Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion

Abstract

This invention relates to the field of mineral exploration technology, specifically to a Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion. The method includes: constructing a multi-dimensional identification marker system encompassing geological, geophysical, and geochemical elements and integrating multi-source data; analyzing the structural framework to extract relevant information elements of deep ore-guiding and shallow ore-hosting structures, and establishing a structural coupling model reflecting the spatial configuration relationship between the two through coupling analysis; quantitatively evaluating the spatial correlation between geochemical anomaly information and the structural coupling model, and constructing a three-dimensional prediction model to delineate gold mineralization anomaly target areas; quantitatively assessing the uncertainty of the model, determining the sensitivity and confidence interval of key parameters, and iteratively optimizing the model to generate the final prediction results. This method can clearly identify effective metallogenic structural systems and quantify the reliability of prediction conclusions, thereby improving the accuracy and practicality of mineral exploration prediction.

Description

Technical Field

[0001] This invention relates to the field of mineral exploration technology, and in particular to a Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion. Background Technology

[0002] Prospecting for Carlin-type gold deposits has long relied on the comprehensive interpretation of geological, geophysical, and geochemical information. Existing methods typically involve overlaying or qualitatively analyzing these multi-source data to delineate anomalous areas. In terms of structural analysis, conventional techniques focus on the morphological description and classification of structural features such as faults and folds, or on using geophysical data to invert deep structural outlines. For 3D modeling, existing techniques mostly construct static solid models based on interpretation results to visualize the spatial relationship between geological bodies and mineralization.

[0003] These existing methods have limitations. Traditional structural analysis fails to effectively distinguish between structural elements of different origins and functions; that is, deep ore-guiding structures and shallow ore-hosting structures are often conflated or simply correlated. This results in the understanding of the spatial configuration relationship between fluid migration channels and sedimentation sites remaining at the conceptual level, unable to quantitatively characterize their coupling mechanisms. Furthermore, conventional 3D geological modeling results are deterministic models, lacking quantitative evaluation of their reliability. The selection of parameters upon which model construction relies is largely based on experience, and the extent to which changes in these parameters affect the prediction results is unclear, making the model's predictive ability uncertain and difficult to guide risk assessment in exploration decisions. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a Carlin-type gold deposit prospecting and prediction method based on the fusion of multi-source geological data.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion, comprising:

[0006] A multi-dimensional identification marker system for ore-guiding and ore-hosting structures is constructed, which covers identification elements in geological, geophysical, and geochemical dimensions.

[0007] Based on the aforementioned multi-dimensional identification system, multi-source geological data of the target area are collected and integrated to form an initial fused data volume;

[0008] The initial fused data volume is subjected to structural lattice analysis to extract deep structural information elements related to ore-guiding structures and shallow structural information elements related to ore-hosting structures. The deep structural information elements and the shallow structural information elements are subjected to coupling analysis to establish a structural coupling model that reflects the spatial configuration relationship between ore-guiding structures and ore-hosting structures.

[0009] Geochemical anomaly information is overlaid with the tectonic coupling model to identify gold mineralization anomaly target areas that have a high degree of agreement with the tectonic coupling model. The gold mineralization anomaly target areas are then characterized in three dimensions using three-dimensional geological modeling technology to generate a three-dimensional prediction model that includes tectonic ore-controlling elements and mineralization indication elements.

[0010] Uncertainty quantification assessment is performed on the three-dimensional prediction model to determine the sensitivity and confidence interval of the key parameters of the model. Based on the uncertainty quantification assessment results, the three-dimensional prediction model is iteratively optimized to form the final mineral exploration prediction results.

[0011] As a further aspect of the present invention, the construction of a multi-dimensional identification marker system for ore-guiding structures and ore-hosting structures includes:

[0012] The system collects regional geological maps, geophysical exploration data, geochemical sampling data, and geological data of known mineral deposits;

[0013] The regional detachment structures and associated faults of compressional anticlines are interpreted from the geological maps of the region as geological dimensional identification markers of ore-guiding structures.

[0014] Gravity anomaly gradient zones and magnetic anomaly linear zones reflecting deep structural characteristics are extracted from the geophysical exploration data as geophysical dimension identification markers for ore-guiding structures.

[0015] Specific elemental combinations anomalies related to deep fluid activity were screened from the geochemical sampling data as geochemical dimension identification markers for ore-guiding structures.

[0016] The spatial distribution characteristics of tectonic alteration forms, the morphology of anticline axes with specific geometric parameters, and the distribution range of dome structures are summarized from the geological data of the known ore deposits as geological dimension identification markers for ore-hosting structures.

[0017] The identification marks obtained from different dimensions will be systematically classified and encoded to form a structured, multi-dimensional identification mark system database.

[0018] As a further aspect of the present invention, the step of constructing a lattice analysis on the initial fused data volume includes:

[0019] Edge detection and texture analysis algorithms are applied to process geophysical data to identify potential linear structures and boundary information;

[0020] The geological map is digitized to construct structural elements and spatial topological relationships, clarifying the junction relationships and order of faults at different levels;

[0021] Geochemical data is gridded and the anisotropy of element spatial distribution is calculated to delineate element enrichment centers and migration trajectories.

[0022] By combining the linear structure and boundary information, the intersection relationship and order of the fractures, and the element enrichment centers and migration trajectories, a spatial geometric model and a causal relationship model reflecting the structural framework of the target area are constructed.

[0023] As a further aspect of the present invention, the extraction of deep structural information elements related to ore-guiding structures and shallow structural information elements related to ore-hosting structures includes:

[0024] In the spatial geometric model and genetic relationship model of the structural lattice, a depth threshold is set to distinguish between deep structural domains and shallow structural domains.

[0025] Within the deep tectonic domain, fault zones and ductile shear zones with regional extension scale and cutting through deep crustal layers are selected, and their geometric morphology, occurrence parameters, and activity period information are extracted as deep tectonic information elements.

