Regional prospecting method and system based on multi-element space coupling of metallogenic system

Abstract

The invention discloses a regional prospecting method and system based on multi-element space coupling of a metallogenic system, and belongs to the technical field of mineral resource prospecting prediction and comprehensive modeling of the metallogenic system. The method comprises the following steps: acquiring multi-source spatial data covering a mine concentration area; according to a process mechanism of a metallogenic system, converting and constructing the multi-source spatial data into a background domain, a transmission domain, an ore containing domain and a target domain to indicate spatial distribution indexes; constructing a comprehensive metallogenic potential index for representing a spatial distribution pattern of the regional metallogenic potential; an uncertainty prediction index is constructed based on the consistency deviation degree between the domain index responses and the local difference, and synchronous credibility quantification is carried out on the comprehensive mineralization potential evaluation result; and in combination with a joint discriminant relationship between the mineralization potential index and the uncertainty index, performing classified screening and target region delineation on the regional mineralization abnormal unit, so as to realize continuous quantitative prediction of the regional scale mineralization potential and optimization of the prospecting target region, and provide a mineral geological exploration service developed by using a high technology.

Description

Technical Field

[0001] This invention relates to the field of mineral resource exploration prediction and comprehensive modeling of metallogenic systems, specifically a regional mineral exploration method and system based on the spatial coupling of multiple elements of a metallogenic system. Background Technology

[0002] Cobalt (Co) is a crucial strategic metal in new energy battery materials, aerospace alloys, and high-end manufacturing. Its deposits are complex in type and diverse in occurrence, generally existing as isomorphous or associated forms within polymetallic deposit systems such as copper, nickel, and iron. Independent cobalt-rich mineral phases are poorly developed, their surface expression is concealed, and mineralization anomalies exhibit poor identifiability and scattered, superimposed characteristics at the ore cluster scale, posing significant technical challenges to regional mineral exploration prediction and precise target area delineation. These types of deposits, represented by cobalt mines, generally exhibit common problems such as weak mineralization indicators, high heterogeneity of anomaly signals, and insufficient spatial constraint.

[0003] Existing mineral exploration prediction technologies mainly rely on methods such as empirical ratios of remote sensing spectra, anomaly enhancement transformation, target spectral matching, or sample classification prediction based on machine learning to perform superposition analysis and statistical discrimination of single-factor or multi-source spatial information. These methods primarily rely on correlation learning and empirical pattern matching, focusing on the extraction and expansion of anomaly features. They lack systematic modeling and quantitative expression of the intrinsic causal relationships between the ore-forming supply background, fluid transport channels, surrounding rock reaction and precipitation environment, and the spatial response of mineralization products. Under complex geological conditions, non-ore-forming geological processes (such as carbonate deposition, weathering and ferrification, exposure of tectonic alteration zones, and vegetation stress effects) can easily produce observational responses similar to mineralization anomalies. This leads to problems such as "high response - low mineralization" and "concentrated misjudgment of anomalies" in prediction results based on empirical thresholds or local superposition discrimination, weakening the stability and interpretability of mineral exploration decisions.

[0004] In addition, existing prediction technologies are mostly based on pixel-level anomaly classification or probability threshold classification, which makes it difficult to depict the continuous spatial transmission, hierarchical structure and convergence law of mineralization processes. They also lack effective means to simultaneously quantify and evaluate the confidence and uncertainty of prediction results, thus limiting the reliability of prediction results in engineering mineral exploration deployment.

[0005] Therefore, there is an urgent need for a prospecting method and system that can take the metallogenic system process as a mechanism constraint, transform multi-source observation information into a multi-domain spatial indication distribution that reflects the metallogenic background, transport conditions, reaction space and target response, and realize continuous prediction of metallogenic potential and simultaneous assessment of uncertainty through spatial coupling and comprehensive modeling, so as to overcome the problems of insufficient spatial interpretation ability and engineering stability of existing experience superposition prediction technology. Summary of the Invention

[0006] To address the aforementioned issues, this invention provides a regional mineral exploration method and system based on the spatial coupling of multiple elements within a mineralization system. This invention focuses on mineralization processes such as ore supply, migration and conduction, reaction and precipitation, and target response. By constructing a multi-domain mineralization indicator distribution system, it transforms multi-source spatial observation indicators into mineralization process variables with physical genetic constraints. Based on this, it achieves continuous quantitative prediction of mineralization potential and simultaneous assessment of prediction uncertainties, thereby improving the stability, interpretability, and reliability of cobalt mineral exploration prediction results for engineering applications.

[0007] To achieve the above objectives, this invention provides a regional mineral exploration method based on the spatial coupling of multiple elements of an ore-forming system, specifically including the following steps: S1. Acquisition and Preprocessing of Multi-Source Spatial Data: Acquire multi-source spatial data covering the target mining area. The data includes at least one or more of the following types: satellite remote sensing image data, geophysical data, geochemical data, digital elevation model data, and geological logging and structural interpretation data. Perform coordinate unification, radiometric or numerical calibration, noise suppression, and spatial registration on the above data to construct a standardized basic spatial dataset.

[0008] S2, Construction of multi-domain indicator distribution of ore-forming system: Based on the process model of the mineralization system, namely "source-channel-reaction site-abnormal response", the multi-source spatial data are transformed into the following four types of mineralization indicator distributions: Background domain indicates distribution: used to characterize the ore-forming supply background and favorable stratigraphic features; Transport domain indication distribution: used to characterize the migration channels of ore-forming fluids and tectonic ore-controlling conditions; Ore-bearing zone indicator distribution: used to reflect the intensity of alteration reaction of the surrounding rock and the spatial conditions of mineralization precipitation; Target domain indication distribution: used to characterize mineralization products and their near-surface integrated response characteristics.

[0009] S3. Normalization and smoothing of indicator distribution: Quantile normalization or extreme value normalization is performed on the indicator distributions of the background domain, transmission domain, ore-bearing domain and special target domain respectively, and they are standardized and mapped to a unified numerical range. A Sigmoid-type mapping function or an equivalent continuous smoothing function is introduced to perform nonlinear transformation on the index response in order to suppress local noise interference, enhance the high response signal related to mineralization and maintain spatial continuity. S4 Comprehensive Metallogenic Potential Index Construction: The standardized four-domain indicator distribution fields are weighted and fused according to preset weights to construct the comprehensive metallogenic potential index MPI (Metallogenic Prospectivity Index), expressed as follows: in: , , and The standardized indicators are categorized into four types: background domain, transmission domain, ore-bearing domain, and special target domain. For any spatial pixel point; , and The power-law weighting coefficients for the first three domains (background domain, transmission domain, and ore-bearing domain) are used to control the coupling contribution weight of each domain to the mineralization potential. To enhance the weighting coefficients of specific target domains and control the degree of reinforcement of the direct mineralization response in the overall potential, target domain indices are introduced as enhancement factors. S5 Uncertainty Index Calculation; Based on the consistency and local variability of the spatial isotopic responses of each domain indicator distribution, a prediction uncertainty index U(x) is constructed to simultaneously quantify the credibility of the prediction results corresponding to MPI. This invention will improve the inter-domain response separation. D(x) Local differences V(x) After standardization and weighted fusion, an uncertainty index is constructed, expressed as follows: in: For inter-domain response separation weighting coefficients, For local variance weighting coefficients, satisfying + =1; Z is the normalization operator, which maps the values ​​of U(x) uniformly to the interval [0,1].

