3D Geological Modeling based Multi-Source Data Fusion— A Case Study of Fuyang Karst Cave Area in Hangzhou

3D Geological Modeling based Multi-Source Data Fusion— A Case Study of Fuyang Karst Cave Area in Hangzhou | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article 3D Geological Modeling based Multi-Source Data Fusion— A Case Study of Fuyang Karst Cave Area in Hangzhou Fanfan Dou, Ting Lei, Huaixue Xing, Zunlong Du This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6524682/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 9 You are reading this latest preprint version Abstract Geological hazards such as uneven settlement of building foundations or roadbed collapses are prone to occur in karst regions. High-precision 3D geological modeling method can visually presents the spatial distribution and morphological characteristics of karst bodies. This study proposes a multi-data fusion methodology for 3D dynamic modeling of concealed karst features, centering on the integration of multi-source data. By comprehensively incorporating multi-dimensional information including surface remote sensing, subsurface geophysical surveys, geological sections, and boreholes, we constructed a high-precision 3D geological model of Fuyang karst cave area. The model effectively integrates diverse datasets such as borehole data, geophysical profiles, and planar geological maps, providing intuitive visualization of subsurface geological structures in karst areas, including caves, faults, and other features. This study offers scientific support for resource exploitation and hazard prevention in karst cave area. Implicit modeling 3D geological modeling Multi-source data integration Dynamic updating Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Introduction In 1993, Canadian scholar Houlding first proposed the concept of "3D geological modeling", defined as a technology that integrates geological data with computer techniques and mathematical methods to characterize geological bodies, structures, or specific geological features within a three-dimensional space. This approach combines spatial data management, geological interpretation, spatial exploration analysis, geostatistics, and graphical visualization to achieve three-dimensional analysis and visualization of geological entities (Houlding 1994). Current datasets involved in 3D geological modeling include regional topographic data, geological mapping data, borehole logging data, geological cross-section data, gravity-magnetic-electrical inversion data, and reflection seismic data, among others. 3D geological modeling enables the inference and interpretation of target geological bodies based on limited known geological and geophysical data. It facilitates qualitative and quantitative characterization of urban surface and subsurface spatial strata, geological structures, and attribute parameters within a 3D framework. This technology provides an intuitive representation of the spatial morphology, distribution characteristics, and interrelationships of subsurface geological bodies, offering decision-making support and technical services for urban planning, development, and management (Xiong et al. 2018 ; Pan et al.2020; He et al.2020; Zhou et al.2021). Prior to urban development, high-precision 3D geological models can be utilized to preliminarily analyze and evaluate geological structural characteristics under various development objectives. For instance, these models enable the examination of burial depths, spatial distributions, and relationships among critical subsurface geological layers (e.g., soft soil, hard soil, and gravel layers) within planned underground areas. They also facilitate statistical analysis of lithology and volumetric parameters of strata traversed by proposed metro tunnel shield routes. Furthermore, quantitative analysis of high-precision 3D geological models can advance research in three-dimensional quantitative comprehensive evaluations for Geological disaster prevention, underground space utilization and related fields (Makana et al.2016; Price et al.2018; Dou et al.2021, 2022). The construction of high-precision 3D geological models relies on appropriate modeling methodologies. Current 3D geological modeling methods can be broadly categorized into two types: explicit contour-based methods and implicit interpolation-based methods. Explicit contour-based 3D geological modeling methods dominated early 3D geological modeling practices. Notable commercial software such as Surpac™, Micromine™, 3Dmine™, and Datamine™ emerged during this period, alongside the development of various explicit modeling techniques tailored to diverse data sources. These methods can be further subdivided into two classes: (1) cross-section-based geological modeling methods and (2) borehole-based geological modeling methods (Qu et al.2008; Tang et al.2015; Bi et al.2016). However, both approaches exhibit limitations in 3D geological modeling. On the one hand, their capacity for integrating multi-source subsurface data remains constrained, making it challenging to simultaneously incorporate heterogeneous data from varying orientations, types, and spatial extents into the construction of models representing complex underground structures. On the other hand, when new borehole data or geological (interpreted) cross-sections are introduced post-modeling—even for localized structural updates—the explicit modeling process typically necessitates partial or complete regional remodeling. This reliance on repetitive human-computer interactions is time-intensive and laborious, significantly hindering the practical efficiency of geological modeling applications. In recent years, an implicit modeling approach that integrates observational data with geological knowledge has undergone significant advancements. Termed the implicit interface modeling method, this technique constructs geological interfaces through isosurfaces derived from an implicit function. Specifically, it leverages a 3D scalar field based on geochronological sequences, where geological and geophysical inferred boundaries and orientations serve as geometric constraints to generate a series of isosurfaces (via grid or tetrahedral mesh tracing) representing geological interfaces (Calcagno et al.2008; Hillier et al.2014). Over the past decade, implicit 3D geological modeling methods have matured, now capable of addressing dynamic data updates, integrating multi-source data, and constructing geologically plausible models even under sparse observational constraints. These advancements have garnered increasing attention from geoscientists (Vollgger et al.2013). Current implicit 3D modeling software incorporating geological knowledge includes Geomodeller, GIS3D, and Leapfrog. Among these, GeoModeller™, developed by Australia’s Intrepid Geophysics, stands out as the most robust and mature platform. This software employs an advanced Dual Cokriging algorithm for interpolating 3D stratigraphic surfaces, enabling rapid and efficient computations. During practical modeling, it effectively integrates diverse geological observations and interpreted data, such as borehole logs, digital elevation models (DEMs), geological maps, interpreted cross-sections, and remote sensing imagery. The methodology embedded in this software eliminates constraints on data orientation, scale, or spatial alignment, allowing seamless integration of multi-scale, heterogeneous geological information. Moreover, it supports model construction under limited observational data. GeoModeller has been extensively applied in underground space exploration and related disciplines (Collon et al.2015; Strati et al.2017; Scott et al.2019; Hassen et al.2021). The geological conditions in karst area are complex, with caves exhibiting diverse shapes and irregular distributions, posing numerous challenges to engineering construction, such as soil instability and karst collapse during subway shield tunneling. 3D geological modeling technology can visually and accurately represent the geological structures in karst areas, providing crucial references for engineering design and construction. Therefore, research on 3D geological modeling of karst regions holds significant practical importance. Xu et al. (2022) integrated multi-source data such as borehole data and high-density electrical method data, and accomplished 3D geological modeling for the complex karst area in the Nanning International Airport Project based on BIM technology. Shang et al. (2019) constructed a karst model for Jinan City by utilizing profile data and based on MapGIS 10 software. In summary, the application of 3D geological modeling in karst area is still in its nascent/initial stage. Moreover, given the relatively simple geomorphology of the area where most relevant projects have been conducted, there is limited application in complex karst areas and a lack of 3D geological modeling methodologies suitable for complex karst area based on multi-source geoscience data. Therefore, this study proposes a multi-data fusion approach centered on multi-source data integration for the purpose of 3D dynamic modeling of concealed karst features. By comprehensively integrating multi-dimensional information, including surface remote sensing data, underground geophysical survey results, geological profile data, and borehole records, a high-precision 3D geological model of the karst cave area in Fuyang has been constructed. This model visually presents the spatial distribution and morphological characteristics of karst bodies. Furthermore, an analysis has been conducted on the location, depth, scale and other attributes of karst features in the study area, along with an evaluation of karst collapse. Overview of the case study The study area is situated at the boundary between Fuyang district and Yuhang district, located northwest of Fuyang's urban center with a straight-line distance of approximately 16 km from Hang zhou (Fig. 1 ). The study area spans 1.6 km in east-west width and 1.4 km in north-south length, covering a total area of 2.24 km². Based on regional geological data and field investigations from this study, the stratigraphic sequence in the study area, ordered chronologically from oldest to youngest, is as follows: Nanhua System (Nh), Sinian System (Z), Cambrian System (Є), Ordovician System (O), Silurian System (S), and Quaternary System (Q). Multiple phases of intrusive rocks are also developed in this area. The exposed Quaternary deposits in the area are primarily classified by genetic types into slope-alluvial deposits and eluvial-colluvial deposits. Eluvial-colluvial deposits is widely distributed on the surface of natural slopes, consisting of silty clay with gravel, composed of clay, loam, sand, and gravel. Slope-alluvial deposits is mainly distributed in piedmont slope-alluvial lands with gentle slopes and low-lying terrain, which have been reclaimed as tea plantations and dry farmland. The lithology is silty clay with gravel-pebble, composed of clay, loam, and gravel. Granite porphyry dike (γπ) intrudes into the Upper Cambrian Xiyangshan Formation and Lower Ordovician Lunshan Formation in a northeast-trending orientation. The study area is characterized by structural erosion-denudation hilly landforms, with the overall ridge distribution trending northeastward. The terrain slopes downward from north to south, ranging in elevation from 121.4m to 425.7m, and exhibits well-developed mountain vegetation. Carbonate rocks are widely distributed in this area, and under the combined processes of water dissolution and erosion, various karst landforms have developed. Multiple factors influence carbonate dissolution processes in the area, including regional tectonic structures, stratigraphic lithology, and groundwater dynamics. The Sinian Bangqiaoshan Formation (Z₂b) and Cambrian Dachengling Formation (Є₁d) and Yangliugang Formation (Є₂y) constitute the primary carbonate strata that provide the material foundation for karst development. Specifically, the interbedded thick-layered limestone and dolomite in the Bangqiaoshan Formation offer high-permeability pathways for groundwater dissolution. Differential corrosion of carbonaceous limestone and dolomite in the Cambrian strata further shapes the complex internal structures of karst caves. Tectonic controls strictly determine the development direction of karst features: the orientation of karst valleys and the long axes of beaded karst depressions align with regional structural lineations. Karst cave development is also governed by tectonic fissures and bedding plane fractures. The study area contains six identified faults, predominantly trending northeast with subordinate NNE- and NW-trending structures, as interpreted from engineering surveys, field investigations, geophysical data, and borehole analyses. According to the types of groundwater occurrence, the groundwater in the study area can be divided into four categories: loose rock pore groundwater, carbonate rock interbedded with clastic rock fissure karst water, structural fissure water, and bedrock fissure water. Loose rock pore water is mainly found in the Quaternary residual slope silty clay, which is mainly distributed in slopes, valleys. The water bearing rock formations of carbonate karst water are mainly composed of Sinian and Cambrian limestone, argillaceous limestone, and dolomitic limestone. The structural fissure water is relatively developed and has good water abundance. Due to the development of faults in the area, most of the structural zones are distributed along the gullies and depressions on both sides of the landforms. Within the affected zone, cracks are developed, rock masses are broken, and the water conductivity is good. The fissure water in bedrock is mainly stored in strongly to weakly weathered bedrock. The strongly weathered fissures are developed, and the rock mass is fragmented with good water storage capacity and small thickness, mostly above the groundwater level. The weakly weathered joint fissures are developed, and the water volume is generally poor. Methodology Given the inherent challenges of integrating diverse and heterogeneous modeling data in karst area, a workflow for 3D dynamic modeling of concealed karst based on multi-source data fusion and Geomodeller™ was proposed (Fig. 2 ). The specific 11 steps are as follows: (1) Determine the geological structure within the study area and establish a unified sequence of modeling strata; (2) Collect borehole data, perform digitization and standardization, encode the data, and establish a borehole database; (3) Carry out preprocessing of borehole data to generate control points for missing geological strata at different borehole locations and control points for geological strata not reached by deep drilling based on the borehole data, addressing issues such as shallow borehole depth and missing local strata; (4) Import geological borehole data and control point data to achieve three-dimensional visualization of boreholes; (5) Collect topographic data, assign three-dimensional coordinate information to the topographic data, and construct a three-dimensional topographic model; (6) Perform spatial registration by combining planar data such as geological maps and bedrock geological maps, vectorize and redraw exposed strata and other structural information, extract planar constraint information, and form planar constraint information in conjunction with topographic data; (7) Perform spatial registration on profile data such as geological sections, vectorize and redraw exposed strata and other structural information, extract profile constraint information, and form profile constraint information in combination with depth information; (8) Number the constraint information belonging to different strata categories based on the division principles of the geological structure in the study area; (9) Utilize the "potential field method" implicit three-dimensional geological modeling method to establish a three-dimensional geological model (including the 3D geological model and a karst model) based on the division principles of the geological structure in the study area, and based on borehole, planar, and profile constraint interface information; (10) Conduct quantitative evaluation of the constructed three-dimensional geological model. Verify the rationality of the 3D geological model in the study area through overall graphical representation and sectioning; simultaneously, perform precise analysis on the geological model at known borehole locations to evaluate in detail the degree of consistency between the three-dimensional geological model and known borehole data, characterizing the accuracy and reliability of the model; (11) Repeat steps 8 to 11 to cyclically verify and revise the 3D geological model, accurately reflecting the geological characteristics, stratigraphic lithology combination, spatial variation, topography, and other content within the study area. 3D Geological Modeling 3.1 Data collection and standardization For the study area, the modeling process involved various data and materials from different sources and scales, primarily including borehole data, Digital Elevation Model (DEM), planar geological maps, geological cross-sections, and other multi-source and multi-dimensional data. Based on the above data and in combination with the implicit 3D modeling method summarized above (Fig. 2 ), the construction of the 3D geological structure model of karst in the study area was carried out using Geomodeller TM software. 3.2 Implicit 3D geological modeling (1) Determine the scope of the modeling work area and the modeling scale. The scope of the modeling work area in this study is consistent with that of the study area (Fig. 1 ). Due to the significant difference between the horizontal and vertical scales of the study area, in order to increase the vertical resolution of the model and avoid the impact of strong topographic mutations caused by large terrain elevation differences, the modeling scale was enlarged by 20 times in this study. (2) Define and divide the stratum modeling units into 18 units based on the actual geological background of the study area mentioned above (Table 1 ). On this basis, the characteristics such as the age, stages, and interrelationships of the strata in the area were further defined, and a modeling stratum age sequence was compiled. Table 1 Stratigraphic pile of the study area Series Reference Relationship Formation Karst cave_Series Bottom Erode Karst cave Q4_Series Bottom Erode Q4 rπ_Series Bottom Erode rπ Nh_Series Bottom Erode Nh2n Z1 _Series Bottom Erode Z1d1、Z1d2 Z2 _Series Bottom Erode Z2b1、Z2b2 Є1_Series Bottom Erode Є1h 、Є1d Є2_Series Bottom Erode E2y E3_Series Bottom Erode E3h、E3x O1_Series Bottom Erode O1l O3_Series Bottom Erode O3w2 S1_Series Bottom Erode S1x、S1h、S1k (3) A total of 11 valid boreholes were collected in this study. The study further sorted, classified, and standardized the borehole data layers based on the established standard stratigraphic sequence of the study area. The borehole data was imported and three-dimensionally plotted using modeling software (Fig. 4 ), and 3D stratum boundary points were extracted. (4) Based on the collected geological maps, and other maps covering the study area, the corresponding planar constraint information such as the boundaries and occurrences of geological units exposed in the study area, as well as the fault trace lines and their occurrences, were extracted and vectorized for redrawing. (5) Construct a 3D terrain model. By extracting elevation points from the collected CAD-format geological maps covering the study area, and performing operations such as data clipping, data mosaicking, format conversion, and longitudinal scale multiplication, the three-dimensional terrain model of the study area was ultimately constructed by importing the data into the modeling software. The topographic differences of geomorphic units such as plains and hills in the study area can be clearly observed in Fig. 3 . (6) Based on the collected data of four geological cross-sections, each cross-section was registered into three-dimensional space according to its endpoints (Fig. 4 ). The three-dimensionally plotted borehole information was projected onto adjacent cross-sections. After fully considering the geological age of the study area and combining the collected geological cross-section data, the vertical constraint information belonging to different categories was compiled according to the stratigraphic age sequence. Simultaneously, based on the collected multiple geophysical exploration interpretation results, each result was also registered into three-dimensional space according to the cross-section endpoints (Fig. 4 ). The 