[0026] Within the shallow structural domain, secondary fractures, densely fractured zones, and folded detachment areas that have a genetic connection or spatial intersection with the deep structural information elements are identified, and their spatial morphology, density, and combination relationship information are extracted as shallow structural information elements.

[0027] As a further aspect of the present invention, the coupling analysis of the deep structural information elements and the shallow structural information elements includes:

[0028] Calculate the angle between the spatial distribution direction of the deep structural information elements and the dominant development direction of the shallow structural information elements;

[0029] The analysis examines the relationship between the multi-phase activity history of fault zones in the deep structural information elements and the formation time sequence of structures in the shallow structural information elements.

[0030] Evaluate the configuration efficiency between the ore-forming fluid transport potential provided by the deep structural information elements and the ore-forming spatial storage capacity provided by the shallow structural information elements;

[0031] Based on the angle relationship, the formation timing matching relationship, and the configuration efficiency, a structural coupling model is established to quantitatively describe the spatial coupling strength between ore-guiding structures and ore-hosting structures.

[0032] As a further aspect of the present invention, the overlay analysis of geochemical anomaly information and the tectonic coupling model includes:

[0033] Trend surface analysis and residual anomaly calculation were performed on the geochemical sampling data to separate gold element anomalies and associated element anomalies related to local mineralization.

[0034] The spatial coordinates, intensity, and scale parameters of the concentration centers of gold and associated element anomalies are spatially superimposed with the spatial location and coupling intensity region of the constructed coupling model.

[0035] Calculate the spatial statistical correlation index between geochemical anomalies and tectonic coupling strength within each stacking unit;

[0036] Based on the preset correlation index threshold, gold mineralization anomaly target areas with high consistency with the structural coupling model are delineated, and the consistency score of each target area is recorded.

[0037] As a further aspect of the present invention, the three-dimensional characterization of the gold mineralization anomaly target area using three-dimensional geological modeling technology includes:

[0038] Based on borehole data, geophysical inversion results, and geological profiles, a three-dimensional geological structure framework model of the target area is constructed.

[0039] The spatial relationship between the ore-guiding structure and the ore-hosting structure in the structural coupling model is integrated into the three-dimensional geological structural framework model to form a three-dimensional geological model containing structural ore-controlling elements.

[0040] The three-dimensional spatial distribution data of geochemical anomalies and the distribution data of alteration mineral assemblages are used as attribute bodies to the three-dimensional geological model containing tectonic ore-controlling elements, thereby generating a three-dimensional prediction model that simultaneously includes tectonic ore-controlling elements and mineralization indication elements.

[0041] As a further aspect of the present invention, the uncertainty quantification evaluation of the three-dimensional prediction model includes:

[0042] Identify the data sources relied upon in the construction of the three-dimensional prediction model, including geological interpretation data, geophysical inversion data, and geochemical interpolation data;

[0043] Evaluate the accuracy, resolution, and coverage of each data source, and analyze the consistency of matching between data.

[0044] The Monte Carlo simulation method was used to perform multiple random perturbation simulations on the key interface morphology, structural spatial location, and mineralization boundary of the three-dimensional prediction model.

[0045] By statistically analyzing the variation range of key model parameters in multiple simulations, the sensitivity and confidence intervals of the key model parameters are determined.

[0046] As a further aspect of the present invention, the iterative optimization of the three-dimensional prediction model based on the uncertainty quantification assessment results includes:

[0047] Based on the confidence intervals of the key parameters of the model, adjust the control point weights and interpolation algorithm parameters of the geological interface during the three-dimensional geological modeling process;

[0048] For key parameters of the model that are highly sensitive, supplementary high-precision data should be collected or targeted geological verification should be carried out to constrain the direction of model optimization.

[0049] The 3D geological modeling process is rerun using the optimized parameters to generate an updated 3D prediction model;

[0050] The differences in the fit between the 3D prediction models before and after the update at known mineral deposits are compared until the model changes tend to stabilize and conform to geological laws, thus forming the final mineral exploration prediction results.

[0051] As a further aspect of the present invention, after forming the final mineral exploration prediction result, it further includes:

[0052] The final mineral exploration prediction results will be dynamically linked with the deployment of a new round of geological exploration projects;

[0053] Based on the actual mineralization information reported by the exploration project, the multi-dimensional identification marker system, the structural coupling model, and the three-dimensional modeling parameters are calibrated.

[0054] By using calibrated parameters and models, the prediction results for unverified areas are continuously updated to achieve a sustained improvement in mineral exploration prediction capabilities.

[0055] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0056] By implementing structural framework analysis, specific information elements of deep ore-guiding structures and shallow ore-hosting structures were extracted, and coupling analysis was conducted on these two types of elements to establish a structural coupling model reflecting their spatial configuration relationship. This technical approach advances structural research from morphology to functional systems, clarifying the complete transport network and ore-controlling framework of ore-forming fluids from the source region to the sedimentation site. Based on this model, the geological basis for target areas selected has been deepened from "existence of structures" to "existence of an effectively coupled structural system," improving the accuracy of identifying truly mineralized potential spaces from complex structural backgrounds.