[0010] The S6 comprehensive assessment of mineralization target areas is based on the joint distribution relationship between MPI and U, and the target mineralization areas are graded and evaluated: areas with high MPI and low U are identified as preferred target areas; areas with high MPI and high U are identified as potential anomaly areas; and areas with low MPI are identified as mineralization background areas; thereby realizing the intelligent delineation and priority ranking of cobalt prospecting target areas.

[0011] This invention provides a regional mineral exploration system based on the spatial coupling of multiple elements of an ore-forming system, specifically including: The system comprises six functional modules: data acquisition and preprocessing, multi-domain indicator distribution construction, indicator distribution normalization and smoothing, comprehensive mineralization potential index construction, uncertainty index calculation, and target area comprehensive judgment and output. During system operation, the raw multi-source spatial data is first standardized by the preprocessing module, outputting spatially aligned data cubes. Subsequently, the multi-domain indicator construction module generates four types of indicator distributions—background domain, transmission domain, ore-bearing domain, and target domain—based on the "source-channel-reaction site-abnormal response" mechanism of the mineralization system. After unification and enhancement by the normalization and smoothing module, the indicators for each domain are input to the mineralization potential index module and the uncertainty index module for parallel calculation, yielding the comprehensive mineralization potential index (MPI) and the uncertainty index (U). Finally, the target area judgment module uses a joint discrimination rule based on MPI and U to achieve graded target area delineation and output results. All modules are sequentially connected and collaboratively operate through standard data interfaces, forming a complete automated workflow of "data input—indicator construction—fusion calculation—target area output," enabling continuous quantitative prediction of mineralization potential and synchronous reliability assessment. The system also supports feedback optimization of weights and thresholds through a parameter verification module, thereby improving prediction stability and adaptability.

[0012] Compared with existing technologies, this invention provides a regional mineral exploration method and system based on the spatial coupling of multiple elements of an ore-forming system, which has the following beneficial effects: (1) This invention establishes a multi-domain spatial indication coupling model based on the ore-forming system mechanism, which breaks through the limitations of traditional empirical spectral matching and statistical correlation analysis, and improves the physical interpretability and traceability of mineral exploration prediction results; (2) This invention achieves continuous quantitative expression of mineralization probability by constructing a comprehensive mineralization potential index, overcoming the problem that traditional discrete anomaly extraction methods cannot reflect enrichment gradient and spatial hierarchical structure; (3) The present invention introduces an uncertainty index to simultaneously evaluate the reliability of the prediction results, thereby enhancing the credibility and risk controllability of mineral exploration results in engineering decision-making applications; (4) This invention realizes multi-domain collaborative constraints of background, channel, ore-bearing capacity and anomaly response, which upgrades mineral exploration prediction from single-factor anomaly identification to systematic prediction driven by mineralization process, and improves the stability and accuracy of target area delineation under complex geological background. (5) This invention introduces the constraints of the ore-forming system mechanism into the multi-source spatial information fusion modeling process, and constructs a continuous prediction model driven by the synergistic effect of the ore-forming process elements. It breaks through the problem that the traditional empirical anomaly superposition method is insufficient in expressing the spatial structure constraints of the ore-forming system. It can realize the precise identification and stable positioning of regional mineral exploration prediction under complex geological background conditions. It has the advantages of strong applicability, high spatial constraint capability and good engineering promotion.

[0013] (6) This invention deeply integrates satellite remote sensing data, image enhancement processing and mineralization system modeling to form a complete high-tech mineral geological exploration service solution, which significantly improves the engineering application capability of regional mineral exploration prediction. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the overall process of the mineral exploration prediction method based on the spatial coupling of multiple elements of the metallogenic system according to the present invention; Figure 2 Figure 1 shows a comparison of the index responses of the background domain, transmission domain, ore-bearing domain, and special target domain of the present invention. Figure 2 shows a comparison of the index response results of the background domain, Figure 3 shows a comparison of the index response results of the transmission domain, Figure 4 shows the index response results of the ore-bearing domain, and Figure 5 shows a comparison of the index response results of the special target domain. Figure 3 Figure 1 shows the detection and spatial response results of the cobalt mineralization spectral endmembers based on constrained energy minimization (CEM) according to the present invention. Figure 2 shows the CEM response results of the cobalt mineralization spectral endmembers in the full spectrum range of 350–2500 nm overlaid with GF-1 panchromatic images. Figure 3 shows the CEM response results of the cobalt mineralization spectral endmembers in the preferred SWIR band of 1100–2500 nm overlaid with GF-1 panchromatic images. Figure 4 Figure (a) is a grayscale map of the mineralization potential index, and Figure (b) is a mineralization potential grading structure map. Figure 5 Figure (a) is a grayscale image of the uncertainty distribution, and Figure (b) is a hierarchical image of the uncertainty distribution. Figure 6 Figure (a) shows the MPI-U joint target classification and delineation results and the predicted target classification map of the present invention, wherein Figure (b) is the joint target classification and delineation result map and Figure (a) is the predicted target classification map. Figure 7 This is a schematic diagram of the overall architecture of the software functional modules of the mineral exploration prediction system of the present invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Please see Figures 1-7A regional mineral exploration method and system based on the spatial coupling of multiple elements of an ore-forming system includes the following steps: S1. Acquisition and preprocessing of multi-source spatial data; In this embodiment, a multi-source spatial data and basic sample data system covering the target mining area and key parts of the mining area is first constructed. Through sample collection, mineralogical and geochemical characterization, experimental spectral library construction and spatial geoscience data acquisition, the basic data conditions for subsequent multi-domain indicator distribution construction and mineralization potential prediction are established.

[0017] Specifically, the following steps are included: S1.1 Select the target mining area and collect mineralized rock samples: This embodiment selects a mining area located in a typical stratabound cobalt-rich zone within the Kangdian metallogenic belt as the verification area. This area belongs to a sedimentary-hydrothermal superimposed metallogenic system, and its mineralization is synergistically controlled by the distribution of ore-bearing strata and regional tectonic transport channels, thus possessing typicality and verifiability.