3D plotted borehole information was projected onto adjacent cross-sections. By combining the karst planar constraint information from geological maps, the vertical constraint information of multiple karst interpretation results in the geophysical exploration cross-sections was compiled and redrawn. (7) An implicit modeling method was employed to integrate extracted borehole, planar, and cross-sectional geological constraint information (Fig. 5 ), initializing the modeling parameters to construct the initial 3D geological model. (8) The 3D geological model was sectioned one by one according to the locations of the four cross-sections. Based on the three-dimensional geological model, the sectioned cross-sections were reconstructed, followed by detailed comparison and revision. The comparison and verification of all four geological cross-sections are shown in Fig. 6 . Steps 3 to 8) were repeated to incorporate the revision results into the model construction process. Simultaneously, a matching analysis was performed on the geological model corresponding to the known borehole locations. The consistency between the three-dimensional geological model and the known borehole data was evaluated in detail to characterize the accuracy and reliability of the model. Upon calculating the matching degree between all boreholes and the actual model, it was found to exceed 98.6%, indicating that the three-dimensional geological model established in this study made full use of the existing geological information and possessed high precision, with good performance and fitting ability in local areas. Furthermore, to accurately reflect the spatial structure and variations of the strata, as well as the topography and landforms within the area, additional geological cross-sections were added to constrain locations with fewer constraints. This step also aimed to observe the presence of missing strata or abnormal stratum thickness. On the basis of completing the aforementioned steps, a high-precision three-dimensional geological structure model of the study area was ultimately established, as shown in Fig. 7 . Model application 4.1 Quantitative analysis The calculation of the area and depth of karst caves holds significant importance for the prevention and control of geological disasters. Based on the established 3D geological model of the study area (Fig. 7 ), the area calculation function provided by the Geomodeller™ software was utilized to obtain the following measurement results for the karst cave model's depth and area: the depth of the karst cave ranges from 0 to 142 meters, with the predominant depth concentrated between 5–20 meters (Fig. 9 ). This depth requires key protection. Meanwhile, the karst model reveals that a total of 12 karst models were constructed and are distributed in moniliform patten (Fig. 8 )‌‌. The karst exhibits a buried depth ranging from 0 to 30 m and widths varying between 20 and 110 m.The total length of the ten geophysical profiles was 3300 meters, the total length of the karst caves was 1160 meters, and the linear density of the karst caves was 35.15%. Among the 11 boreholes, ‌ 6 boreholes drilled karst formations‌, yielding a ‌cavity encounter rate of 54.5%‌. Integrated quantitative analysis results indicate that the karst in the study area is relatively well-developed. 4.2 Spatial structure analysis To comprehensively observe the spatial distribution of strata throughout the entire study area, this study established four uniform barrier planes both horizontally and vertically within the software. By utilizing these barrier planes to slice through the model, a three-dimensional perspective cross-sectional view of the geological model in the study area was obtained, as shown in the Fig. 6 . The model enables a visual, three-dimensional, and intuitive representation of the spatial distribution patterns of strata, faults and karst caves. The mainly gained the following insights: The migration pathway of water is a necessary prerequisite for karst formation. Fault zones and their adjacent areas serve as favorable conduits for water flow. According to the three-dimensional visualization results (Fig. 10 ), fault structures are the main controlling factors for karst development in the study area. The areas traversed by five fault structures (eg. F1, F2, F4, F5, and F6) exhibit relatively developed karst features. In particular, fault intersections (eg. F4, F5, and F6) profoundly influence the direction of karst development in the study area, as well as the scale and size of its development. The formation of karst caves primarily depends on the presence of soluble rocks, such as limestone and dolomite. As can be seen from the 3D geological model (Fig. 7 ), karst is mainly hosted in the first member of the Banqiaoshan Formation of the Sinian System, and the second member of the Banqiaoshan Formation of the Sinian System. The lithology of the formations is mainly composed of calcium carbonate. These rocks can form caves, known as karst caves, under the long-term corrosion of water containing carbon dioxide. Additionally, the model facilitates the imaginative construction of a three-dimensional geological spatial concept for the entire survey area, aiding in geoscience education and providing geological evidence for the implementation of engineering projects. 4.3 3D evaluation of karst collapse Combining the above Quantitative and spatial structure analysis, this study selected five evaluation indicators: Fault intersection distance, Fault distance, Z 2 b 1 and Z 2 b 1 stratigraphic units distance, and terrain slope as key evaluation indicators affecting karst collapse in the study area. The Analytic Hierarchy Process (AHP) was employed to determine indicator weights, and the improved TOPSIS method (Dou et al. 2022 ) was adopted to establish a karst collapse susceptibility assessment model. Supported by 3D spatial analysis results of five evaluation indicators(Fig. 11 (a)~(e)), the karst collapse susceptibility in the study area was systematically evaluated, with zoning results categorized into four susceptibility levels (I-IV) based on evaluation scores (Fig. 11 (f)). The spatial distribution reveals that Zone IV (highest susceptibility) occupies 6.57% of the total study area, followed by Zone III (12.57%), Zone II (27.88%), and Zone I (lowest susceptibility, 52.98%). High-risk areas (Zones III and IV) exhibit limited spatial extent, primarily concentrated in identified karst development zones and adjacent regions. These areas are characterized by proximity to fault structures and their intersections, steep terrain gradients (> 15°), and distribution within limestone strata. The susceptibility zoning results of karst collapse provide critical data support for disaster prevention and mitigation planning. Specifically, groundwater extraction from karst aquifers should be strictly prohibited in Zones III and IV. Any engineering activities within these zones must be preceded by specialized karst geological surveys. Furthermore, the establishment of a regional karst water monitoring network is recommended to acquire dynamic hydrogeological data, enabling scientifically-grounded early warning systems for karst collapse events. Conclusion This study, based on the Geomodeller TM implicit dynamic 3D geological modeling software and combined with the proposed multi-data fusion method for 3D dynamic modeling of concealed karst, achieved the effective fusion of diverse data sources such as boreholes, geophysical profiles, and planar geological maps. Consequently, a detailed 3D model of karst and its geological settings in Fuyang was constructed. This work visually presents the underground geological structures in karst areas, including karst caves, faults, and other features, assisting relevant personnel in more accurately identifying potential collapse risk zones. It also provides a corresponding basis for subsequent monitoring and early warning of potential karst collapse risk areas, as well as for ensuring the development safety of related engineering constructions. Declarations Author contributions Conceptualization: D. F. F.; Methodology: X.H.X.; Formal analysis and investigation: L.T.; Writing - original draft preparation: D.Z.L.; Writing - review and editing: L.T.; Funding acquisition: D.F.F. and D.Z.L.; Resources: X.H.X. Funding This study was supported by the Natural Science Foundation of the Jiangsu Higher Education Institutions of China, grant number 23KJD170001 and The Project of Research on Karst Exploration and Risk Prevention in Typical Areas of Huzhou, grant number 2024ZJDZ023. Competing interests The authors declare no competing interests. References Bi Lin，Liu Xiaoming，Chen Xin (2016) An utomatic 3D modeling method based on orebody contours.Geomatics and Information Sciemce of Wuhan University，41（10）：1359-1365 Calcagno P., Chilès J.P., Courrioux G., Guillen A (2008) Geological modelling from field data and geological knowledge: Part I. Modelling method coupling 3D potential-field interpolation and geological rules. Physics of the Earth and Planetary Interiors, 171(1-4): 147-157. Collon P., Steckiewicz-Laurent W., Pellerin J., Laurent G., Caumon G., Reichart G (2015) 3D geomodelling combining implicit surfaces and Voronoi-based remeshing: A case study in the Lorraine Coal Basin (France). Computers & Geosciences, 77: 29-43. Dou F.F., Xing H.X., Li X.H., Y F., Lu Z.T., Li X.L (2022) 3D geological suitability evaluation for urban underground space development based on combined weighting and improved TOPSIS. Natural Resources Research, 31: 693–711. Hassen I., Fauchard C., Antoine R (2021) 3D geological modelling of a coastal area: case study of the Vaches Noires Cliffs, Normandy, France. Bulletin of Engineering Geology and the Environment, 80: 1375-1388. He H.H., He J., Xiao J.Z., Zhou Y.X., Liu Y., Li C (2020) 3D geological modeling and engineering properties of shallow superficial deposits: A case study in Beijing, China. Tunnelling and Underground Space Technology, 100: 103390. Hillier M.J., Schetselaar E.M., de Kemp E.A., Perron G (2014) Three-dimensional modelling of geological surfaces using generalized interpolation with radial basis functions. Mathematical Geosciences, 46(8): 931-953. Hou W.S., Yang L., Deng D.C., YL., Keith Clarke., YZ.J (2016) Assessing quality of urban underground spaces by coupling 3D geological models: The case study of Foshan city, South China. Computers & Geosciences, 89:1-11. Houlding S.W.