[0057] The generated 3D prediction model undergoes uncertainty quantification assessment to determine the sensitivity and confidence intervals of its key parameters, and the model is iteratively optimized based on the assessment results. This process transforms the static geological model into a dynamic and evaluable prediction system. Sensitivity analysis reveals which parameters have the greatest impact on the prediction results, thus guiding the optimization direction of data acquisition and interpretation; the confidence interval provides the credible range of the prediction conclusions. Through iterative feedback, the model can continuously integrate new insights and data, and its prediction results are no longer single conclusions, but decision support information including reliability measures and optimization paths, enhancing the practical value and risk controllability of the prediction results in actual exploration applications. Attached Figure Description

[0058] Figure 1 This is a flowchart of the Carlin-type gold prospecting prediction method based on multi-source geological data fusion as described in this invention;

[0059] Figure 2 A flowchart for constructing a multi-dimensional identification system;

[0060] Figure 3 A flowchart for extracting structural information elements from deep and shallow structures;

[0061] Figure 4 A bar chart showing the perturbation parameters of the Monte Carlo simulation for a three-dimensional prediction model of Carlin-type gold deposits;

[0062] Figure 5 A comparative curve of structural intensity versus depth in Carlin-type gold prospecting prediction. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0064] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0065] See Figure 1This invention provides a Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion. The specific implementation steps are as follows: A multi-dimensional identification marker system for ore-guiding and ore-hosting structures is constructed. This system covers identification elements from geological, geophysical, and geochemical dimensions. Based on the multi-dimensional identification marker system, multi-source geological data of the target area is collected and integrated to form an initial fused data volume. Structural framework analysis is performed on the initial fused data volume to extract deep structural information elements related to ore-guiding structures and shallow structural information elements related to ore-hosting structures. The deep and shallow structural information elements are then analyzed... Coupling analysis was conducted to establish a structural coupling model reflecting the spatial configuration relationship between ore-guiding and ore-hosting structures. Geochemical anomaly information was overlaid with the structural coupling model to identify gold mineralization anomaly target areas with high consistency with the structural coupling model. Three-dimensional geological modeling technology was used to perform three-dimensional characterization of the gold mineralization anomaly target areas, generating a three-dimensional prediction model containing structural ore-controlling elements and mineralization indication elements. Uncertainty quantification assessment of the three-dimensional prediction model was conducted to determine the sensitivity and confidence interval of key model parameters. Based on the uncertainty quantification assessment results, the three-dimensional prediction model was iteratively optimized to form the final mineral exploration prediction results.

[0066] See Figure 2 In one embodiment of the present invention, the system collects regional geological maps, geophysical exploration data, geochemical sampling data, and geological data of known mineral deposits. Regional detachment structures and associated faults of compressional anticlines are interpreted from the regional geological maps as geological dimension identification markers for ore-guiding structures. Gravity anomaly gradient zones and magnetic anomaly linear zones reflecting deep structural characteristics are extracted from the geophysical exploration data as geophysical dimension identification markers for ore-guiding structures. Specific elemental combination anomalies related to deep fluid activity are screened from the geochemical sampling data as geochemical dimension identification markers for ore-guiding structures. The spatial distribution characteristics of tectonic alteration forms, the morphology of anticline axes with specific geometric parameters, and the distribution range of dome structures are summarized from the geological data of known mineral deposits as geological dimension identification markers for ore-hosting structures. The identification markers obtained from different dimensions are systematically classified and coded to form a structured multi-dimensional identification marker system database. Edge detection and texture analysis algorithms are applied to process geophysical data to identify potential linear structures and boundary information. Geological maps are digitized to construct structural elements and spatial topological relationships, clarifying the junction relationships and order of faults at different levels. Geochemical data are gridded and the anisotropy of element spatial distribution is calculated to delineate element enrichment centers and migration trajectories. By integrating linear structures and boundary information, fault junction relationships and order, and element enrichment centers and migration trajectories, a spatial geometric model and genetic relationship model reflecting the structural framework of the target area are constructed.

[0067] In the specific implementation, a Carlin-type gold deposit cluster in Guizhou Province was used as an example area. A multi-dimensional data collection was systematically undertaken, including a 1:50,000 regional geological map, 1:200,000 regional gravity and aeromagnetic exploration data, 1:50,000 soil geochemical sampling data, and exploration reports and comprehensive research results from multiple known gold deposits. From the 1:50,000 regional geological map, a northwest-trending regional detachment tectonic zone and a series of fault systems associated with northeast-trending compressional anticlines were interpreted. These structures served as geological dimension identification markers for ore-guiding structures. From the 1:200,000 Bouguer gravity anomaly data, a northwest-trending gravity anomaly gradient zone running longitudinally through the study area was extracted. Its horizontal gradient value ranged from 3.5 to 5.0 milligal per kilometer. Corresponding linear magnetic anomaly zones were identified from the aeromagnetic ΔT polarization anomaly data. These features served as geophysical dimension identification markers for ore-guiding structures. From 1:50,000 soil geochemical sampling data, factor analysis was used to screen for combined anomalies of arsenic, antimony, and mercury. The spatial distribution of these combined anomalies is correlated with deep fluid activity and was used as a geochemical dimension identification marker for ore-guiding structures. From exploration reports of known gold deposits, the banded and lenticular spatial distribution characteristics of silicification, pyritization, and clay alteration structures were summarized. Curvature radii and dip angles of anticline axes were measured from geological profiles of the deposits, and the area distribution of dome structures on the plane was statistically analyzed. These characteristics were used as geological dimension identification markers for ore-hosting structures. The identification markers obtained from geological, geophysical, and geochemical dimensions will be systematically classified and coded according to structural type, spatial coordinates, geometric parameters, and geophysical and geochemical properties, and stored in a relational database to form a structured multi-dimensional identification marker system database.

[0068] In some embodiments, the Canny edge detection algorithm and the gray-level co-occurrence matrix texture analysis algorithm are applied to process the Bouguer gravity anomaly data and aeromagnetic ΔT polarization anomaly data of the above-mentioned areas to identify potential linear tectonic traces and boundary information of different geological bodies. For example, linear boundaries with a length exceeding 10 kilometers and a strike of 320° NW are identified on the gravity vertical second derivative map. Tectonic elements of the 1:50,000 regional geological map are digitized, transforming elements such as fault lines and fold axis traces into vector layers with topological relationships, clarifying the intersection relationship between main faults and secondary faults, such as clarifying the order of NW-trending faults cutting NE-trending faults. The gold, arsenic, and antimony element contents in the soil geochemical data are interpolated using Kriging grids to generate a raster map of elemental spatial distribution, and the anisotropy of each element's distribution is calculated. This calculation is achieved by comparing variograms in different directions, using the following formula:

[0069]

[0070] in: Indicates the anisotropy ratio index. This represents the variogram value along the principal axis. This represents the variogram value perpendicular to the principal axis. Based on this calculation, the enrichment center of gold and the migration trajectory of arsenic along a specific direction are delineated. By integrating linear structural and boundary information identified from geophysical data, fault junction relationships and sequences constructed from geological maps, and element enrichment centers and migration trajectories delineated from geochemical data, spatial overlay and correlation analysis are performed on a GIS platform to construct a target area structural framework model that both expresses the spatial geometry of the structure and reveals the genetic relationship between the structure and element migration.

