Sarn deep mineralized zone positioning method and readable storage medium

Abstract

The invention discloses a skarn deep mineralized zone positioning method, which comprises the steps of generating a contact zone working area and a unique grid reference based on a contact zone geological boundary, a structure line and a common coverage range; based on the hyperspectral data, generating an oxidation thickness map and a demagnetization clue initial map; based on the oxidation thickness map and the demagnetization clue initial map, introducing a physical constraint of near-surface demagnetization caused by strong oxidation, and outputting a demagnetization risk map; constructing a depth bias function according to the demagnetization risk map, and generating joint inversion configuration; carrying out gravity-magnetic joint inversion according to the joint inversion configuration, outputting a density body and a magnetic susceptibility body, and carrying out pairing marking; the continuity of the deep preferential mineralization zone is checked, consistency verification is carried out, and preferential zone labels and low-priority zone labels are output; and the pairing marks, the priority belt marks and the low-priority belt marks are overlapped to form a hole arrangement reference graph, and suggested drilling positions and depth intervals are given in a contact belt working area. According to the scheme, the skarn mineralization zone can be stably positioned, and hole arrangement deployment can be served.

Description

Technical Field

[0001] This invention belongs to the field of geophysical applied geology technology, and in particular relates to a method for locating deep mineralization zones in skarn and a readable storage medium. Background Technology

[0002] In the contact zone between carbonate rocks and intermediate-acidic magmatic rocks, skarn-type metallic mineral deposits are distributed in bands along the strike, with deep mineralization often coexisting with strong weathering and oxidation phenomena in the shallow layers. Airborne magnetics and airborne gravity surveys, due to their wide coverage and sensitivity to magnetic susceptibility and density respectively, have become important tools for deep mineral exploration. Hyperspectral technology can identify hematite, limonite, goethite, and hydrous silicate minerals such as chlorite, clay, and sericite, providing reliable evidence for delineating oxidation and alteration zones. However, strong near-surface oxidation causes demagnetization of ferromagnetic minerals such as magnetite, significantly weakening the response of shallow magnetic anomalies and easily leading to shallow bias in conventional magnetic interpretation, thus masking the true information of deep mineralization bodies. Controlling the geometric morphology of the contact zone and its strike and normal orientation requires data fusion and constraint settings within a unified spatial reference and orientation framework to ensure the reliability of deep-shallow correlation and stable target positioning along the strike.

[0003] Existing technologies typically unfold along two main lines:

[0004] One method is to use inversion or joint inversion based on airborne magnetics and airborne gravity, employing structural coupling techniques such as fixed depth weighting, isotropic or simplified anisotropic smoothing, and cross gradients to recover the magnetic susceptibility volume and density volume, and then indicate favorable regions through anomalous superposition.

[0005] Secondly, hyperspectral techniques are used to identify the mineral characteristics associated with oxidation and skarn, delineating alteration zones and possible mineralization trends. Regarding data fusion, interpolation and overlay at different resolutions and coordinates are often employed, or the working domain is trimmed with a coarse buffer zone, followed by anomaly comparison. To address the effects of demagnetization, some methods employ continuation, polarization, and remanent magnetization correction, or suppress shallow interference through upward continuation and shallow-deep stratification interpretation. However, these are mostly empirical treatments, failing to explicitly introduce the causal relationship between oxidation and demagnetization as an inversion control. Some studies have attempted to use strike preference regularization or qualitative alteration grading for constraints, but lack a mechanism to quantify mineral spectral characteristics, weathering, and topographic factors into location-related controls and integrate them throughout the entire process.

[0006] However, the aforementioned existing technologies have failed to effectively transform the demagnetization caused by strong near-surface oxidation into core control quantities that permeate data fusion, constraint generation, and joint inversion. Specifically, depth weighting and regularization are mostly location-independent or weakly oriented, making it difficult to effectively suppress shallow misleading and enhance deep responses in high-demagnetization-risk areas; anisotropic constraints along the contact zone direction and normal direction are not coupled with oxidation control, easily leading to discontinuous stripes or sharp normal changes; multi-source data lack unified rasterization processing and common coverage pruning, making it difficult to compare deep and shallow layers and express paired labels; hyperspectral alteration information is mostly limited to qualitative delineation, failing to form a mapping chain between physically guided risk fields and subsequent constraints, making it difficult to stably guide the location and pore placement of deep mineralization zones in contact zone scenarios. Summary of the Invention

[0007] The purpose of this invention is to provide a method for locating deep mineralized zones in skarn and a readable storage medium.