(1994).3D Geoscience Modeling: Computer Techniques for Geological Characterization. Berlin: Springer-Verlag. Pan D.D., Xu Z.H., Lu X.M., Zhou L.Q., Li H.Y (2020) 3D scene and geological modeling using integrated multi-source spatial data: Methodology, challenges, and suggestions. Tunnelling and Underground Space Technology, 100: 103393. Price S.J., Terrington R.L., Busbyc J., Bricker S., Berry T (2018) 3D ground-use optimisation for sustainable urban development planning: a case-study from Earls Court, London, UK. Tunnelling and Underground Space Technology, 81: 144–164. Qu Honggang, Pan Mao,Ming Jing (2008) An Efficient Method for High- Precision 3D Geological Modeling from Intersected Folded Cross- Sections. Acta Scientiarum Naturalium Universitatis Pekinensis, (06):915-920. Scott S.W., Covell C., Júlíusson E., Valfells Á., Newson J., Hrafnkelsson B (2019) A probabilistic geologic model of the Krafla geothermal system constrained by gravimetric data. Geothermal Energy, 7(1): 1-30. Shang Hao，Song Xiaoshuai，Li Hu，Yan Shanshan (2019) Three-dimensional coupling model of multi-source and multi-scale data and its application in karst zone of Jinan spring region. Journal of Geology, 43(03):385-392. Strati V., Wipperfurth S.A., Baldoncini M., McDonough W.F., Mantovani F (2017) Perceiving the crust in 3‐D: A model integrating geological, geochemical, and geophysical data. Geochemistry, Geophysics, Geosystems, 18(12): 4326-4341. Tang Bingyin, Wu Chonglong, Li Xinchuan (2015) A fast progressive 3D geological modeling method based on borehole data. Rock and Soil Mechanics, 36(12):3633-3638. Vollgger S.A., Cruden A.R., Cowan J.E ( 2013) 3D implicit geological modelling of a gold deposit from a structural geologist's point of view. In 12th SGA Biennial Meeting—Mineral Deposit Research for a High-Tech World: 1-4 . Xiong Z.M., Guo J.T., Xia Y.P., L H., Wang M.Y., Shi S.S ( 2018) A 3D multi-scale geology modeling method for tunnel engineering risk assessment. Tunnelling and Underground Space Technology, 73: 71-81. Xu Tao, Mo Fan, Lin Xitao, Gao Zongli, Peng Bolun, Liang Huan, Meng Xianglin (2022) Construction Technology, 51(11):42-44+77. Zhou F., Li M., Huang C., Liang H., Liu Y.J., Zhang J.L (2022) Lithology-Based 3D Modeling of Urban Geological Attributes and Their Engineering Application: A Case Study of Guang’an City, SW China. Frontiers in Earth Science: 1070. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 18 Feb, 2026 Reviews received at journal 13 Jan, 2026 Reviewers agreed at journal 29 Dec, 2025 Reviews received at journal 05 Dec, 2025 Reviewers agreed at journal 30 Nov, 2025 Reviewers invited by journal 10 Aug, 2025 Editor assigned by journal 04 May, 2025 Submission checks completed at journal 25 Apr, 2025 First submitted to journal 24 Apr, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6524682","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":501175537,"identity":"218fefcc-d818-4c68-b56b-4a261bbbc82b","order_by":0,"name":"Fanfan Dou","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/UlEQVRIie3QsUoDQRCA4TkW1maStHMo6iOMBBKElTyIzUjgOttgkeJAWMu0+hZXCXYnC1uJtnZZG+vYCIKIm5TCnrGz2L/ej50ZgFzuP+buAwHsA6gWQC4IB0q50En8lCMZAmgBCA9mr7zSFW9NihdbGX7CQ+oSfa+K5zfLB2OtXoNoh0OHwDA3pylSeqWObywf3Vk9ZkGHI9drA/jqvE4QXi78bs9+Fs2yHpHQmvSFi9qliVc6Ep40fuedhONgl8i0DTlrPMZfpEJWv5DNLtePPI1kRtIaJBePLB27bC62mvFJHOy2/PiiyWDhXFjNTZIkkr89z+VyudyPvgH2jlk2FmXwbgAAAABJRU5ErkJggg==","orcid":"","institution":"Jiangsu Second Normal University","correspondingAuthor":true,"prefix":"","firstName":"Fanfan","middleName":"","lastName":"Dou","suffix":""},{"id":501175538,"identity":"1af16b6a-ad22-4409-bb92-86f10e6f7df4","order_by":1,"name":"Ting Lei","email":"","orcid":"","institution":"China Geological Survey","correspondingAuthor":false,"prefix":"","firstName":"Ting","middleName":"","lastName":"Lei","suffix":""},{"id":501175539,"identity":"9fd5920f-92f9-4f5d-b38b-1a3dd334644e","order_by":2,"name":"Huaixue Xing","email":"","orcid":"","institution":"China Geological Survey","correspondingAuthor":false,"prefix":"","firstName":"Huaixue","middleName":"","lastName":"Xing","suffix":""},{"id":501175540,"identity":"e8387e3a-df79-42f3-b5bf-62cc1225b9df","order_by":3,"name":"Zunlong Du","email":"","orcid":"","institution":"Zhejiang Provincial Natural Resources Group","correspondingAuthor":false,"prefix":"","firstName":"Zunlong","middleName":"","lastName":"Du","suffix":""}],"badges":[],"createdAt":"2025-04-25 03:08:06","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6524682/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6524682/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":89290335,"identity":"710b3beb-47b0-4e12-bc34-1d4181006ed5","added_by":"auto","created_at":"2025-08-18 12:13:24","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":229302,"visible":true,"origin":"","legend":"\u003cp\u003eGeographic map of the study area\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6524682/v1/d3c9eb038671d4cdd63fced2.png"},{"id":89290333,"identity":"d2305d7e-8b3d-486e-8abe-e3051467024a","added_by":"auto","created_at":"2025-08-18 12:13:24","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":100312,"visible":true,"origin":"","legend":"\u003cp\u003eThe workflow of 3D Geological Modeling based Multi-Source Data Fusion\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6524682/v1/2d0cd7ebbea57e8657a62e3b.png"},{"id":89290882,"identity":"6b84b6fb-697b-45e0-ad97-f1af836b0176","added_by":"auto","created_at":"2025-08-18 12:21:24","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":309259,"visible":true,"origin":"","legend":"\u003cp\u003e3D terrain model of the study area\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6524682/v1/abb058565e7eab424f19fa02.png"},{"id":89290336,"identity":"2b208c00-3d5b-4fcc-9d3f-eb221fb0442b","added_by":"auto","created_at":"2025-08-18 12:13:24","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":459866,"visible":true,"origin":"","legend":"\u003cp\u003e3D overlay display of section maps and boreholes\u003c/p\u003e\n\u003cp\u003e(a) geological section map and boreholes\u003c/p\u003e\n\u003cp\u003e(b) geophysical section map and boreholes\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6524682/v1/f4ec092a9772333404678ba8.png"},{"id":89291633,"identity":"0c877e33-d72d-4b46-b32c-e17f77d2ef8e","added_by":"auto","created_at":"2025-08-18 12:29:24","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":205400,"visible":true,"origin":"","legend":"\u003cp\u003eMulti-source and multi-dimensional geological constraint information\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6524682/v1/cf1ed4979786b212b6ff96b3.png"},{"id":89290342,"identity":"9d3bab53-6d6e-4d00-b9ae-a3f8fe81bf98","added_by":"auto","created_at":"2025-08-18 12:13:24","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":116379,"visible":true,"origin":"","legend":"\u003cp\u003eCross-sections generated from the 3D gelogical model\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-6524682/v1/8e5daba720c5d07f0490ff81.png"},{"id":89290345,"identity":"fcfa817b-65a0-41e4-ba6d-541fd3c6e3a4","added_by":"auto","created_at":"2025-08-18 12:13:24","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":209570,"visible":true,"origin":"","legend":"\u003cp\u003e3D gelogical model of the study area\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-6524682/v1/4f5dcd4b30ab146d76005ec6.png"},{"id":89290352,"identity":"8bec0162-2b86-4e4e-a834-7bbcc22bc0e9","added_by":"auto","created_at":"2025-08-18 12:13:24","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":195388,"visible":true,"origin":"","legend":"\u003cp\u003e3D karst cave model of the study area\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-6524682/v1/cb2088e4aa1654170f62c927.png"},{"id":89290357,"identity":"0be2c5b9-a380-4f1c-bd58-ff607f0226c8","added_by":"auto","created_at":"2025-08-18 12:13:24","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":88183,"visible":true,"origin":"","legend":"\u003cp\u003eThe depth statistic of 3D karst cave\u003c/p\u003e","description":"","filename":"9.png","url":"https://assets-eu.researchsquare.com/files/rs-6524682/v1/3b18ae6b9ee495c96ba27e64.png"},{"id":89290888,"identity":"2ffa9ef9-b6b3-456d-9113-ae0380c8b793","added_by":"auto","created_at":"2025-08-18 12:21:24","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":234951,"visible":true,"origin":"","legend":"\u003cp\u003e3D Fault Model of the study area\u003c/p\u003e","description":"","filename":"10.png","url":"https://assets-eu.researchsquare.com/files/rs-6524682/v1/8449d1c60c75bccd390a2714.png"},{"id":89290347,"identity":"83a1020a-e864-4f11-8e15-44bb8f56649b","added_by":"auto","created_at":"2025-08-18 12:13:24","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":492593,"visible":true,"origin":"","legend":"\u003cp\u003e3D evaluation results of the study area\u003c/p\u003e\n\u003cp\u003e(a) Fault intersection distance; (b ) Fault distance; (c ) Z\u003csub\u003e2\u003c/sub\u003eb\u003csup\u003e1\u003c/sup\u003e distance; (d ) Z\u003csub\u003e2\u003c/sub\u003eb\u003csup\u003e1\u003c/sup\u003e distance ;\u0026nbsp; (e ) Terrain slope; (f) 3D evaluation results\u003c/p\u003e","description":"","filename":"11.png","url":"https://assets-eu.researchsquare.com/files/rs-6524682/v1/c4b49181fce9da0429fd2211.png"},{"id":89292848,"identity":"f5de16fc-6981-4079-9c6b-c55419e175cc","added_by":"auto","created_at":"2025-08-18 12:45:30","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3183260,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6524682/v1/078b2488-417d-4ed8-a038-8fc116120a1c.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"3D Geological Modeling based Multi-Source Data Fusion— A Case Study of Fuyang Karst Cave Area in Hangzhou","fulltext":[{"header":"Introduction","content":"\u003cp\u003eIn 1993, Canadian scholar Houlding first proposed the concept of \"3D geological modeling\", defined as a technology that integrates geological data with computer techniques and mathematical methods to characterize geological bodies, structures, or specific geological features within a three-dimensional space. This approach combines spatial data management, geological interpretation, spatial exploration analysis, geostatistics, and graphical visualization to achieve three-dimensional analysis and visualization of geological entities (Houlding 1994). Current datasets involved in 3D geological modeling include regional topographic data, geological mapping data, borehole logging data, geological cross-section data, gravity-magnetic-electrical inversion data, and reflection seismic data, among others. 