[0071] See Figure 3 In one embodiment of the present invention, in the spatial geometric model and genetic relationship model of the structural framework, a depth threshold is set to distinguish between deep structural domains and shallow structural domains. Within the deep structural domain, fault zones and ductile shear zones with regional extension scale and cutting deep crustal layers are screened, and their geometric morphology, occurrence parameters, and activity period information are extracted as deep structural information elements. Within the shallow structural domain, secondary faults, fracture-dense zones, and folded detachment sites that have genetic relationships or spatial intersections with the deep structural information elements are identified, and their spatial morphology, density, and combination relationship information are extracted as shallow structural information elements. The angle between the spatial distribution direction of deep structural information elements and the dominant development direction of shallow structural information elements is calculated. The multi-phase activity history of fault zones in deep structural information elements and the formation time sequence matching relationship of structures in shallow structural information elements are analyzed. The configuration efficiency between the ore-forming fluid transport potential provided by deep structural information elements and the ore-forming spatial storage capacity provided by shallow structural information elements is evaluated. Based on the angle relationship, formation time sequence matching relationship and configuration efficiency, a structural coupling model is established to quantitatively describe the spatial coupling strength between ore-guiding structures and ore-hosting structures.

[0072] In specific implementation, based on the spatial geometric model and genetic relationship model of the structural framework constructed in the embodiments, a depth threshold of 5 km underground is set as the depth threshold to distinguish between deep and shallow structural domains. Within the deep structural domain, two NW-trending regional fault zones with an extension length greater than 50 km and an inferred cutting depth of more than 10 km are selected. The geometric morphology of these two fault zones is extracted as a gently wavy spatial trajectory, an average dip of 60° NE, a dip angle between 50° and 70°, and information on the occurrence of at least two periods of activity, the Indosinian and Yanshanian periods. This information is extracted as deep structural information elements. Within the shallow tectonic domain, a group of northeast-trending secondary faults that intersect the aforementioned northwest-trending regional fault zone in a spatial "in" shape, a dense zone of extensional fractures located in the core of the anticline, and a folded detachment region of a short-axis anticline were identified. The displacement and length data of these secondary faults, the spatial distribution of fracture surface density within the dense fracture zone, and the information on the combined intersecting relationships between different tectonic elements were extracted. This information was extracted as shallow tectonic information elements.

[0073] In some embodiments, the average spatial distribution direction of northwest-trending regional fault zones in deep structural information elements is calculated as 310°, and the angle between this angle and the dominant development direction of northeast-trending secondary faults in shallow structural information elements is 55°, resulting in an angle of 75°. The Indosinian and Yanshanian activity histories of fault zones in deep structural information elements are analyzed and matched with the Yanshanian formation sequence of northeast-trending secondary faults in shallow structural information elements, confirming a temporal match between the two during the Yanshanian period. The ore-forming fluid transport potential provided by deep structural information elements is evaluated using the product of fault zone width and extension length as the potential index. The ore-forming spatial storage capacity provided by shallow structural information elements is calculated using the product of fracture density and favorable structural geometric volume as the capacity index. The configuration efficiency between the potential index and the capacity index is calculated through a ratio relationship.

[0074] It is understandable that, based on the calculated 75° angle relationship, the confirmed temporal matching relationship of the Yanshanian activity, and the calculated configuration efficiency ratio, a tectonic coupling model is established to quantitatively describe the spatial coupling strength between ore-guiding and ore-hosting structures. The spatial coupling strength of the tectonic coupling model... Quantization is achieved through a comprehensive function, expressed by the following formula:

[0075]

[0076] in: Indicates the spatial coupling strength value. This represents the formation time sequence matching coefficient between deep and shallow structures (value 1 when matching, value 0 when not matching). This represents the ratio of the configuration efficiency between the fluid transport potential index of deep structures and the mineralization space capacity index of shallow structures. This represents the angle (in degrees) between the direction of deep structural distribution and the direction of dominant shallow structural development. The structural coupling model is expressed in the form of a raster diagram, with each raster cell containing a calculated spatial coupling strength value.

[0077] Optionally, during the calculation process of constructing the coupled model, the included angle is... Standardization is performed; when the included angle is close to 90 degrees, the denominator term... Approaching 0 will lead to a spatial coupling strength value If the value increases abnormally, this can be smoothed by introducing a small constant term, for example, by changing the denominator to... To avoid computational singularities while preserving the spatial coupling strength value The negative correlation between the spatial coupling strength and the rationality of the structural configuration means that the closer the included angle is to orthogonal, the stronger the spatial coupling. The larger.

[0078] In one embodiment of the present invention, trend surface analysis and residual anomaly calculation are performed on geochemical sampling data to separate gold element anomalies and associated element anomalies related to local mineralization. The spatial coordinates, anomaly intensity, and anomaly scale parameters of the concentration centers of gold element anomalies and associated element anomalies are spatially superimposed with the spatial location and coupling strength region of the tectonic coupling model. The spatial statistical correlation index between geochemical anomalies and tectonic coupling strength in each superimposed unit is calculated. Based on a preset correlation index threshold, gold mineralization anomaly target areas with high consistency with the tectonic coupling model are delineated, and the consistency score of each target area is recorded. A three-dimensional geological structure framework model of the target area is constructed based on borehole data, geophysical inversion results, and geological profile maps. The spatial relationship between ore-guiding structures and ore-hosting structures in the tectonic coupling model is integrated into the three-dimensional geological structure framework model to form a three-dimensional geological model containing tectonic ore-controlling elements. The three-dimensional spatial distribution data of geochemical anomalies and the distribution data of alteration mineral assemblages are used as attribute bodies to the three-dimensional geological model containing tectonic ore-controlling elements, generating a three-dimensional prediction model that simultaneously contains tectonic ore-controlling elements and mineralization indication elements.