[0018] Sampling points were set up along known mineralization concentrations within the target mining area, and several outcrops and mineralized rock samples were collected, prioritizing samples with high mineralization intensity and typical mineral assemblage characteristics. The collected samples were mainly hosted in dolomite and siliceous dolomite strata, commonly enriched in cobalt travertine, and associated with sulfide minerals such as chalcopyrite and chalcocite. The mineralization morphology was mainly characterized by a composite distribution of veins, nodules, and disseminated minerals.

[0019] The samples were used for subsequent mineralogical identification, elemental composition analysis, and experimental spectral endmember construction.

[0020] S1.2 Based on mineralized rock samples, mineralogical identification and geochemical testing were conducted to obtain a quantitative table of mineral composition and a geochemical dataset; To constrain the composition and spectral response mechanism of ore-forming materials at the experimental scale, mineralogical analysis and geochemical tests were performed on the collected samples, specifically including the following steps: (1) Mineralogical identification: X-ray diffraction (XRD) was used to identify and quantify the mineral phases of the samples, determining the main mineral composition and assemblage characteristics. The identified minerals include, but are not limited to: Cobalt-based minerals such as cobaltite and cobalt arsenide; associated sulfide minerals such as chalcopyrite and chalcocite. Carbonate minerals such as calcite and dolomite; alteration minerals such as sericite and chlorite.

[0021] (2) Geochemical tests: Using X-ray fluorescence spectrometry (XRF), the contents of various major and trace elements, including but not limited to, were determined after the sample surfaces were grouped according to their color and mineralization characterization. Al, As, Ca, Co, Cr, Cu, Fe, K, Mg, Mn, Ni, P, S, Si, Ti, Zr and other elements.

[0022] The testing process was conducted under standard analytical conditions, with detection limits controlled within the ppm range to ensure effective characterization of trace element distribution and enrichment. Experimental data were normalized and significant outliers were removed to establish quantitative comparability between sample component types, serving as a reference for subsequent spectral endmember classification, verification, and calibration.

[0023] S1.3 Based on mineralized rock samples, spectral data were acquired and endmembers were constructed to obtain a cobalt ore spectral endmember library; Under laboratory conditions, a spectral measurement device (such as an ASD FieldSpec type or equivalent device) with visible-near-infrared-short-wave infrared band coverage and a microscopic hyperspectral imaging system are used to perform reflectance spectral measurements on samples in situ and in micro-regions, obtaining continuous reflectance spectral data with a band coverage range of 350 nm to 2500 nm, and thus obtaining hyperspectral imaging data.

[0024] Based on the mineral composition, elemental abundance characteristics, and mineralization classification results of the samples, the experimental spectral data were classified and analyzed, and an experimental spectral endmember model was established, including: High cobalt end-members: typical spectral response characteristics of mineral assemblages corresponding to Co–As and Co–S systems; Low cobalt or background endmembers: corresponding to the spectral characteristics of host rock carbonate minerals, clay minerals, and weak alteration zones.

[0025] The constructed endmembers are processed by band resampling and sensor response function matching to convert them into equivalent spectral endmembers with the same spectral band setting structure as the space observation images, so as to realize the comparability, transferability and joint use between experimental scale endmembers and regional remote sensing observation data.

[0026] S1.4 Multi-source spatial data acquisition and preprocessing; Acquire a multi-source spatial dataset covering the target mining area, wherein the data includes at least one or a combination of the following: remote sensing optical or hyperspectral image data; digital elevation model (DEM) data; tectonic linear bodies, fault distribution data and regional geological logging data; optional gravity, magnetic and electromagnetic detection data; geochemical spatial raster data.

[0027] Remote sensing optical or hyperspectral imagery can originate from airborne platforms, satellite platforms, or ground imaging systems; DEMs can be obtained based on radar mapping, stereo image pair measurements, or publicly available elevation products.

[0028] Before entering predictive modeling, all spatial data undergoes the following preprocessing steps: unified coordinate projection system; geometric fine registration; optical image radiometric calibration, atmospheric correction, and terrain effect correction; data resampling and spatial grid format conversion, thereby constructing a pixel-level multi-source data cube under the same spatial reference frame, providing standardized input data conditions for the construction of four-domain indicator distribution and spatial fusion analysis.

[0029] S2. Based on multi-source spatial data, construct four-domain spatial indicators to obtain background domain indicator field, transmission domain indicator field, mineral storage domain indicator field and special target domain indicator field. After completing the acquisition of S1 multi-source spatial data, such as Figure 2 As shown, based on the "source-channel-site-response" process model of the sedimentary-hydrothermal superimposed cobalt mineralization system, four types of spatial index fields are constructed: background domain, transport domain, ore-hosting domain, and specific target domain. The index design for each mineralization domain is based on the mineral spectral absorption mechanism and remote sensing detectable characteristics, and specifically includes the following steps.

[0030] S2.1 Based on multi-source spatial data, construct the background domain indicator field to obtain the background domain indicator field; S2.1.1 Establishment of Background Mineral Spectral Reference: Based on the stratabound sedimentary-metamorphic formation characteristics of the target mining area, dolomite, calcite, carbonaceous shale, and gypsum were selected as representative mineral assemblages of the background domain. Using mineral spectral libraries and / or measured spectral data, their range in the 2.315–2.350 μm (CO3²⁻) region was determined. - The characteristic absorption behaviors in the bands of vibration, 1.90 μm (H2O absorption) and 0.65–0.80 μm (organic matter C–C π–π* absorption) are used as spectral references for constructing background domain indices.

[0031] S2.1.2 Background Spectral Index Construction: Based on the above absorption band locations and diagnostic characteristics, at least one or more background spectral indices are constructed on multispectral or hyperspectral images, including but not limited to: a) Carbonate index: Constructed using one or more bands around 2.30–2.35 μm to characterize the CO3² of the Dol–Cal combination, representing reflectivity ratios, absorption depths, or continuous deenvelopment depths. - a) Absorption intensity and peak position variation; b) Hydration index: A hydration index is constructed using an absorption band of approximately 1.90 μm to distinguish between gypsum-type evaporative formations and non-evaporative formations; c) Carbonaceous rock index: A reflectance ratio or slope-type index is constructed by combining the overall decrease in reflectance and the gentle red edge characteristics of visible-near infrared (0.5–1.0 μm) to identify carbonaceous formations and stratabound reducing environments.

[0032] The specific band combinations and coefficients can be adapted and adjusted according to the center wavelength and bandwidth of the sensor used.