3D geological modeling enables the inference and interpretation of target geological bodies based on limited known geological and geophysical data. It facilitates qualitative and quantitative characterization of urban surface and subsurface spatial strata, geological structures, and attribute parameters within a 3D framework. This technology provides an intuitive representation of the spatial morphology, distribution characteristics, and interrelationships of subsurface geological bodies, offering decision-making support and technical services for urban planning, development, and management (Xiong et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Pan et al.2020; He et al.2020; Zhou et al.2021). Prior to urban development, high-precision 3D geological models can be utilized to preliminarily analyze and evaluate geological structural characteristics under various development objectives. For instance, these models enable the examination of burial depths, spatial distributions, and relationships among critical subsurface geological layers (e.g., soft soil, hard soil, and gravel layers) within planned underground areas. They also facilitate statistical analysis of lithology and volumetric parameters of strata traversed by proposed metro tunnel shield routes. Furthermore, quantitative analysis of high-precision 3D geological models can advance research in three-dimensional quantitative comprehensive evaluations for Geological disaster prevention, underground space utilization and related fields (Makana et al.2016; Price et al.2018; Dou et al.2021, 2022).\u003c/p\u003e\u003cp\u003eThe construction of high-precision 3D geological models relies on appropriate modeling methodologies. Current 3D geological modeling methods can be broadly categorized into two types: explicit contour-based methods and implicit interpolation-based methods. Explicit contour-based 3D geological modeling methods dominated early 3D geological modeling practices. Notable commercial software such as Surpac\u0026trade;, Micromine\u0026trade;, 3Dmine\u0026trade;, and Datamine\u0026trade; emerged during this period, alongside the development of various explicit modeling techniques tailored to diverse data sources. These methods can be further subdivided into two classes: (1) cross-section-based geological modeling methods and (2) borehole-based geological modeling methods (Qu et al.2008; Tang et al.2015; Bi et al.2016). However, both approaches exhibit limitations in 3D geological modeling. On the one hand, their capacity for integrating multi-source subsurface data remains constrained, making it challenging to simultaneously incorporate heterogeneous data from varying orientations, types, and spatial extents into the construction of models representing complex underground structures. On the other hand, when new borehole data or geological (interpreted) cross-sections are introduced post-modeling\u0026mdash;even for localized structural updates\u0026mdash;the explicit modeling process typically necessitates partial or complete regional remodeling. This reliance on repetitive human-computer interactions is time-intensive and laborious, significantly hindering the practical efficiency of geological modeling applications.\u003c/p\u003e\u003cp\u003eIn recent years, an implicit modeling approach that integrates observational data with geological knowledge has undergone significant advancements. Termed the implicit interface modeling method, this technique constructs geological interfaces through isosurfaces derived from an implicit function. Specifically, it leverages a 3D scalar field based on geochronological sequences, where geological and geophysical inferred boundaries and orientations serve as geometric constraints to generate a series of isosurfaces (via grid or tetrahedral mesh tracing) representing geological interfaces (Calcagno et al.2008; Hillier et al.2014). Over the past decade, implicit 3D geological modeling methods have matured, now capable of addressing dynamic data updates, integrating multi-source data, and constructing geologically plausible models even under sparse observational constraints. These advancements have garnered increasing attention from geoscientists (Vollgger et al.2013). Current implicit 3D modeling software incorporating geological knowledge includes Geomodeller, GIS3D, and Leapfrog. Among these, GeoModeller\u0026trade;, developed by Australia\u0026rsquo;s Intrepid Geophysics, stands out as the most robust and mature platform. This software employs an advanced Dual Cokriging algorithm for interpolating 3D stratigraphic surfaces, enabling rapid and efficient computations. During practical modeling, it effectively integrates diverse geological observations and interpreted data, such as borehole logs, digital elevation models (DEMs), geological maps, interpreted cross-sections, and remote sensing imagery. The methodology embedded in this software eliminates constraints on data orientation, scale, or spatial alignment, allowing seamless integration of multi-scale, heterogeneous geological information. Moreover, it supports model construction under limited observational data. GeoModeller has been extensively applied in underground space exploration and related disciplines (Collon et al.2015; Strati et al.2017; Scott et al.2019; Hassen et al.2021).\u003c/p\u003e\u003cp\u003eThe geological conditions in karst area are complex, with caves exhibiting diverse shapes and irregular distributions, posing numerous challenges to engineering construction, such as soil instability and karst collapse during subway shield tunneling. 3D geological modeling technology can visually and accurately represent the geological structures in karst areas, providing crucial references for engineering design and construction. Therefore, research on 3D geological modeling of karst regions holds significant practical importance. Xu et al. (2022) integrated multi-source data such as borehole data and high-density electrical method data, and accomplished 3D geological modeling for the complex karst area in the Nanning International Airport Project based on BIM technology. Shang et al. (2019) constructed a karst model for Jinan City by utilizing profile data and based on MapGIS 10 software. In summary, the application of 3D geological modeling in karst area is still in its nascent/initial stage. Moreover, given the relatively simple geomorphology of the area where most relevant projects have been conducted, there is limited application in complex karst areas and a lack of 3D geological modeling methodologies suitable for complex karst area based on multi-source geoscience data.\u003c/p\u003e\u003cp\u003eTherefore, this study proposes a multi-data fusion approach centered on multi-source data integration for the purpose of 3D dynamic modeling of concealed karst features. By comprehensively integrating multi-dimensional information, including surface remote sensing data, underground geophysical survey results, geological profile data, and borehole records, a high-precision 3D geological model of the karst cave area in Fuyang has been constructed. This model visually presents the spatial distribution and morphological characteristics of karst bodies. Furthermore, an analysis has been conducted on the location, depth, scale and other attributes of karst features in the study area, along with an evaluation of karst collapse.\u003c/p\u003e"},{"header":"Overview of the case study","content":"\u003cp\u003eThe study area is situated at the boundary between Fuyang district and Yuhang district, located northwest of Fuyang's urban center with a straight-line distance of approximately 16 km from Hang zhou (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e ). The study area spans 1.6 km in east-west width and 1.4 km in north-south length, covering a total area of 2.24 km².\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eBased on regional geological data and field investigations from this study, the stratigraphic sequence in the study area, ordered chronologically from oldest to youngest, is as follows: Nanhua System (Nh), Sinian System (Z), Cambrian System (Є), Ordovician System (O), Silurian System (S), and Quaternary System (Q). Multiple phases of intrusive rocks are also developed in this area. The exposed Quaternary deposits in the area are primarily classified by genetic types into slope-alluvial deposits and eluvial-colluvial deposits. Eluvial-colluvial deposits is widely distributed on the surface of natural slopes, consisting of silty clay with gravel, composed of clay, loam, sand, and gravel. Slope-alluvial deposits is mainly distributed in piedmont slope-alluvial lands with gentle slopes and low-lying terrain, which have been reclaimed as tea plantations and dry farmland. The lithology is silty clay with gravel-pebble, composed of clay, loam, and gravel. Granite porphyry dike (γπ) intrudes into the Upper Cambrian Xiyangshan Formation and Lower Ordovician Lunshan Formation in a northeast-trending orientation.\u003c/p\u003e\u003cp\u003eThe study area is characterized by structural erosion-denudation hilly landforms, with the overall ridge distribution trending northeastward. The terrain slopes downward from north to south, ranging in elevation from 121.4m to 425.7m, and exhibits well-developed mountain vegetation. Carbonate rocks are widely distributed in this area, and under the combined processes of water dissolution and erosion, various karst landforms have developed. Multiple factors influence carbonate dissolution processes in the area, including regional tectonic structures, stratigraphic lithology, and groundwater dynamics. The Sinian Bangqiaoshan Formation (Z₂b) and Cambrian Dachengling Formation (Є₁d) and Yangliugang Formation (Є₂y) constitute the primary carbonate strata that provide the material foundation for karst development. Specifically, the interbedded thick-layered limestone and dolomite in the Bangqiaoshan Formation offer high-permeability pathways for groundwater dissolution. Differential corrosion of carbonaceous limestone and dolomite in the Cambrian strata further shapes the complex internal structures of karst caves. Tectonic controls strictly determine the development direction of karst features: the orientation of karst valleys and the long axes of beaded karst depressions align with regional structural lineations. Karst cave development is also governed by tectonic fissures and bedding plane fractures. The study area contains six identified faults, predominantly trending northeast with subordinate NNE- and NW-trending structures, as interpreted from engineering surveys, field investigations, geophysical data, and borehole analyses.