[0079] In the specific implementation, trend surface analysis and residual anomaly calculation were performed on the 1:50,000 soil geochemical sampling data of the study area. A cubic polynomial was used to fit the regional background trend, separating the regional background value and local residual anomalies from the original gold content data. At the same time, the data of associated elements such as arsenic, antimony, and mercury were processed to separate gold anomalies and associated element anomalies of arsenic, antimony, and mercury related to local mineralization. The spatial coordinates of the concentration centers of gold anomalies and associated element anomalies of arsenic, antimony, and mercury, the anomaly intensity represented by the residual anomaly value, and the anomaly scale parameter represented by the anomaly area were spatially overlaid on the spatial location distribution map and the tectonic coupling intensity regional division map of the tectonic coupling model established in the example on the GIS platform. The overlay operation used a 100m × 100m grid cell as the basic analysis unit.

[0080] In some embodiments, the spatial statistical correlation index of geochemical anomalies and tectonic coupling strength within each 100m × 100m grid overlay unit is calculated. Specifically, for each grid unit, the average value of gold element anomaly intensity and the tectonic coupling strength value within the unit are extracted, and the relationship between their product and their respective sum of squares is calculated. Based on a preset correlation index threshold of 0.65, gold mineralization anomaly target areas with high consistency with the tectonic coupling model are delineated. For example, three target areas are delineated, where the spatial statistical correlation index of target area A is 0.78, that of target area B is 0.71, and that of target area C is 0.68. The consistency score for each target area is recorded, with target area A being grade A and target areas B and C being grade B. The formula for calculating the spatial statistical correlation index is expressed as follows:

[0081]

[0082] in: The spatial statistical correlation index represents the strength of the coupling between the geochemical anomaly and the tectonic structure in the i-th grid cell. This represents the normalized geochemical anomaly intensity value within the grid cell. This represents the structural coupling strength value after standardization within the grid cell.

[0083] It is understandable that a three-dimensional geological structure framework model of the target areas is constructed based on borehole core data, borehole logging data, magnetotelluric sounding inversion results, and existing geological profile maps of the three gold mineralization anomaly target areas (A, B, and C). The three-dimensional geological structure framework model includes elements such as stratigraphic sequence boundaries and major fault surfaces. The spatial location and occurrence information of ore-guiding structures and the spatial morphology and combination relationships of ore-hosting structures in the tectonic coupling model are integrated into the three-dimensional geological structure framework model, forming a three-dimensional geological model containing tectonic ore-controlling elements. In this model, northwest-trending regional faults extend in three-dimensional space as ore-guiding channels, while northeast-trending secondary faults and anticline detachment areas are specifically characterized as ore-hosting spaces.

[0084] Optionally, the three-dimensional spatial distribution data of geochemical anomalies are kriging interpolated to generate a three-dimensional attribute volume of gold content. The distribution data of alteration mineral assemblages are then used to generate three-dimensional attribute volumes of silicification and pyrite mineralization intensity based on borehole logging information. These three-dimensional attribute volumes are then used as mineralization display elements and applied to a three-dimensional geological model containing tectonic ore-controlling elements. In a visualization environment, the generated three-dimensional prediction model can simultaneously display ore-guiding structural surfaces, ore-hosting structural entities, gold anomaly halos, and the distribution range of alteration mineral assemblages, achieving an integrated three-dimensional representation of tectonic ore-controlling elements and mineralization display elements.

[0085] In one embodiment of the present invention, the data sources relied upon in the construction of the 3D prediction model are identified, including geological interpretation data, geophysical inversion data, and geochemical interpolation data. The accuracy, resolution, and coverage of each data source are evaluated, and the matching consistency between data is analyzed. Monte Carlo simulation is used to conduct multiple random perturbation simulations on the key interface morphology, structural spatial location, and mineralization boundary of the 3D prediction model. The variation range of key model parameters in the multiple simulation results is statistically analyzed to determine the sensitivity and confidence interval of the key model parameters. Based on the confidence interval of the key model parameters, the control point weights and interpolation algorithm parameters of the geological interface in the 3D geological modeling process are adjusted. For key model parameters with high sensitivity, high-precision data is collected or targeted geological verification is performed to constrain the model optimization direction. The 3D geological modeling process is rerun using the optimized parameters to generate an updated 3D prediction model. The difference in the fitting degree of the 3D prediction model before and after the update at known mineral deposits is compared until the model changes tend to stabilize and conform to geological laws, forming the final mineral exploration prediction result.

[0086] In the specific implementation, the data sources relied upon in the construction of the 3D prediction model were identified. Geological interpretation data came from core logging and comprehensive geological profile interpretation of 42 boreholes in the study area. Geophysical inversion data came from resistivity sections obtained by CSAMT 2D inversion. Geochemical interpolation data came from the gold and arsenic content analysis of primary halo samples from boreholes. The data accuracy of each data source was evaluated. Borehole logging data had high vertical accuracy, but its planar distribution was limited by the borehole spacing of 500 to 1000 meters. The vertical resolution of CSAMT inversion data was approximately 50 meters in shallow areas (<500 meters) and decreased to 150 meters in deep areas (>500 meters). The extrapolation uncertainty of geochemical interpolation data increased significantly in areas without borehole control. Analysis of the consistency between data revealed that, at known orebody locations, there was a 70% spatial overlap between high resistivity anomaly areas and high gold value areas.