[0033] S2.1.3 Background Domain Comprehensive Score Calculation: After normalizing each individual background index, a linear or fuzzy weighted average is performed according to preset weights to obtain the background domain comprehensive score for each spatial pixel. This characterizes the mineralization supply background, favorable stratigraphic assemblages, and geochemical baseline conditions at this location. The comprehensive score function expression for the background domain is as follows: in, Representing spatial pixels The overall score of the background domain; For the first Background domain indicators at the pixel The response value at the specified location; Z(·) is the normalization operator; The total number of background domain indicators; For the preset weighting coefficients, satisfy The weighting coefficients can be adaptively determined based on regional geological weighting experience, sample calibration relationships, or the entropy weighting method.

[0034] S2.2 Based on multi-source spatial data, construct the transmission domain index field to obtain the transmission domain index field; S2.2.1 Establishment of Spectral References for Transport Minerals: Muscovite, sericite, and chlorite were selected as indicator mineral assemblages for fluid migration-alteration zone formation. Based on mineral spectral libraries or measured spectra, the center wavelength λ_c, full width at half maximum (FWHM), and absorption depth ratio BD of the Al–OH and Mg–OH binding absorption bands in the range of 2.195–2.350 μm were determined. 2.33 / BD 2.20 The variation pattern is as follows: during enhanced alteration, λ_c shifts to the right from approximately 2.20 μm to approximately 2.31 μm (Δλ≈0.13 μm), FWHM broadens, and BD... 2.33 / BD 2.20 Increase.

[0035] S2.2.2 Construction of the transmission domain spectral gradient index: This is constructed on remote sensing imagery using one or more bands covering 2.20–2.35 μm. a) Al–OH / Mg–OH absorption center migration index, used to characterize the spatial transition from muscovite-rich to chlorite-rich alteration zones; b) Absorption zone width or shape index, used to characterize regions of enhanced alteration intensity; c) BD 2.33 / BD 2.20The ratio index is used to characterize the regions where the Mg–OH component is relatively enhanced.

[0036] The above indicators are obtained through the band ratio method, the continuous envelope removal method, or the spectral curvature analysis method, without limiting the specific algorithm form.

[0037] S2.2.3 Construction of geometric channel indicators: Based on the DEM, topographic factors such as slope and curvature are derived, the slope break density and slope aspect consistency index are calculated, and the fracture density, linear volume aggregation degree and fracture intersection node distribution are obtained by combining the linear structure extraction results, which are used to quantify the geometric skeleton of potential fluid transport channels.

[0038] S2.2.4 Calculation of Transmission Domain Composite Score: The spectral gradient index and the construction geometry index are normalized and weighted on a uniform grid to obtain the transmission domain composite score. This is used to characterize the development level of fluid migration channels and tectonic ore-controlling conditions at this location. The expression for the transport domain comprehensive score function is as follows: in, It represents the comprehensive score of the transmission domain at pixel x; it is used to quantitatively characterize the development degree of fluid migration channels and tectonic ore-controlling conditions within the target ore area. For the first The response value of the spectral gradient or geometric channel index at pixel x; Z(·) is the normalization operator; This represents the total number of spectral gradients. For the preset weighting coefficients, satisfy = 1.

[0039] S2.3. Based on multi-source spatial data, construct the ore-bearing domain index field to obtain the ore-bearing domain index field; S2.3.1 Establishment of spectral references for ore-bearing minerals: Dolomite + Calcite, chlorite, and SiO2 silicified mineral assemblage were used as representative indicator minerals of the ore-hosting domain. Based on mineral spectral libraries and / or measured data, their spectral signature in the 2.25–2.35 μm band (CO3²) was determined. - Characteristic response behavior at the 1.90 μm band (H2O absorption) and the OH group complex absorption band.

[0040] When the fluid-surround rock reaction is enhanced, it manifests as CO3² - Enhanced absorption, weakened H2O absorption, and increased overall reflectivity form a stable bright-colored alteration band pattern, providing physical spectroscopic basis for remote sensing identification of metal precipitation reaction interfaces.

[0041] S2.3.2 Construction of Spectral Indices for Ore-Bearing Areas: At least one or more ore-hosting domain characterization indices are constructed by combining multispectral or hyperspectral image data, including: a) Dol+Cal Index: based on the absorption depth, spectral broadening, or continuous envelope curvature parameters of the 2.30–2.35 μm carbonate absorption band, comprehensively characterizing the intensity of carbonate metasomatism and recrystallization reactions; b) Chlorite Index: a mineral abundance index constructed using the 2.25–2.30 μm Mg–OH absorption characteristics, used to characterize the spatial distribution characteristics of the carbonate-chloritization metasomatism reaction zone; c) Hydration Suppression Index: a response index constructed based on the weakening degree of the 1.90 μm hydration absorption band, used to identify the areas where dehydration-precipitation reactions occur and the changes in the degree of reaction.

[0042] The specific band combination and coefficient settings of the various indicators can be adapted and adjusted according to the center wavelength, bandwidth and signal-to-noise ratio characteristics of the remote sensing sensor used, without limiting the specific algorithm implementation.

[0043] S2.3.3 Calculation of Comprehensive Score for Ore-bearing Area: After standardized processing by unified spatial grid registration, the above-mentioned individual ore-bearing area indicators are superimposed and calculated according to preset weights or fuzzy fusion rules to construct the comprehensive score function for ore-bearing area, as shown in the following expression: in: Represents a pixel The comprehensive score of the ore-bearing area is used to quantitatively characterize the fluid-surrounding rock reaction intensity, the location of the metal precipitation interface, and the degree of development of effective ore-bearing space within the target mining area. Indicates the first Project capacity mineral indicators in pixels The response value at the specified location; Z(·) is the normalization operator; This represents the total number of ore-holding indicators. For the weighting coefficients, satisfying = 1.

[0044] S2.4 Based on the cobalt ore spectral endmember library and multi-source spatial data, a specific target domain index field is constructed to obtain the specific target domain index field; The specific target domain is used to characterize spatial anomalies that are directly or indirectly related to cobalt mineralization products and their near-surface secondary responses. The target domain indices are divided into two categories: direct target domain indices and indirect target domain indices.

[0045] S2.4.1 Direct Target Domain: Construction of cobalt-based endmember spectral indices, specifically including the following steps: S2.4.1.1 Construction of cobalt ore spectral endmember library; Based on laboratory spectral and microscopic hyperspectral imaging data obtained from S1, erythrite, cobaltite, and other representative minerals of the Co–As–S system were selected as target spectral endmember samples to establish a cobalt spectral endmember reference library, such as... Figure 3 As shown.

