking above- and below-ground ecophysiological processes with remote sensing observations. Following Open Science and FAIR (Findable, Accessible, Interoperable, Reusable) principles, both for data and research software, we present the building blocks of the soil-plant digital twin. It emphasizes the importance of FAIR-enabling digital technologies to translate research needs and developments into reproducible and reusable data, software and knowledge.
{"title":"Towards an open soil-plant digital twin based on STEMMUS-SCOPE model following open science","authors":"Yijian Zeng , Fakhereh Alidoost , Bart Schilperoort , Yang Liu , Stefan Verhoeven , Meiert Willem Grootes , Yunfei Wang , Zengjing Song , Danyang Yu , Enting Tang , Qianqian Han , Lianyu Yu , Mostafa Gomaa Daoud , Prajwal Khanal , Yunfei Chen , Christiaan van der Tol , Raúl Zurita-Milla , Serkan Girgin , Bas Retsios , Niels Drost , Zhongbo Su","doi":"10.1016/j.cageo.2025.106013","DOIUrl":"10.1016/j.cageo.2025.106013","url":null,"abstract":"<div><div>Droughts and heatwaves jeopardize terrestrial ecosystem services. The development of an open digital twin of the soil-plant system can help monitor and predict the impact of these extreme events on ecosystem functioning. We illustrate how our recently developed STEMMUS-SCOPE model—STEMMUS, Simultaneous Transfer of Energy, Mass and Momentum in Unsaturated Soil; SCOPE, Soil Canopy Observation of Photosynthesis and Energy fluxes—links soil-plant processes to novel satellite observables (e.g. solar-induced chlorophyll fluorescence), contributing to such a digital twin. This soil-plant digital twin allows a mechanistic window for tracking above- and below-ground ecophysiological processes with remote sensing observations. Following Open Science and FAIR (Findable, Accessible, Interoperable, Reusable) principles, both for data and research software, we present the building blocks of the soil-plant digital twin. It emphasizes the importance of FAIR-enabling digital technologies to translate research needs and developments into reproducible and reusable data, software and knowledge.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"205 ","pages":"Article 106013"},"PeriodicalIF":4.2,"publicationDate":"2025-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144686901","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-11DOI: 10.1016/j.cageo.2025.106012
Yuhan Ren, Hongming Zhang, Liang Dong, Huanyu Yang, Hongyi Li, Lu Du, Qiankun Chen, Songyuan Li
The accurate extraction of valley bottoms from digital elevation models (DEMs) is crucial for hydrological and geomorphological analyses of mountainous landscapes. However, threshold settings rely on manual intervention; roads near valley bottoms resemble valley-bottom features, and thresholds cannot effectively adapt to valleys of various shapes, leading to low extraction accuracy in existing methods, particularly in narrow V-shaped valleys. To address these issues, this study developed a semantic segmentation approach called a Dense-based Attention Merging Context Guided Network (DAM-CGNet). Without relying on thresholds, this method effectively excludes roads on hillslopes and enhances the recognition of steep feature changes at valley boundaries, enabling the extraction of valley bottoms of various shapes. Key improvements include: (1) incorporating the convolutional block attention module (CBAM) to enhance feature reuse in the information flow, employing attention mechanisms to suppress irrelevant feature responses and focus on valley boundary features; (2) using the dense connection strategy of DenseNet to rebuild the feature flow, helping the model keep important valley-bottom details in deep layers and better recognize small and narrow valleys; and (3) addressing the limitations of single-channel DEM representation by evaluating various input combinations, ultimately selecting DEM, topographic position index (TPI), and slope as effective inputs for valley-bottom extraction. Experiments using semantic segmentation models and conventional methods validated the effectiveness of the proposed method. Specifically, DAM-CGNet achieved high accuracy on the test set (MPA: 90.15 %, MIoU: 84.18 %, FWIoU: 92.99 %) and outperformed conventional methods in extracting valley bottoms of various shapes. This method, without a manual threshold setting as in conventional approaches, enhances valley bottom extraction precision and provides a new perspective for subsequent valley bottom width calculations.