\u003c/p\u003e\u003cp\u003eAccording to the types of groundwater occurrence, the groundwater in the study area can be divided into four categories: loose rock pore groundwater, carbonate rock interbedded with clastic rock fissure karst water, structural fissure water, and bedrock fissure water. Loose rock pore water is mainly found in the Quaternary residual slope silty clay, which is mainly distributed in slopes, valleys. The water bearing rock formations of carbonate karst water are mainly composed of Sinian and Cambrian limestone, argillaceous limestone, and dolomitic limestone. The structural fissure water is relatively developed and has good water abundance. Due to the development of faults in the area, most of the structural zones are distributed along the gullies and depressions on both sides of the landforms. Within the affected zone, cracks are developed, rock masses are broken, and the water conductivity is good. The fissure water in bedrock is mainly stored in strongly to weakly weathered bedrock. The strongly weathered fissures are developed, and the rock mass is fragmented with good water storage capacity and small thickness, mostly above the groundwater level. The weakly weathered joint fissures are developed, and the water volume is generally poor.\u003c/p\u003e\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"Methodology","content":"\u003cp\u003eGiven the inherent challenges of integrating diverse and heterogeneous modeling data in karst area, a workflow for 3D dynamic modeling of concealed karst based on multi-source data fusion and Geomodeller™ was proposed (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e ). The specific 11 steps are as follows:\u003c/p\u003e\u003cp\u003e(1) Determine the geological structure within the study area and establish a unified sequence of modeling strata;\u003c/p\u003e\u003cp\u003e(2) Collect borehole data, perform digitization and standardization, encode the data, and establish a borehole database;\u003c/p\u003e\u003cp\u003e(3) Carry out preprocessing of borehole data to generate control points for missing geological strata at different borehole locations and control points for geological strata not reached by deep drilling based on the borehole data, addressing issues such as shallow borehole depth and missing local strata;\u003c/p\u003e\u003cp\u003e(4) Import geological borehole data and control point data to achieve three-dimensional visualization of boreholes;\u003c/p\u003e\u003cp\u003e(5) Collect topographic data, assign three-dimensional coordinate information to the topographic data, and construct a three-dimensional topographic model;\u003c/p\u003e\u003cp\u003e(6) Perform spatial registration by combining planar data such as geological maps and bedrock geological maps, vectorize and redraw exposed strata and other structural information, extract planar constraint information, and form planar constraint information in conjunction with topographic data;\u003c/p\u003e\u003cp\u003e(7) Perform spatial registration on profile data such as geological sections, vectorize and redraw exposed strata and other structural information, extract profile constraint information, and form profile constraint information in combination with depth information;\u003c/p\u003e\u003cp\u003e(8) Number the constraint information belonging to different strata categories based on the division principles of the geological structure in the study area;\u003c/p\u003e\u003cp\u003e(9) Utilize the \"potential field method\" implicit three-dimensional geological modeling method to establish a three-dimensional geological model (including the 3D geological model and a karst model) based on the division principles of the geological structure in the study area, and based on borehole, planar, and profile constraint interface information;\u003c/p\u003e\u003cp\u003e(10) Conduct quantitative evaluation of the constructed three-dimensional geological model. Verify the rationality of the 3D geological model in the study area through overall graphical representation and sectioning; simultaneously, perform precise analysis on the geological model at known borehole locations to evaluate in detail the degree of consistency between the three-dimensional geological model and known borehole data, characterizing the accuracy and reliability of the model;\u003c/p\u003e\u003cp\u003e(11) Repeat steps 8 to 11 to cyclically verify and revise the 3D geological model, accurately reflecting the geological characteristics, stratigraphic lithology combination, spatial variation, topography, and other content within the study area.\u003c/p\u003e"},{"header":"3D Geological Modeling","content":"\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n \u003ch2\u003e3.1 Data collection and standardization\u003c/h2\u003e\n \u003cp\u003eFor the study area, the modeling process involved various data and materials from different sources and scales, primarily including borehole data, Digital Elevation Model (DEM), planar geological maps, geological cross-sections, and other multi-source and multi-dimensional data. Based on the above data and in combination with the implicit 3D modeling method summarized above (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e), the construction of the 3D geological structure model of karst in the study area was carried out using Geomodeller \u003csup\u003eTM\u003c/sup\u003e software.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003e3.2 Implicit 3D geological modeling\u003c/h3\u003e\n\u003cp\u003e(1) Determine the scope of the modeling work area and the modeling scale. The scope of the modeling work area in this study is consistent with that of the study area (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). Due to the significant difference between the horizontal and vertical scales of the study area, in order to increase the vertical resolution of the model and avoid the impact of strong topographic mutations caused by large terrain elevation differences, the modeling scale was enlarged by 20 times in this study.\u003c/p\u003e\n\u003cp\u003e(2) Define and divide the stratum modeling units into 18 units based on the actual geological background of the study area mentioned above (Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). On this basis, the characteristics such as the age, stages, and interrelationships of the strata in the area were further defined, and a modeling stratum age sequence was compiled.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab1\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eStratigraphic pile of the study area\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSeries\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eReference\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eRelationship\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eFormation\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eKarst cave_Series\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBottom\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eErode\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eKarst cave\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eQ4_Series\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBottom\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eErode\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003er\u0026pi;_Series\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBottom\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eErode\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003er\u0026pi;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNh_Series\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBottom\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eErode\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNh2n\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eZ1 _Series\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBottom\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eErode\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eZ1d1、Z1d2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eZ2 _Series\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBottom\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eErode\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eZ2b1、Z2b2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eЄ1_Series\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBottom\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eErode\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eЄ1h 、Є1d\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eЄ2_Series\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBottom\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eErode\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eE2y\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eE3_Series\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBottom\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eErode\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eE3h、E3x\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eO1_Series\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBottom\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eErode\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eO1l\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eO3_Series\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBottom\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eErode\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eO3w2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eS1_Series\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBottom\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eErode\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eS1x、S1h、S1k\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e(3) A total of 11 valid boreholes were collected in this study. The study further sorted, classified, and standardized the borehole data layers based on the established standard stratigraphic sequence of the study area. The borehole data was imported and three-dimensionally plotted using modeling software (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e), and 3D stratum boundary points were extracted.