[0046] The main spectral diagnostic features of the cobalt-based endmembers include, but are not limited to: Co² at approximately 0.56 μm + –O² - Electronic transition absorption; a transition metal ion coordination absorption band of approximately 1.0 μm; Co–O characteristic vibrational absorption of approximately 2.12 μm; and As–O and Co–S complex vibrational absorption clusters in the 2.34–2.48 μm range.

[0047] The cobalt ore endmember spectrum is resampled and processed by sensor response function matching, and then converted into an equivalent endmember vector consistent with the target multispectral or hyperspectral image band system for subsequent space target detection calculations.

[0048] S2.4.1.2 Extraction of the cobalt-based target response field; In remotely sensed hyperspectral imagery, using cobalt endmember vectors as target detection spectral elements, constrained energy minimization (CEM) or other target enhancement detection algorithms (including but not limited to spectral angle matching, spectral feature fitting, minimum subspace demixing, or depth feature mapping methods) are employed to perform pixel-by-pixel calculations on the image pixels to obtain a raster of spatial response intensity distribution of cobalt endmembers. .

[0049] in, This represents the response intensity value of pixel x in the cobalt ore target detection operation, used to quantitatively characterize the degree of direct spectral anomaly indication of cobalt mineralization.

[0050] When multiple cobalt-based endmembers (such as Co–As type endmembers, Co–S type endmembers, etc.) are used, the corresponding response fields can be calculated separately and the comprehensive target response intensity can be obtained by weighted synthesis to reflect the comprehensive spatial distribution characteristics of multiple types of cobalt mineralization.

[0051] S2.4.2 Indirect Target Domain: Construction of Secondary Anomalies and Environmental Response Indicators, specifically including the following steps: S2.4.2.1 Construction of spectral indices for oxide capping: Based on the remote sensing response characteristics of near-surface iron-rich environments, and utilizing Fe³⁺ from goethite, limonite, and other minerals...+ Oxides in: The wide absorption valley formed in the 0.86–0.95 μm range and the hydration absorption correlation bands at 1.40 μm and 1.90 μm were used to construct an iron oxide index or iron cap indicator index to quantitatively characterize the spatial development of surface oxide caps and ferritic alteration in mineralized areas.

[0052] S2.4.2.2 Construction of vegetation anomaly indicators: In vegetation-covered areas, based on the spectral anomalies in vegetation caused by heavy metal stress, biogeochemical indirect indicators are constructed, including but not limited to: the Normalized Difference Vegetation Index (NDVI) plateau shift index; the red edge position shift index; and the near-infrared reflectance suppression index.

[0053] The above indicators are used to characterize the remotely identifiable disturbance effect of mineralization anomalies on the physiological state of surface vegetation.

[0054] S2.4.3 Calculation of Comprehensive Score for Specific Target Domain: Response grid for direct target Iron oxide indicator and vegetation abnormality indicators A unified spatial registration and standardization process is performed, and then a fusion weighting is applied based on preset weight coefficients to construct a comprehensive response function for the specific target domain. in: Z(·) represents the comprehensive score of the specific target domain at pixel x; Z(·) is the standardization operator; For direct target response raster weighting coefficients; The weighting coefficient for the iron oxide indicator; The weighting coefficients for vegetation anomaly indicators; and satisfying the following conditions: + + = 1. Among them, high weights can be assigned to indicators in the direct target domain, and adjustment weights can be set for indicators in the indirect target domain, so as to take into account the fusion judgment of the direct response and secondary characterization information of the ore body.

[0055] The background domain is formed through the above steps. Transmission domain Mineral storage area With specific target domain The four types of spatial index distribution fields serve as the unified input data basis for subsequent S3 index normalization and smoothing processing and the construction of the S4 comprehensive mineralization potential index.

[0056] S3, index distribution normalization and smoothing; To ensure the comparability and fusion of multi-domain indicators with different sources and dimensions under a unified spatial scale, the constructed background domain is analyzed before conducting comprehensive modeling of mineralization potential. Transmission domain Mineral storage area With specific target domain The distribution of the four types of indicators undergoes standardization, normalization, and continuous smoothing, specifically including the following steps: S3.1, Index field normalization processing; S3.1.1 Standardized mapping of four-domain index values: Distribution of the four types of raw indicators: , The index values ​​are linearly mapped to a unified standard interval of [0,1] using either quantile normalization or extreme value normalization methods. Preferably, the following quantile normalization method is used: in: Indicates the metallogenic domain In pixels The original index value at the location; , These are the 5th and 95th quantiles of index k within the target mining area, respectively. This is the normalized dimensionless index value, whose numerical range is limited to the interval [0,1].

[0057] When the noise of the original index is low and the dynamic range is stable, the extreme value standardization method can be used instead of the quantile method, and its form is as follows: in: , Indicators The minimum and maximum values ​​within the target mining area. Through the above dimensionless processing, a unified scale expression is achieved for indicators from different observed physical quantity sources (spectral absorption intensity, tectonic density, topographic geometry, geochemical concentration, etc.), providing a numerical basis for subsequent continuous fusion and calculation of mineralization potential index.

[0058] S3.1.2 Spatial outlier suppression and numerical robustness: During the index normalization process, extreme high and low value pixels exceeding the preset quantile interval are subject to truncation constraints, uniformly limiting them within the quantile boundary range. This suppresses the nonlinear amplification effect of isolated outliers or local noise pixels on subsequent weighted fusion operations, thereby: Reduce the impact of local noise disturbances on the overall prediction stability; prevent isolated anomalies from dominating the spatial distribution of mineralization potential; and enhance the continuity and reliability of index distribution at the regional scale.

[0059] S3.2, Continuous smoothing of indicators and response enhancement mapping, specifically includes the following steps: S3.2.1 Continuous nonlinear mapping processing; To enhance the weight contribution of high-response areas related to mineralization in the fusion modeling, while suppressing unnecessary interference from medium- and low-response areas on the potential index, the standardized index values ​​were adjusted. Apply a continuous nonlinear smoothing transformation.

[0060] Preferably, the Sigmoid mapping function is used: in: This is the index response function after smoothing mapping; is the slope adjustment parameter, used to control the steepness of the mapping curve; b is the threshold shift parameter, used to set the position of the response enhancement interval, usually set in the median or high quantile region of the statistical distribution of the index.

[0061] When the function form is not limited, any of the following equivalent smoothing functions can be used to achieve continuous mapping, including but not limited to: hyperbolic tangent function (tanh); power function mapping; sliding weighted kernel smoothing; and continuous normalized stretching function.