{"title":"DAM-CGNet: Semantic segmentation-based approach for valley-bottom extraction from digital elevation models","authors":"Yuhan Ren, Hongming Zhang, Liang Dong, Huanyu Yang, Hongyi Li, Lu Du, Qiankun Chen, Songyuan Li","doi":"10.1016/j.cageo.2025.106012","DOIUrl":"10.1016/j.cageo.2025.106012","url":null,"abstract":"<div><div>The accurate extraction of valley bottoms from digital elevation models (DEMs) is crucial for hydrological and geomorphological analyses of mountainous landscapes. However, threshold settings rely on manual intervention; roads near valley bottoms resemble valley-bottom features, and thresholds cannot effectively adapt to valleys of various shapes, leading to low extraction accuracy in existing methods, particularly in narrow V-shaped valleys. To address these issues, this study developed a semantic segmentation approach called a Dense-based Attention Merging Context Guided Network (DAM-CGNet). Without relying on thresholds, this method effectively excludes roads on hillslopes and enhances the recognition of steep feature changes at valley boundaries, enabling the extraction of valley bottoms of various shapes. Key improvements include: (1) incorporating the convolutional block attention module (CBAM) to enhance feature reuse in the information flow, employing attention mechanisms to suppress irrelevant feature responses and focus on valley boundary features; (2) using the dense connection strategy of DenseNet to rebuild the feature flow, helping the model keep important valley-bottom details in deep layers and better recognize small and narrow valleys; and (3) addressing the limitations of single-channel DEM representation by evaluating various input combinations, ultimately selecting DEM, topographic position index (TPI), and slope as effective inputs for valley-bottom extraction. Experiments using semantic segmentation models and conventional methods validated the effectiveness of the proposed method. Specifically, DAM-CGNet achieved high accuracy on the test set (MPA: 90.15 %, MIoU: 84.18 %, FWIoU: 92.99 %) and outperformed conventional methods in extracting valley bottoms of various shapes. This method, without a manual threshold setting as in conventional approaches, enhances valley bottom extraction precision and provides a new perspective for subsequent valley bottom width calculations.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"205 ","pages":"Article 106012"},"PeriodicalIF":4.2,"publicationDate":"2025-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144672332","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Precipitation forecast is critical in flood management, agricultural planning, water resource allocation, and weather warnings. Despite significant advancements in Numerical Weather Prediction (NWP) models, these systems often exhibit substantial biases and errors, particularly at high spatial and temporal resolutions. To address these limitations, we develop and evaluate uncertainty-aware deep learning ensemble architectures, focusing on characterizing forecast uncertainties while achieving high accuracy and an optimal balance between sharpness and reliability. This study presents SDE U-Net, a novel adaptation of SDE-Net designed specifically for segmentation tasks in precipitation forecasting. We conduct a comprehensive evaluation of state-of-the-art ensemble architectures, including SDE U-Net, and compare their forecast uncertainty against that of a Poor Man’s Ensemble (PME, i.e. NWPs forecast average) across diverse meteorological conditions, ranging from non-intense precipitation patterns to intense weather events. As an example case, we focus on predicting daily cumulative precipitation in northwest Italy, though our approach is broadly generalizable. Our findings demonstrate that all the evaluated probabilistic deep learning models outperform the PME benchmark in terms of median RMSE for both non-intense and intense precipitation events. Among them, SDE U-Net achieves the best overall performance, delivering the lowest RMSE for intense events () and demonstrating a more stable error distribution compared to other models. For non-intense events, SDE U-Net perform comparably to other deep learning models, still notably surpassing the baselines. Moreover, SDE U-Net effectively balances sharpness and reliability, making it particularly suitable for operational forecasting of extreme weather. Integrating uncertainty-aware models like SDE U-Net into forecasting workflows can enhance decision-making and preparedness for weather-related hazards.
{"title":"Uncertainty-aware methods for enhancing rainfall prediction with deep-learning based post-processing segmentation","authors":"Simone Monaco , Luca Monaco , Daniele Apiletti , Roberto Cremonini , Secondo Barbero","doi":"10.1016/j.cageo.2025.105992","DOIUrl":"10.1016/j.cageo.2025.105992","url":null,"abstract":"<div><div>Precipitation forecast is critical in flood management, agricultural planning, water resource allocation, and weather warnings. Despite significant advancements in Numerical Weather Prediction (NWP) models, these systems often exhibit substantial biases and errors, particularly at high spatial and temporal resolutions. To address these limitations, we develop and evaluate uncertainty-aware deep learning ensemble architectures, focusing on characterizing forecast uncertainties while achieving high accuracy and an optimal balance between sharpness and reliability. This study presents SDE U-Net, a novel adaptation of SDE-Net designed specifically for segmentation tasks in precipitation forecasting. We conduct a comprehensive evaluation of state-of-the-art ensemble architectures, including SDE U-Net, and compare their forecast uncertainty against that of a Poor Man’s Ensemble (PME, i.e. NWPs forecast average) across diverse meteorological conditions, ranging from non-intense precipitation patterns to intense weather events. As an example case, we focus on predicting daily cumulative precipitation in northwest Italy, though our approach is broadly generalizable. Our findings demonstrate that all the evaluated probabilistic deep learning models outperform the PME benchmark in terms of median RMSE for both non-intense and intense precipitation events. Among them, SDE U-Net achieves the best overall performance, delivering the lowest RMSE for intense events (<span><math><mrow><mn>2</mn><mo>.</mo><mn>637</mn><mo>×</mo><mn>1</mn><msup><mrow><mn>0</mn></mrow><mrow><mo>−</mo><mn>2</mn></mrow></msup></mrow></math></span>) and demonstrating a more stable error distribution compared to other models. For non-intense events, SDE U-Net perform comparably to other deep learning models, still notably surpassing the baselines. Moreover, SDE U-Net effectively balances sharpness and reliability, making it particularly suitable for operational forecasting of extreme weather. Integrating uncertainty-aware models like SDE U-Net into forecasting workflows can enhance decision-making and preparedness for weather-related hazards.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"205 ","pages":"Article 105992"},"PeriodicalIF":4.2,"publicationDate":"2025-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144614778","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号