\u003c/p\u003e\n\u003cp\u003e(4) Based on the collected geological maps, and other maps covering the study area, the corresponding planar constraint information such as the boundaries and occurrences of geological units exposed in the study area, as well as the fault trace lines and their occurrences, were extracted and vectorized for redrawing.\u003c/p\u003e\n\u003cp\u003e(5) Construct a 3D terrain model. By extracting elevation points from the collected CAD-format geological maps covering the study area, and performing operations such as data clipping, data mosaicking, format conversion, and longitudinal scale multiplication, the three-dimensional terrain model of the study area was ultimately constructed by importing the data into the modeling software. The topographic differences of geomorphic units such as plains and hills in the study area can be clearly observed in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e.\u003c/p\u003e\n\u003cp\u003e(6) Based on the collected data of four geological cross-sections, each cross-section was registered into three-dimensional space according to its endpoints (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e). The three-dimensionally plotted borehole information was projected onto adjacent cross-sections. After fully considering the geological age of the study area and combining the collected geological cross-section data, the vertical constraint information belonging to different categories was compiled according to the stratigraphic age sequence.\u003c/p\u003e\n\u003cp\u003eSimultaneously, based on the collected multiple geophysical exploration interpretation results, each result was also registered into three-dimensional space according to the cross-section endpoints (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e). The 3D plotted borehole information was projected onto adjacent cross-sections. By combining the karst planar constraint information from geological maps, the vertical constraint information of multiple karst interpretation results in the geophysical exploration cross-sections was compiled and redrawn.\u003c/p\u003e\n\u003cp\u003e(7) An implicit modeling method was employed to integrate extracted borehole, planar, and cross-sectional geological constraint information (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e), initializing the modeling parameters to construct the initial 3D geological model.\u003c/p\u003e\n\u003cp\u003e(8) The 3D geological model was sectioned one by one according to the locations of the four cross-sections. Based on the three-dimensional geological model, the sectioned cross-sections were reconstructed, followed by detailed comparison and revision. The comparison and verification of all four geological cross-sections are shown in Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e. Steps 3 to 8) were repeated to incorporate the revision results into the model construction process.\u003c/p\u003e\n\u003cp\u003eSimultaneously, a matching analysis was performed on the geological model corresponding to the known borehole locations. The consistency between the three-dimensional geological model and the known borehole data was evaluated in detail to characterize the accuracy and reliability of the model. Upon calculating the matching degree between all boreholes and the actual model, it was found to exceed 98.6%, indicating that the three-dimensional geological model established in this study made full use of the existing geological information and possessed high precision, with good performance and fitting ability in local areas.\u003c/p\u003e\n\u003cp\u003eFurthermore, to accurately reflect the spatial structure and variations of the strata, as well as the topography and landforms within the area, additional geological cross-sections were added to constrain locations with fewer constraints. This step also aimed to observe the presence of missing strata or abnormal stratum thickness. On the basis of completing the aforementioned steps, a high-precision three-dimensional geological structure model of the study area was ultimately established, as shown in Fig. \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e.\u003c/p\u003e"},{"header":"Model application","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003ch2\u003e4.1 Quantitative analysis\u003c/h2\u003e\n \u003cp\u003eThe calculation of the area and depth of karst caves holds significant importance for the prevention and control of geological disasters. Based on the established 3D geological model of the study area (Fig. \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e), the area calculation function provided by the Geomodeller\u0026trade; software was utilized to obtain the following measurement results for the karst cave model\u0026apos;s depth and area: the depth of the karst cave ranges from 0 to 142 meters, with the predominant depth concentrated between 5\u0026ndash;20 meters (Fig. \u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003e). This depth requires key protection. Meanwhile, the karst model reveals that a total of 12 karst models were constructed and are distributed in moniliform patten (Fig. \u003cspan class=\"InternalRef\"\u003e8\u003c/span\u003e)\u0026zwnj;\u0026zwnj;. The karst exhibits a buried depth ranging from 0 to 30 m and widths varying between 20 and 110 m.The total length of the ten geophysical profiles was 3300 meters, the total length of the karst caves was 1160 meters, and the linear density of the karst caves was 35.15%. Among the 11 boreholes, \u0026zwnj; 6 boreholes drilled karst formations\u0026zwnj;, yielding a \u0026zwnj;cavity encounter rate of 54.5%\u0026zwnj;. Integrated quantitative analysis results indicate that the karst in the study area is relatively well-developed.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003e4.2 Spatial structure analysis\u003c/h3\u003e\n\u003cp\u003eTo comprehensively observe the spatial distribution of strata throughout the entire study area, this study established four uniform barrier planes both horizontally and vertically within the software. By utilizing these barrier planes to slice through the model, a three-dimensional perspective cross-sectional view of the geological model in the study area was obtained, as shown in the Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e. The model enables a visual, three-dimensional, and intuitive representation of the spatial distribution patterns of strata, faults and karst caves. The mainly gained the following insights:\u003c/p\u003e\n\u003cp\u003eThe migration pathway of water is a necessary prerequisite for karst formation. Fault zones and their adjacent areas serve as favorable conduits for water flow. According to the three-dimensional visualization results (Fig. \u003cspan class=\"InternalRef\"\u003e10\u003c/span\u003e), fault structures are the main controlling factors for karst development in the study area. The areas traversed by five fault structures (eg. F1, F2, F4, F5, and F6) exhibit relatively developed karst features. In particular, fault intersections (eg. F4, F5, and F6) profoundly influence the direction of karst development in the study area, as well as the scale and size of its development.\u003c/p\u003e\n\u003cp\u003eThe formation of karst caves primarily depends on the presence of soluble rocks, such as limestone and dolomite. As can be seen from the 3D geological model (Fig. \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e), karst is mainly hosted in the first member of the Banqiaoshan Formation of the Sinian System, and the second member of the Banqiaoshan Formation of the Sinian System. The lithology of the formations is mainly composed of calcium carbonate. These rocks can form caves, known as karst caves, under the long-term corrosion of water containing carbon dioxide.\u003c/p\u003e\n\u003cp\u003eAdditionally, the model facilitates the imaginative construction of a three-dimensional geological spatial concept for the entire survey area, aiding in geoscience education and providing geological evidence for the implementation of engineering projects.\u003c/p\u003e\n\u003ch3\u003e4.3 3D evaluation of karst collapse\u003c/h3\u003e\n\u003cp\u003eCombining the above Quantitative and spatial structure analysis, this study selected five evaluation indicators: Fault intersection distance, Fault distance, Z\u003csub\u003e2\u003c/sub\u003eb\u003csup\u003e1\u003c/sup\u003e and Z\u003csub\u003e2\u003c/sub\u003eb\u003csup\u003e1\u003c/sup\u003e stratigraphic units distance, and terrain slope as key evaluation indicators affecting karst collapse in the study area. The Analytic Hierarchy Process (AHP) was employed to determine indicator weights, and the improved TOPSIS method (Dou et al. \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e) was adopted to establish a karst collapse susceptibility assessment model. Supported by 3D spatial analysis results of five evaluation indicators(Fig. \u003cspan class=\"InternalRef\"\u003e11\u003c/span\u003e(a)~(e)), the karst collapse susceptibility in the study area was systematically evaluated, with zoning results categorized into four susceptibility levels (I-IV) based on evaluation scores (Fig. \u003cspan class=\"InternalRef\"\u003e11\u003c/span\u003e (f)).\u003c/p\u003e\n\u003cp\u003eThe spatial distribution reveals that Zone IV (highest susceptibility) occupies 6.57% of the total study area, followed by Zone III (12.57%), Zone II (27.88%), and Zone I (lowest susceptibility, 52.98%). High-risk areas (Zones III and IV) exhibit limited spatial extent, primarily concentrated in identified karst development zones and adjacent regions. These areas are characterized by proximity to fault structures and their intersections, steep terrain gradients (\u0026gt;\u0026thinsp;15\u0026deg;), and distribution within limestone strata.