[0062] S3.2.2, Enhancement of high-response mineralization signals; The following effect is achieved through the above nonlinear smoothing mapping: Low-value noise compression: Low-value responses to background noise and non-ore-forming disturbances are compressed close to 0, reducing their impact on the ore-forming potential index; High-value mineralization is amplified: the mid-to-high value responses related to ore-forming material enrichment, enhanced fluid channels, reaction insitu, and mineralization products are boosted, increasing their weight in subsequent MPI calculations; Spatial gradient continuity: The smoothing function maintains the continuity of response changes between pixels, avoids spatial "faults" caused by hard thresholds, and preserves the hierarchical transition characteristics of the mineralization system in space.

[0063] After processing in step S3, the following four types of continuous, dimensionless, noise-suppressed standardized response index distribution fields are formed: , , and The standardized distribution serves as the unified input data basis for constructing the mineralization potential index MPI and calculating the uncertainty index U.

[0064] S3.3, Inter-domain gating and non-metallic area removal (optional implementation steps); To avoid generating inflated potential or pseudo-anomaly responses in areas that clearly lack mineralization background or effective fluid transport conditions, this invention can introduce a spatial gating mechanism with background and transport domains as the main constraints, based on index normalization and continuous smoothing, to further limit the calculation area of ​​subsequent mineralization potential index.

[0065] S3.3.1 Definition of Space Gating Function; Preferably, a gate function is defined. as follows: in: , These are the values ​​of the background domain index and the transmission domain index after standardization and continuous smoothing processes S3.1–S3.2, respectively. , The response thresholds for the background and transmission domains are determined by using quantiles, adaptive thresholds, or empirical critical values ​​based on the index distribution characteristics. For example, the 60th percentile value or its equivalent interval of each index field is preferred.

[0066] S3.3.2, Gating Applications and Functions; In the subsequent construction and calculation of the mineralization potential index MPI and the uncertainty index U, only those satisfying the following conditions are considered: For spatial pixels with a value of 1, effective fusion calculations are performed, and regions that do not meet the gating conditions are assigned a masking weight or are directly removed.

[0067] S3.3.3, Technical effects of the gating mechanism; By using inter-domain gating constraints, the following technical effects are achieved: areas without mineralization background are eliminated, specifically as follows: (1) Implement overall shielding for areas lacking ore-forming material supply conditions or ore-bearing strata constraints; (2) Weaken ineffective transport channels. Reduce the potential contribution weight of space units that do not have fluid transport channels or structural control conditions; (3) Limit the working range of the ore-forming system to ensure that subsequent potential index and uncertainty calculations are carried out only in the action space of the potential ore-forming system with complete "material background + fluid transport" basic conditions, thereby reducing the nonlinear interference of heterogeneous geological responses on the prediction results.

[0068] S3.3.4, Input data formation; After processing through steps S3.1–S3.3, the distributions of the four types of standardized indicators are as follows: , , and Under the constraints of the gate function G(x), they together constitute: A standardized input index system with noise suppression, response enhancement and physical-geological process constraints serves as the unified input data basis for the subsequent calculation of the comprehensive mineralization potential index MPI (S4) and the simultaneous evaluation of the uncertainty index U (S5) based on physical priors.

[0069] S4, Construction of the Comprehensive Metallogenic Potential Index (MPI); After completing the standardization, smoothing, and gating of the four-domain indicators (S3), the remaining effective pixels are fused according to a multi-domain coupling model with physical prior constraints to construct the Metallogenic Prospectivity Index (MPI). (See [link to relevant documentation]). Figure 4 It is used to characterize the continuous spatial distribution pattern of cobalt mineralization potential.

[0070] S4.1 Construction of a multi-domain weighted coupling model; Preferably, the four types of standardized indicators—background domain, transport domain, ore-hosting domain, and specific target domain—are regarded as numerical projections of different geological process variables of the ore-forming system in space, and are respectively denoted as: , , and ; The Comprehensive Mineralization Potential Index (MPI) is defined as follows: in: For any spatial pixel point; , and The power-law weighting coefficients for the first three domains (background domain, transmission domain, and ore-bearing domain) are used to control the coupling contribution weight of each domain to the mineralization potential. To enhance the weighting coefficients of specific target domains and control the degree of reinforcement of the direct mineralization response in the overall potential, target domain indices are introduced as enhancement factors. The aforementioned weighting coefficients are determined based on sample calibration, expert weighting, or statistical robust optimization methods, and their value range is not fixed.

[0071] S4.2, Fusion Mechanism and Physical Prior Constraints; In the MPI construction model, the ore-forming domains are not simply linearly superimposed, but follow the physical operation logic of the ore-forming system of "background constraints - channel transport - reaction precipitation - target response enhancement", specifically manifested as follows: Background domain With transmission domain It has a prerequisite constraint effect on mineralization behavior, and its weights are entered into the model in a multiplicative structure to limit the spatial boundary of potential mineralization systems. Mineral storage area As a characterization factor of the actual metal precipitation space, the superposition results of the first three domains are subjected to reaction coupling modulation; Specific target domain The direct expression signals reflecting mineralization products and their surface anomalies are introduced by weighted amplification terms to form probabilistically enhanced responses for spatial units that already possess the structural basis of a mineralization system.

[0072] The aforementioned fusion model is based on the thermodynamic equilibrium conditions and reaction kinetic constraints of the open system of mineralization-alteration. It transforms the deterministic mapping relationship between mineral structure evolution and spectral response into a combination of weighted power exponents and enhancement coefficients, so that the mineralization process mechanism is directly introduced into the construction rules of the mineralization potential index. This achieves spatial probability inversion driven by mechanism, rather than exponential superposition based on simple empirical correlation.

[0073] S4.3, Spatial generation of mineralization potential index; The MPI model is applied to the set of all effective pixels Ω after gating, and the comprehensive mineralization potential index value is calculated pixel by pixel: MPI(x). x∈Ω; thus obtaining a continuous mineralization potential distribution field covering the target mining area.

[0074] Wherein Ω represents the effective set of computational spatial pixels formed after multi-source data registration, index normalization processing and gating screening, which is used to limit the scope of the calculation object of the mineralization potential index.

[0075] The resulting MPI field is used to: quantitatively characterize the magnitude and classification of cobalt mineralization probability in each spatial unit; identify the continuous spatial orientation and accumulation core regions of potential mineralization systems; and provide a core numerical basis for subsequent uncertainty assessment and target delineation.

[0076] S5, Construction of the Uncertainty Index (U); After completing the spatial inversion of the S4 Integrated Metallogenic Potential Index (MPI), to simultaneously assess the spatial stability and reliability of the prediction results, this invention further constructs a metallogenic uncertainty index U, which is used to quantitatively characterize the degree of consistency deviation and local variation characteristics among the responses of each metallogenic domain. (See [link to relevant documentation]). Figure 5 .