[0087] In some embodiments, the Monte Carlo simulation method is used to perform multiple random perturbation simulations on the key interface morphology, structural spatial location, and mineralization boundaries of the three-dimensional prediction model. For the 25 borehole control points controlling the stratigraphic interfaces in the geological interpretation data, a random displacement following a normal distribution with a mean of 0 and a standard deviation of 15 meters is applied in the direction perpendicular to the stratigraphic strike. For the main fault plane interpreted by CSAMT, a random displacement following a normal distribution with a mean of 0 and a standard deviation of 10 meters is applied in its dip direction. For the gold mineralization boundaries delineated based on borehole data interpolation, their boundary grade thresholds are uniformly and randomly sampled within the range of 0.8 g / t to 1.2 g / t. The total number of Monte Carlo simulations is set to 1000.

[0088] It is understandable that the variation range of key model parameters in 1000 Monte Carlo simulations is statistically analyzed. Key parameters of the 3D prediction model include the depth of the main fault plane at -500 meters above sea level, the volume of favorable ore-bearing space in the anticline core, and the total volume of gold mineralization. The sensitivity and confidence intervals of the key model parameters are determined. Sensitivity is evaluated by calculating the coefficient of variation for each parameter in the simulation results. The formula for calculating the coefficient of variation is expressed as:

[0089]

[0090] in: The sensitivity index (i.e., coefficient of variation) represents the key parameters of the model. This represents the standard deviation of the parameter across 1000 Monte Carlo simulations. This represents the arithmetic mean of the parameter across 1000 Monte Carlo simulations. See Table 1 for the statistical characteristics of the key parameters in the simulation results.

[0091] Table 1: Statistical Results of Key Parameters in Monte Carlo Simulation

[0092] Key parameters mean Standard deviation Sensitivity Index 95% confidence interval Depth of the main fault plane (meters) -512 18.5 0.036 [-548,-476] Ore-containing space volume (million cubic meters) 5.8 1.2 0.207 [3.5,8.1] Volume of gold mineralization (million cubic meters) 1.5 0.45 0.300 [0.6,2.4]

[0093] Based on the confidence intervals of the key parameters of the model, the weights of the control points and the interpolation algorithm parameters of the geological interfaces are adjusted during the 3D geological modeling process. For the depth of the main fault plane, its confidence interval is narrow and its sensitivity index is low, so the borehole control points are given higher weights during modeling, and the Kriging interpolation algorithm with smaller errors is adopted. For the volume of gold mineralization, its confidence interval is wide and its sensitivity index is high, so the weight of the single grade threshold boundary is reduced during modeling, and the sequential indicator simulation algorithm that can characterize soft boundaries is adopted.

[0094] Optionally, for the gold mineralization boundary, a key model parameter with high sensitivity, core sample analysis data from three additional verification boreholes were collected in areas with sparse existing borehole control. This was done to obtain more accurate mineralization boundary location information and constrain the direction of model optimization. The 3D geological modeling process was then rerun using the optimized control point weights and interpolation algorithm parameters to generate an updated 3D prediction model. The fit of the 3D prediction model before and after the update at the known boreholes of five industrial ore bodies was compared. The previous model fitted three ore bodies, while the updated model fitted all five. Furthermore, the consistency between the ore body morphology and the observed grade distribution improved. The evaluation and optimization process was repeated until the newly added verification data no longer caused significant changes in the model's key parameters beyond the confidence interval, and the model structure conformed to regional geological patterns, thus forming the final mineral exploration prediction result.

[0095] See Figure 4 This is a bar chart showing the perturbation parameters of a Monte Carlo simulation of a 3D prediction model for Carlin-type gold deposits, clearly presenting the differences in perturbation intensity among three key elements. The chart reflects the perturbation characteristics of different elements in 3D geological modeling, with numerical differences corresponding to the degree of contribution of each element to the model's uncertainty. For stratigraphic interface control points (high uncertainty), their spatial location can be constrained by supplementing with high-precision borehole data, reducing model fluctuations. For the grade of gold mineralization boundaries (low uncertainty), soft boundary interpolation algorithms can be used to more accurately characterize the mineralization range. This type of chart is a core outcome of the uncertainty quantification assessment stage in gold prospecting prediction, providing direct quantitative basis for subsequent model iteration and optimization. Based on this chart, high-precision borehole or geophysical data can be selectively supplemented in areas with high uncertainty, avoiding resource waste on low-risk elements and optimizing the cost-effectiveness of exploration investment.

[0096] In one embodiment of the present invention, the final mineral exploration prediction results are dynamically linked with the deployment of a new round of geological exploration projects. Based on the actual mineralization information fed back by the exploration projects, the multi-dimensional identification marker system, structural coupling model and three-dimensional modeling parameters are calibrated. The calibrated parameters and models are used to continuously update the prediction results of unverified areas, thereby achieving a continuous improvement in mineral exploration prediction capabilities.

[0097] In practice, the final mineral exploration prediction results were dynamically linked with the deployment of the new round of geological exploration projects. Within the Class A prediction target area delineated in the final mineral exploration prediction results, verification boreholes ZK01, ZK02, and ZK03 were designed and constructed. The borehole layout strictly followed the overlapping areas of structural coupling high-value zones and geochemical anomalies indicated by the three-dimensional prediction model. According to the actual mineralization information reported by the exploration project, borehole ZK01 revealed silicified and pyrite alteration at a depth of 320 to 350 meters, with a gold grade of 1.5 g / t. Borehole ZK02 revealed a NE-trending fault fracture zone at a depth of 280 meters, but the gold grade was less than 0.1 g / t. Borehole ZK03 did not show significant mineralization at the predetermined depth. The spatial location, mineralization intensity, and structural attributes of these observed mineralization information were systematically recorded and entered into the database.

