Scholarlink

<p>For urban underground projects such as Qingdao Metro Line 15, reasonably characterizing subsurface geological structures and concealed anomalies is important for construction safety and risk assessment. Conventional three-dimensional geological models mainly describe stratigraphic geometry, while their ability to represent shallow concealed anomalies is limited. To address this limitation, this study develops a multi-source data-driven three-dimensional geological attribute modeling workflow that integrates borehole data, GNSS-RTK measurements, and ground-penetrating radar (GPR) information. Based on standardized borehole data, an initial 3D geological model was constructed to represent the main stratigraphic framework of the study area. GPR data were then preprocessed to enhance anomaly-related reflection features. A U-Net model embedded with a Convolutional Block Attention Module (CBAM) was used to assist the segmentation of pipeline-type anomalies in GPR images, while cavities and loose zones were interpreted manually according to their radar reflection characteristics. The extracted anomaly information was spatially registered using GNSS-RTK data and subsequently integrated into the borehole-based 3D geological model to construct a 3D geological attribute model. Validation results show that the constructed model can generally reflect the main stratigraphic distribution characteristics of the study area. Among the 15 validation points, 10 showed validation consistency values above 70%, with five exceeding 80%, indicating that the model has potential for describing stratigraphic trends under the constraint of available borehole data. Compared with conventional borehole-only geological models, the proposed workflow enables approximate 3D positioning and visualization of concealed anomalies, such as pipelines and cavities, by integrating GPR-derived anomaly information. The results provide supplementary geological information for preliminary risk identification in shield tunneling and offer a feasible workflow for multi-source data fusion in refined subsurface geological characterization of urban underground space.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Construction and application of multi-source data-based 3D geological attribute models

  • 期刊论文

  • 发表日期：

    2026年6月19日

查看全文
  • Peng He,
  • Haoran Ding,
  • Kang Shi,
  • Gang Wang,
  • Lantian Wang,
  • Kun Leng

摘要

For urban underground projects such as Qingdao Metro Line 15, reasonably characterizing subsurface geological structures and concealed anomalies is important for construction safety and risk assessment. Conventional three-dimensional geological models mainly describe stratigraphic geometry, while their ability to represent shallow concealed anomalies is limited. To address this limitation, this study develops a multi-source data-driven three-dimensional geological attribute modeling workflow that integrates borehole data, GNSS-RTK measurements, and ground-penetrating radar (GPR) information. Based on standardized borehole data, an initial 3D geological model was constructed to represent the main stratigraphic framework of the study area. GPR data were then preprocessed to enhance anomaly-related reflection features. A U-Net model embedded with a Convolutional Block Attention Module (CBAM) was used to assist the segmentation of pipeline-type anomalies in GPR images, while cavities and loose zones were interpreted manually according to their radar reflection characteristics. The extracted anomaly information was spatially registered using GNSS-RTK data and subsequently integrated into the borehole-based 3D geological model to construct a 3D geological attribute model. Validation results show that the constructed model can generally reflect the main stratigraphic distribution characteristics of the study area. Among the 15 validation points, 10 showed validation consistency values above 70%, with five exceeding 80%, indicating that the model has potential for describing stratigraphic trends under the constraint of available borehole data. Compared with conventional borehole-only geological models, the proposed workflow enables approximate 3D positioning and visualization of concealed anomalies, such as pipelines and cavities, by integrating GPR-derived anomaly information. The results provide supplementary geological information for preliminary risk identification in shield tunneling and offer a feasible workflow for multi-source data fusion in refined subsurface geological characterization of urban underground space.