[0077] S5.1 Calculation of inter-domain response separation; For any pixel x within the effective pixel set Ω, select a background domain index for material supply constraints. Transport domain indices constrained by fluid transport and mineralization reaction-constrained ore-bearing zone index Construct a three-domain response set, as shown in the following expression: The inter-domain extremum separation degree is calculated as follows: in: D(x) This indicates the relative degree of separation between the three domain responses; ε A very small positive stabilizing factor is introduced to prevent the denominator from being zero.

[0078] S5.2 Calculation of local locality; To characterize the local uncertainty caused by spatial noise, mixing effects, and inconsistencies in observation scales, the local locality of each domain index is further calculated within its spatial neighborhood scale, as shown in the following expression: in: N(x) Indicates the current pixel x A local spatial neighborhood window constructed around the center; Var[ ] represents the variance operation of a local cell set; V(x) Used to characterize the level of local disturbance in the spatial continuity and stability of indicators in each domain.

[0079] S5.3, Comprehensive construction of the uncertainty index U; Inter-domain response separation D(x) Local differences V(x) After standardization and weighted fusion, an uncertainty index is constructed, expressed as follows: in: For inter-domain response separation weighting coefficients, For local variance weighting coefficients, satisfying + =1; Z is the normalization operator, which maps the values ​​of U(x) uniformly to the interval [0,1].

[0080] S5.4 Interpretation of uncertainty criteria; The constructed uncertainty index U(x) reflects the consistency and stability of the mineralization indicator under physical prior constraints. Its physical meaning is defined as follows: when the responses of the background domain, transport domain and ore-bearing domain tend to be consistent in space, the synergy is enhanced, and the local local variance is small, U(x)→0, indicating that the coupling relationship of the mineralization system in this region is complete, and the prediction results are stable and highly reliable. When the three-domain responses show significant spatial separation, or when local noise and mixing effects lead to an increase in the spatial dispersion of the domain index, U(x)→1, indicating that the constraints of the mineralization mechanism in this region are weakened, the prediction results are unstable, and the confidence level is reduced.

[0081] S5.5 Spatial Representation of Multi-Domain Closure Constraints; By introducing the uncertainty index U, this invention transforms the degree of synergy among multiple process elements of the mineralization system, namely "source-channel-reaction in situ," into a measurable numerical variable, thereby enabling a spatial characterization of the closed state of the mineralization process. This upgrades the comprehensive mineralization potential spatial field from a single intensity expression to a unified expression mode of "potential-confidence."

[0082] Thus, MPI and U form a coupled criterion space field, achieving the following functions: MPI characterizes the magnitude of mineralization intensity; U characterizes the degree of satisfaction of mechanistic constraints and the reliability of prediction results. The joint distribution of the two provides dual quantitative basis for subsequent selection and classification evaluation of mineralization target areas.

[0083] S6. Joint identification and classification of mineralized target areas; After obtaining the comprehensive mineralization potential index field MPI(x) and the uncertainty index field U(x), this invention constructs a two-dimensional discriminant space based on their joint distribution relationship to achieve the hierarchical division and delineation of mineralization target areas in mineralized clusters. Specifically, the steps include: S6.1, MPI–U Joint Discriminant Space Construction; Map each effective cell x∈Ω to a two-dimensional space consisting of the mineralization potential index and the uncertainty index: Wherein: MPI(x) represents the overall mineralization potential intensity of the pixel; U(x) represents the degree to which the mineralization mechanism constraints of the pixel are satisfied and the uncertainty level of the prediction results.

[0084] The joint discrimination space reflects the comprehensive state of each pixel under the dual scales of "mineralization indication intensity - mechanism consistency", serving as the direct criterion for target classification and delineation.

[0085] S6.2 Threshold classification and target classification rules; Based on the statistical quantile characteristics or empirical thresholds of MPI and U, a mineralization potential threshold T is set. MPI With uncertainty threshold T U : T MPI High value discrimination threshold of MPI index (preferably the upper percentile threshold, such as 70%–85%). T UThe low confidence threshold of the U index (preferably the lower percentile threshold, such as 25%–40% percentile).

[0086] Based on the above thresholds, target classification and discrimination rules are constructed as follows: (1) Level I preferred target area: Identify stable prospecting target areas with strong synergy of multi-domain indicators, high degree of mechanism closure and concentrated mineralization potential, and deploy them as priority exploration and engineering verification areas.

[0087] (2) Level II potential anomaly zone: characterized by outstanding mineralization potential but insufficient multi-domain constraints. It may be affected by local structural complexity, alteration superposition or scale mixing effects, and is a potential area that needs further fine verification or supplementary observation.

[0088] (3) Level III background or non-metallogenic area: The overall mineralization indicator factors are weakened and the area does not meet the system mineralization conditions. It is classified as a low priority or non-target area.

[0089] S6.3 Spatial delineation and vectorized representation of ore-forming target areas; like Figure 6 As shown, the pixel-level target area classification results are clustered and morphologically connected in a spatial grid. Continuous mineralized unit patches are extracted through connected component analysis to achieve regionalized representation of the target area contour.

[0090] The extracted target area polygons should meet the following requirements: the area scale reaches the preset minimum mineral exploration unit threshold; the morphological structure is continuous; and the polygons have spatial suitability with the high-value principal axes of the background and channel domains. After processing, a vector map of the mineralized target area is generated, which is used to support subsequent ground verification, survey line layout, drilling deployment, and comprehensive mineral exploration decision-making.

[0091] S6.4 Target area results verification and parameter optimization; Spatial overlay analysis was performed on the target area results, existing mineralization points, geochemical anomaly sites, and measured cobalt enrichment data to assess the spatial agreement between the target area and known mineralization information. Weighting parameters were adjusted based on the validation results. , , and and threshold The MPI-U discrimination system is modified and optimized to maintain its stability, adaptability and portability under different mineralization areas or different mineralization types.

[0092] S6.5, Final mineral exploration decision output; Through the aforementioned joint judgment mechanism, this invention ultimately achieves: the transformation from multi-domain indicators to a continuous potential field (S4); the transformation from a potential field to a credibility constraint assessment (S5); and the transformation from continuous field expression to engineering target area delineation (S6). This forms a closed-loop mineralization prediction and mineral exploration decision-making process comprised of mechanism-driven, spatially quantified, hierarchical output, and verifiable verification.