[0098] In some embodiments, based on the observed mineralization information from the exploration project, the multi-dimensional identification marker system, the structural coupling model, and the 3D modeling parameters are calibrated. The calibration of the multi-dimensional identification marker system involves adding the specific indicator of an "arsenic-antimony ratio greater than 5" to the geochemical dimension identification markers. This is because the average arsenic-antimony ratio in the primary halo samples of the mineralized section of borehole ZK01 reached 8.7, while in the non-mineralized borehole ZK02, this ratio was less than 2. The calibration of the structural coupling model involves adjusting the weighting factor of the angle relationship between deep and shallow structures in the model. Based on the observed mineralization information, the structural angles corresponding to the mineralized areas are not strictly orthogonal at 90 degrees, but rather distributed in the range of 75 to 105 degrees. Therefore, the constant term of the angle sensitivity function in the model is adjusted. The calibration of the 3D modeling parameters mainly targets the variogram model in kriging interpolation. Based on the actual thickness and strike variations of the newly revealed orebody, the range parameter of the spherical model is corrected from 150 meters to 200 meters.

[0099] It is understandable that the calibrated parameters and models are used to continuously update the prediction results for unverified areas. Taking the Yiyuan Scenic Area, which has not yet been drilled in the study area, as an example, the calibrated geochemical indicator of "arsenic-antimony ratio greater than 5" is substituted into the multi-dimensional identification indicator system for re-identification. The adjusted tectonic coupling model weighting factor is substituted into the calculation of the spatial coupling strength of the Yiyuan Scenic Area, and the corrected variogram model is used for the three-dimensional geological modeling of the Yiyuan Scenic Area. By running the updated complete prediction process, an updated three-dimensional prediction model of the Yiyuan Scenic Area is generated. Compared with the model before calibration, the spatial coupling strength value of a previously C-level prediction area in the updated model has been improved and reclassified as a B-level prediction area. At the same time, a new favorable fracture-dense zone is delineated.

[0100] Optionally, continuous improvement in mineral exploration prediction capabilities can be achieved by establishing a quantitative model prediction accuracy feedback calibration mechanism. This mechanism calculates the consistency index between the model prediction results before and after each calibration and the results of subsequent validation projects. The consistency index... The calculation formula is expressed as follows:

[0101]

[0102] in: The consistency index represents the accuracy of model predictions. This indicates the number of cells correctly predicted by the calibrated model in subsequent validation processes. This indicates the total number of engineering units involved in the verification. After the model update for the B-side scenic area was completed, the results of two subsequent inspection boreholes showed that... The value is 2. The value is also 2, therefore the consistency index after this calibration is... The value was 1.0, which is higher than the consistency index of 0.67 in target area A before calibration, indicating that the mineral exploration prediction capability has been continuously improved through dynamic correlation and rolling updates.

[0103] See Figure 5This is a comparative curve showing the variation of structural intensity with depth in Carlin-type gold prospecting prediction. It clearly demonstrates the spatial distribution of structural intensity between shallow and deep layers. The structural intensity intersection zone at 200–250 meters is the core area of ​​spatial coupling between ore-guiding and ore-hosting structures, providing direct quantitative evidence for the depth range of prospecting target areas. In shallow layers (<200 meters), the focus can be on verifying ore-hosting structures; in mid-deep layers (200–400 meters), the coupling relationship between ore-guiding and ore-hosting structures needs to be considered; and in deep layers (>400 meters), the focus is on tracing ore-guiding structures. This curve can serve as a key parameter constraining the spatial distribution of structures in 3D geological models, improving the accuracy of the model's representation of structural ore-controlling laws. The overlapping area of ​​the curves (purple area) can quantitatively reflect the coupling strength between shallow and deep structures, providing visual support for structural coupling models.

[0104] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion, characterized in that, Includes the following steps: A multi-dimensional identification marker system for ore-guiding and ore-hosting structures is constructed, which covers identification elements in geological, geophysical, and geochemical dimensions. Based on the aforementioned multi-dimensional identification system, multi-source geological data of the target area are collected and integrated to form an initial fused data volume; The initial fused data volume is subjected to structural lattice analysis to extract deep structural information elements related to ore-guiding structures and shallow structural information elements related to ore-hosting structures. The deep structural information elements and the shallow structural information elements are subjected to coupling analysis to establish a structural coupling model that reflects the spatial configuration relationship between ore-guiding structures and ore-hosting structures. Geochemical anomaly information is overlaid with the tectonic coupling model to identify gold mineralization anomaly target areas that have a high degree of agreement with the tectonic coupling model. The gold mineralization anomaly target areas are then characterized in three dimensions using three-dimensional geological modeling technology to generate a three-dimensional prediction model that includes tectonic ore-controlling elements and mineralization indication elements. Uncertainty quantification assessment is performed on the three-dimensional prediction model to determine the sensitivity and confidence interval of the key parameters of the model. Based on the uncertainty quantification assessment results, the three-dimensional prediction model is iteratively optimized to form the final mineral exploration prediction results. The coupling analysis of the deep structural information elements and the shallow structural information elements includes: Calculate the angle between the spatial distribution direction of the deep structural information elements and the dominant development direction of the shallow structural information elements; The analysis examines the relationship between the multi-phase activity history of fault zones in the deep structural information elements and the formation time sequence of structures in the shallow structural information elements. Evaluate the configuration efficiency between the ore-forming fluid transport potential provided by the deep structural information elements and the ore-forming spatial storage capacity provided by the shallow structural information elements; Based on the angle relationship, the formation timing matching relationship, and the configuration efficiency, a structural coupling model is established to quantitatively describe the spatial coupling strength between ore-guiding structures and ore-hosting structures. Spatial coupling strength of the construction coupling model Quantization is achieved through a comprehensive function, expressed by the following formula: ； in, Indicates the spatial coupling strength value. This represents the temporal matching coefficient between deep and shallow structures. This represents the ratio of the configuration efficiency between the fluid transport potential index of deep structures and the mineralization space capacity index of shallow structures. It represents the angle between the direction of deep structure distribution and the direction of dominant development of shallow structure.