\u003c/p\u003e\n\u003cp\u003eThe susceptibility zoning results of karst collapse provide critical data support for disaster prevention and mitigation planning. Specifically, groundwater extraction from karst aquifers should be strictly prohibited in Zones III and IV. Any engineering activities within these zones must be preceded by specialized karst geological surveys. Furthermore, the establishment of a regional karst water monitoring network is recommended to acquire dynamic hydrogeological data, enabling scientifically-grounded early warning systems for karst collapse events.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study, based on the Geomodeller\u003csup\u003eTM\u003c/sup\u003e implicit dynamic 3D geological modeling software and combined with the proposed multi-data fusion method for 3D dynamic modeling of concealed karst, achieved the effective fusion of diverse data sources such as boreholes, geophysical profiles, and planar geological maps. Consequently, a detailed 3D model of karst and its geological settings in Fuyang was constructed. This work visually presents the underground geological structures in karst areas, including karst caves, faults, and other features, assisting relevant personnel in more accurately identifying potential collapse risk zones. It also provides a corresponding basis for subsequent monitoring and early warning of potential karst collapse risk areas, as well as for ensuring the development safety of related engineering constructions.\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003eConceptualization: D. F. F.; Methodology: X.H.X.; Formal analysis and investigation: L.T.; Writing - original draft preparation: D.Z.L.; Writing - review and editing: L.T.; Funding acquisition: D.F.F. and D.Z.L.; Resources: X.H.X.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u0026nbsp; This study was supported by the Natural Science Foundation of the Jiangsu Higher Education Institutions of China, grant number 23KJD170001 and The Project of Research on Karst Exploration and Risk Prevention in Typical Areas of Huzhou, grant number 2024ZJDZ023.\u003c/p\u003e\n\u003cp\u003eCompeting interests \u0026nbsp;The authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBi Lin，Liu Xiaoming，Chen Xin (2016) An utomatic 3D modeling method based on orebody contours.Geomatics and Information Sciemce of Wuhan University，41（10）：1359-1365\u003c/li\u003e\n\u003cli\u003eCalcagno P., Chil\u0026egrave;s J.P., Courrioux G., Guillen A (2008) Geological modelling from field data and geological knowledge: Part I. Modelling method coupling 3D potential-field interpolation and geological rules. Physics of the Earth and Planetary Interiors, 171(1-4): 147-157.\u003c/li\u003e\n\u003cli\u003eCollon P., Steckiewicz-Laurent W., Pellerin J., Laurent G., Caumon G., Reichart G (2015) 3D geomodelling combining implicit surfaces and Voronoi-based remeshing: A case study in the Lorraine Coal Basin (France). Computers \u0026amp; Geosciences, 77: 29-43.\u003c/li\u003e\n\u003cli\u003eDou F.F., Xing H.X., Li X.H., Y F., Lu Z.T., Li X.L (2022) 3D geological suitability evaluation for urban underground space development based on combined weighting and improved TOPSIS. Natural Resources Research, 31: 693\u0026ndash;711.\u003c/li\u003e\n\u003cli\u003eHassen I., Fauchard C., Antoine R (2021) 3D geological modelling of a coastal area: case study of the Vaches Noires Cliffs, Normandy, France. Bulletin of Engineering Geology and the Environment, 80: 1375-1388.\u003c/li\u003e\n\u003cli\u003eHe H.H., He J., Xiao J.Z., Zhou Y.X., Liu Y., Li C (2020) 3D geological modeling and engineering properties of shallow superficial deposits: A case study in Beijing, China. Tunnelling and Underground Space Technology, 100: 103390.\u003c/li\u003e\n\u003cli\u003eHillier M.J., Schetselaar E.M., de Kemp E.A., Perron G (2014) Three-dimensional modelling of geological surfaces using generalized interpolation with radial basis functions. Mathematical Geosciences, 46(8): 931-953.\u003c/li\u003e\n\u003cli\u003eHou W.S., Yang L., Deng D.C., YL., Keith Clarke., YZ.J (2016) Assessing quality of urban underground spaces by coupling 3D geological models: The case study of Foshan city, South China. Computers \u0026amp; Geosciences, 89:1-11.\u003c/li\u003e\n\u003cli\u003eHoulding S.W.(1994).3D Geoscience Modeling: Computer Techniques for Geological Characterization. Berlin: Springer-Verlag.\u003c/li\u003e\n\u003cli\u003ePan D.D., Xu Z.H., Lu X.M., Zhou L.Q., Li H.Y (2020) 3D scene and geological modeling using integrated multi-source spatial data: Methodology, challenges, and suggestions. Tunnelling and Underground Space Technology, 100: 103393.\u003c/li\u003e\n\u003cli\u003ePrice S.J., Terrington R.L., Busbyc J., Bricker S., Berry T (2018) 3D ground-use optimisation for sustainable urban development planning: a case-study from Earls Court, London, UK. Tunnelling and Underground Space Technology, 81: 144\u0026ndash;164.\u003c/li\u003e\n\u003cli\u003eQu Honggang, Pan Mao,Ming Jing (2008) An Efficient Method for High- Precision 3D Geological Modeling from Intersected Folded Cross- Sections. Acta Scientiarum Naturalium Universitatis Pekinensis, (06):915-920.\u003c/li\u003e\n\u003cli\u003eScott S.W., Covell C., J\u0026uacute;l\u0026iacute;usson E., Valfells \u0026Aacute;., Newson J., Hrafnkelsson B (2019) A probabilistic geologic model of the Krafla geothermal system constrained by gravimetric data. Geothermal Energy, 7(1): 1-30.\u003c/li\u003e\n\u003cli\u003eShang Hao，Song Xiaoshuai，Li Hu，Yan Shanshan (2019) Three-dimensional coupling model of multi-source and multi-scale data and its application in karst zone of Jinan spring region. Journal of Geology, 43(03):385-392.\u003c/li\u003e\n\u003cli\u003eStrati V., Wipperfurth S.A., Baldoncini M., McDonough W.F., Mantovani F (2017) Perceiving the crust in 3‐D: A model integrating geological, geochemical, and geophysical data. Geochemistry, Geophysics, Geosystems, 18(12): 4326-4341.\u003c/li\u003e\n\u003cli\u003eTang Bingyin, Wu Chonglong, Li Xinchuan (2015) A fast progressive 3D geological modeling method based on borehole data. Rock and Soil Mechanics, 36(12):3633-3638.\u003c/li\u003e\n\u003cli\u003eVollgger S.A., Cruden A.R., Cowan J.E ( 2013) 3D implicit geological modelling of a gold deposit from a structural geologist\u0026apos;s point of view. In 12th SGA Biennial Meeting\u0026mdash;Mineral Deposit Research for a High-Tech World: 1-4 .\u003c/li\u003e\n\u003cli\u003eXiong Z.M., Guo J.T., Xia Y.P., L H., Wang M.Y., Shi S.S ( 2018) A 3D multi-scale geology modeling method for tunnel engineering risk assessment. Tunnelling and Underground Space Technology, 73: 71-81.\u003c/li\u003e\n\u003cli\u003eXu Tao, Mo Fan, Lin Xitao, Gao Zongli, Peng Bolun, Liang Huan, Meng Xianglin (2022) Construction Technology, 51(11):42-44+77.\u003c/li\u003e\n\u003cli\u003eZhou F., Li M., Huang C., Liang H., Liu Y.J., Zhang J.L (2022) Lithology-Based 3D Modeling of Urban Geological Attributes and Their Engineering Application: A Case Study of Guang\u0026rsquo;an City, SW China. Frontiers in Earth Science: 1070. \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"carbonates-and-evaporites","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"caev","sideBox":"Learn more about [Carbonates and Evaporites](http://link.springer.com/journal/13146)","snPcode":"13146","submissionUrl":"https://submission.nature.com/new-submission/13146/3","title":"Carbonates and Evaporites","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Implicit modeling, 3D geological modeling, Multi-source data integration, Dynamic updating","lastPublishedDoi":"10.21203/rs.3.rs-6524682/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6524682/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eGeological hazards such as uneven settlement of building foundations or roadbed collapses are prone to occur in karst regions. High-precision 3D geological modeling method can visually presents the spatial distribution and morphological characteristics of karst bodies. This study proposes a multi-data fusion methodology for 3D dynamic modeling of concealed karst features, centering on the integration of multi-source data. By comprehensively incorporating multi-dimensional information including surface remote sensing, subsurface geophysical surveys, geological sections, and boreholes, we constructed a high-precision 3D geological model of Fuyang karst cave area. The model effectively integrates diverse datasets such as borehole data, geophysical profiles, and planar geological maps, providing intuitive visualization of subsurface geological structures in karst areas, including caves, faults, and other features. This study offers scientific support for resource exploitation and hazard prevention in karst cave area.\u003c/p\u003e","manuscriptTitle":"3D Geological Modeling based Multi-Source Data Fusion— A Case Study of Fuyang Karst Cave Area in Hangzhou","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-18 12:13:19","doi":"10.21203/rs.3.rs-6524682/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-02-18T16:07:26+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-13T13:32:30+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"296394012660637671566039084717054755276","date":"2025-12-30T03:29:52+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-05T07:06:56+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"277051012634639916518969395019749685398","date":"2025-12-01T00:28:37+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-08-11T02:37:26+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-05-04T14:40:35+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-04-25T05:41:18+00:00","index":"","fulltext":""},{"type":"submitted","content":"Carbonates and Evaporites","date":"2025-04-25T02:54:15+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"carbonates-and-evaporites","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"caev","sideBox":"Learn more about [Carbonates and Evaporites](http://link.springer.com/journal/13146)","snPcode":"13146","submissionUrl":"https://submission.nature.com/new-submission/13146/3","title":"Carbonates and Evaporites","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"1dd3fe54-7144-45a8-ba8c-86a1e35b23d3","owner":[],"postedDate":"August 18th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-05-10T14:54:15+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-18 12:13:19","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6524682","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6524682","identity":"rs-6524682","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}