[0093] like Figure 7 As shown, this system adopts a layered architecture design, forming an overall technical architecture consisting of a user interface and external system layer, a system management and operation support layer, a four-domain mineralization index and field construction layer, and a mineralization prediction model calculation layer. This enables fully automated and quantitative analysis from multi-source data input to mineralization target area output. Specifically: (1) Calculation layer of mineralization prediction model The mineralization prediction model calculation layer is located at the bottom layer of the system and is the core calculation unit, corresponding to the mineralization potential index calculation module, the uncertainty index calculation module, and related model calculation functions. Based on the normalized multi-domain mineralization indicators, this layer completes the parallel calculation, spatial fusion, and result output of the mineralization potential index (MPI) and the uncertainty index (U), providing quantitative basis for target area determination.

[0094] (2) Metallogenic indices and field structure layers in four regions The four-domain metallogenic index and field construction layer corresponds to a multi-domain indicator distribution construction module, which is used to transform multi-source spatial data and geoscientific information into a metallogenic indicator field with a consistent structure. According to the metallogenic system theory, this layer divides the indicator information into four categories of metallogenic index distributions: background domain, transmission domain, ore-bearing domain, and target domain, providing a unified data representation basis for subsequent model calculations.

[0095] (3) System Management and Operation Support Layer The system management and operation support layer serves as the system's operational assurance layer, encompassing functional modules such as project and engineering management, parameter and model configuration management, user and permission management, and log and result tracking. This layer is responsible for maintaining the system's operational status, configuring model parameters, recording the calculation process, and tracing results, providing support for the system's stable operation and continuous optimization.

[0096] (4) User interface and external system layer The user interface and external system layer supports access from both desktop and web clients and achieves seamless integration with GIS / geology platforms. This layer is used for data loading, result visualization, interactive analysis, and output, supporting cross-platform applications and integration with external systems.

[0097] (5) Inter-layer collaboration and overall system function The system enables collaborative work between different levels through standardized data interfaces, organically integrating data acquisition and preprocessing, multi-domain indicator construction, model calculation and result expression, thereby achieving continuous quantitative prediction of mineralization potential and synchronous assessment of target area credibility.

[0098] Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications or substitutions can be made to the above embodiments without departing from the principles and spirit of the present invention, and the scope of protection of the present invention shall be determined by the appended claims and their equivalents.

Claims

1. A regional mineral exploration method based on the spatial coupling of multiple elements of an ore-forming system, characterized in that, Includes the following steps: (1) Obtain multi-source spatial data covering the target mining area, and perform coordinate unification, calibration, noise suppression and spatial registration on the multi-source spatial data to construct a basic spatial dataset; (2) Based on the process model of the ore-forming system "source-channel-reaction site-abnormal response", the background domain indicator distribution, transport domain indicator distribution, ore-bearing domain indicator distribution and target domain indicator distribution are constructed from the basic spatial dataset respectively; (3) Normalize and continuously smooth the distribution of each domain indicator to generate a standardized index distribution that is comparable within a unified numerical range; (4) According to the preset multi-domain coupling fusion model, the standardized background domain, transmission domain, ore-bearing domain and target domain index distributions are integrated and calculated to generate the comprehensive metallogenic potential index field MPI of the target ore area; (5) Based on the response differences and local spatial variations of the background domain, transmission domain and ore-bearing domain indices at their respective spatial locations, an uncertainty index field U is constructed; (6) Based on the joint discrimination relationship between MPI and U, the target mineral area is graded and evaluated, and the area with high MPI and low U is determined as the preferred mineralization target area, thereby realizing the delineation and ranking of mineralization target areas in the mineral cluster area.

2. The regional mineral exploration method based on the spatial coupling of multiple elements of an ore-forming system according to claim 1, characterized in that, The multi-source spatial data includes at least one of the following: Remote sensing optical or hyperspectral image data, geochemical data, geophysical data, digital elevation model data, and geological logging and structural interpretation data.

3. The regional mineral exploration method based on the spatial coupling of multiple elements of an ore-forming system according to claim 1, characterized in that: The background domain is used to characterize the mineralization supply background and favorable stratigraphic conditions; The transmission domain is used to characterize the fracture structure distribution and fluid transport channel features. The ore-bearing zone is used to reflect the spatial conditions of fluid-wall rock reaction precipitation. The target domain is used to characterize mineralization products and their near-surface anomaly responses.

4. A regional mineral exploration method based on the spatial coupling of multiple elements of an ore-forming system according to claim 1 or 3, characterized in that: The normalization process includes mapping the distribution of each domain indicator to a unified standard numerical range; The continuous smooth transformation includes using a nonlinear continuous mapping function to enhance or compress the index in order to suppress local noise and strengthen the high response zone related to mineralization.

5. The method according to any one of claims 1 to 4, characterized in that, The Comprehensive Metallogenic Potential Index (MPI) is formed by multiplicatively coupling the background domain, transport domain, and ore-bearing domain indices, and introducing the target domain index as an enhancement factor, so that the comprehensive result simultaneously reflects the synergistic constraint effect of multi-domain metallogenic processes.

6. The method according to any one of claims 1 to 5, characterized in that: The uncertainty index U is constructed in the following way: the response separation degree of the background domain, transmission domain and ore-bearing domain indicators at the same spatial location is calculated, the local deviation of each domain indicator is calculated, and the response separation degree and local deviation are weighted and fused to obtain the uncertainty index, which is used to characterize the consistency state and the reliability of the prediction results among the multi-domain mineralization indicators.

7. The method according to any one of claims 1 to 5, characterized in that: The target domain index distribution includes at least the target detection results based on the spectral endmembers of the target mineral, and can be weighted and fused with the iron oxidation anomaly index and the vegetation anomaly index.

8. A regional mineral exploration system based on the spatial coupling of multiple elements of an ore-forming system, characterized in that, include: The data acquisition and preprocessing module is used to perform step (1) of claim 1; the multi-domain indicator distribution construction module is used to perform step (2) of claim 1; the index normalization and smoothing module is used to perform step (3) of claim 1; the mineralization potential index calculation module is used to perform step (4) of claim 1; the uncertainty index calculation module is used to perform step (5) of claim 1; and the mineralization target area determination and result output module is used to perform step (6) of claim 1.

9. The system according to claim 8, characterized in that, The system also includes a model parameter configuration and verification module, which is used to compare the predicted target area with known mineral deposits or abnormal locations, and to adaptively correct the fusion weights and discrimination thresholds.

10. The system according to claim 8 or 9, characterized in that, The system is applicable to mineral exploration prediction in ore clusters of sedimentary, hydrothermal, and sedimentary-hydrothermal superposition cobalt deposits and associated polymetallic deposits. For other types of metal deposits, the index construction methods of the background domain, transport domain, ore-bearing domain, and target domain need to be adjusted according to their metallogenic system mechanism before it can be applied.

Images