2. The Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion as described in claim 1, characterized in that, The construction of a multi-dimensional identification marker system for ore-guiding and ore-hosting structures includes: The system collects regional geological maps, geophysical exploration data, geochemical sampling data, and geological data of known mineral deposits; The regional detachment structures and associated faults of compressional anticlines are interpreted from the geological maps of the region as geological dimensional identification markers of ore-guiding structures. Gravity anomaly gradient zones and magnetic anomaly linear zones reflecting deep structural characteristics are extracted from the geophysical exploration data as geophysical dimension identification markers for ore-guiding structures. Specific elemental combinations anomalies related to deep fluid activity were screened from the geochemical sampling data as geochemical dimension identification markers for ore-guiding structures. The spatial distribution characteristics of tectonic alteration forms, the morphology of anticline axes with specific geometric parameters, and the distribution range of dome structures are summarized from the geological data of the known ore deposits as geological dimension identification markers for ore-hosting structures. The identification marks obtained from different dimensions will be systematically classified and encoded to form a structured, multi-dimensional identification mark system database.

3. The Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion as described in claim 1, characterized in that, The process of constructing a lattice for parsing the initial fused data volume includes: Edge detection and texture analysis algorithms are applied to process geophysical data to identify potential linear structures and boundary information; The geological map is digitized to construct structural elements and spatial topological relationships, clarifying the junction relationships and order of faults at different levels; Geochemical data is gridded and the anisotropy of element spatial distribution is calculated to delineate element enrichment centers and migration trajectories. By combining the linear structure and boundary information, the intersection relationship and order of the fractures, and the element enrichment centers and migration trajectories, a spatial geometric model and a causal relationship model reflecting the structural framework of the target area are constructed.

4. The Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion as described in claim 3, characterized in that, The extraction of deep structural information elements related to ore-guiding structures and shallow structural information elements related to ore-hosting structures includes: In the spatial geometric model and genetic relationship model of the structural lattice, a depth threshold is set to distinguish between deep structural domains and shallow structural domains. Within the deep tectonic domain, fault zones and ductile shear zones with regional extension scale and cutting through deep crustal layers are selected, and their geometric morphology, occurrence parameters, and activity period information are extracted as deep tectonic information elements. Within the shallow structural domain, secondary fractures, densely fractured zones, and folded detachment areas that have a genetic connection or spatial intersection with the deep structural information elements are identified, and their spatial morphology, density, and combination relationship information are extracted as shallow structural information elements.

5. The Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion as described in claim 4, characterized in that, The overlay analysis of geochemical anomaly information with the tectonic coupling model includes: Trend surface analysis and residual anomaly calculation were performed on regional geochemical data to separate gold anomalies and associated element anomalies related to local mineralization. The spatial coordinates, intensity, and scale parameters of the concentration centers of gold and associated element anomalies are spatially superimposed with the spatial location and coupling intensity region of the constructed coupling model. Calculate the spatial statistical correlation index between geochemical anomalies and tectonic coupling strength within each stacking unit; Based on the preset correlation index threshold, gold mineralization anomaly target areas with high consistency with the structural coupling model are delineated, and the consistency score of each target area is recorded.

6. The Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion as described in claim 5, characterized in that, The method of using three-dimensional geological modeling technology to perform three-dimensional characterization of the gold mineralization anomaly target area includes: Based on borehole data, geophysical inversion results, and geological profiles, a three-dimensional geological structure framework model of the target area is constructed. The spatial relationship between the ore-guiding structure and the ore-hosting structure in the structural coupling model is integrated into the three-dimensional geological structural framework model to form a three-dimensional geological model containing structural ore-controlling elements. The three-dimensional spatial distribution data of geochemical anomalies and the distribution data of alteration mineral assemblages are used as attribute bodies to the three-dimensional geological model containing tectonic ore-controlling elements, thereby generating a three-dimensional prediction model that simultaneously includes tectonic ore-controlling elements and mineralization indication elements.

7. The Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion as described in claim 6, characterized in that, The uncertainty quantification assessment of the three-dimensional prediction model includes: Identify the data sources relied upon in the construction of the three-dimensional prediction model, including geological interpretation data, geophysical inversion data, and geochemical interpolation data; Evaluate the accuracy, resolution, and coverage of each data source, and analyze the consistency of matching between data. The Monte Carlo simulation method was used to perform multiple random perturbation simulations on the key interface morphology, structural spatial location, and mineralization boundary of the three-dimensional prediction model. By statistically analyzing the variation range of key model parameters in multiple simulations, the sensitivity and confidence intervals of the key model parameters are determined.

8. The Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion as described in claim 7, characterized in that, The iterative optimization of the three-dimensional prediction model based on the uncertainty quantification assessment results includes: Based on the confidence intervals of the key parameters of the model, adjust the control point weights and interpolation algorithm parameters of the geological interface during the three-dimensional geological modeling process; For key parameters of the model that are highly sensitive, supplementary high-precision data should be collected or targeted geological verification should be carried out to constrain the direction of model optimization. The 3D geological modeling process is rerun using the optimized parameters to generate an updated 3D prediction model; The differences in the fit between the 3D prediction models before and after the update at known mineral deposits are compared until the model changes tend to stabilize and conform to geological laws, thus forming the final mineral exploration prediction results.

9. The Carlin-type gold deposit prospecting prediction method based on multi-source geological data fusion as described in claim 8, characterized in that, After the final mineral exploration prediction results are formed, the process also includes: The final mineral exploration prediction results will be dynamically linked with the deployment of a new round of geological exploration projects; Based on the actual mineralization information reported by the exploration project, the multi-dimensional identification marker system, the structural coupling model, and the three-dimensional modeling parameters are calibrated. By using calibrated parameters and models, the prediction results for unverified areas are continuously updated to achieve a sustained improvement in mineral exploration prediction capabilities.

Images