[0008] To achieve the above objectives, in a first aspect, the present invention provides a method for locating deep mineralization zones in skarn, comprising the following steps:

[0009] S1. Based on the geological boundary and tectonic line of the contact zone between carbonate rocks and intermediate-acidic intrusive bodies, and the common coverage of airborne magnetic, airborne gravity and hyperspectral data, the contact zone working domain and unique grid reference are generated.

[0010] S2. Within the working domain of the contact zone, using a unique grid reference, extract iron oxide and hydrous mineral indicators based on hyperspectral data, and combine topographic and weathering zone information to generate an oxide thickness map and a preliminary map of demagnetization clues.

[0011] S3. Based on the oxide thickness map and the initial demagnetization clue map, introduce the physical constraint that strong oxidation leads to near-surface demagnetization, perform physical-guided conditional diffusion, and output the demagnetization risk map.

[0012] S4. Based on the demagnetization risk map, construct a position-related depth bias function at each grid point to achieve adaptive depth weighting, and set anisotropic focus constraints along the contact zone direction to generate a joint inversion configuration.

[0013] S5. Within the working domain of the contact zone, using a unique grid reference, perform gravity-magnetic joint inversion of airborne gravity and airborne magnetic methods according to the joint inversion configuration, output density volume and magnetic susceptibility volume, and pair and label deep preferential mineralization zones and shallow demagnetization suspected areas under the unique grid reference.

[0014] S6. Check the continuity of the deep priority mineralization zone along the contact zone direction on the paired labels, and verify the consistency with the spatial variation of skarn-related mineral indicators, and output the priority zone label and the low priority zone label.

[0015] S7. Overlay the paired labels with the priority zone labels and low priority zone labels to form a borehole layout reference map, and provide suggested borehole locations and depth ranges within the contact zone working domain, outputting the results for drilling deployment.

[0016] In one embodiment, S1 generates a contact zone working domain and a unique raster reference based on the geological boundary and tectonic line of the contact zone between carbonate rocks and intermediate-acidic intrusive bodies, and the combined coverage of airborne magnetic, airborne gravity, and hyperspectral data, including:

[0017] Obtain the geological boundaries and tectonic lines of the contact zone;

[0018] Select common coverage area;

[0019] Using the geological boundary of the contact zone as the centerline, the strike and normal direction are determined based on the structural lines.

[0020] The initial domain is formed by setting the width of the buffer zone along the direction of travel and the normal direction, respectively.

[0021] The initial domain is trimmed according to the common coverage area to obtain the contact zone working domain;

[0022] The grid outline is defined by the outer boundary of the common coverage area, and the specification resolution is determined according to the spatial sampling interval of airborne magnetic, airborne gravity and hyperspectral data.

[0023] A unique grid reference is generated based on the grid outline and specification resolution.

[0024] In one embodiment, S2 uses a unique grid reference within the contact zone working domain to extract iron oxide and hydrous mineral indicators based on hyperspectral data, and combines this with topographic and weathering zone information to generate an oxide thickness map and a preliminary demagnetization clue map, including:

[0025] Using the contact zone working domain and the unique grid reference as boundaries, the ratio of iron oxide absorption depth indicator and the ratio of water-bearing mineral absorption depth indicator are calculated at each grid point within the contact zone working domain to form a mineral indicator grid layer.

[0026] The topographic slope, erosion morphology and weathering layer thickness level in the weathering zone information are read from the same grid point in the contact zone working domain to form a weathering constraint grid layer.

[0027] The mineral indicator raster layer and the weathering constraint raster layer are superimposed under orientation and normal constraints to output a set of grid features.

[0028] At each grid point in the grid feature set, the oxide layer thickness is estimated based on the correspondence between iron oxide and water-bearing mineral indicators and weathering zone information; the thickness variation is smoothed along the strike and normal to conform to the geometric continuity of the contact zone; the thickness value or thickness level is written on the unique grid reference, and the oxide thickness map is output.

[0029] At each grid point in the grid feature set, the demagnetization trend value is calculated based on the mapping relationship between oxide thickness and magnetic susceptibility decay. The trend continuity is maintained along the contact zone direction, non-causal abrupt changes are suppressed in the normal direction, and the trend value or trend level is written on the unique grid reference to output the initial demagnetization clue map.

[0030] In one embodiment, S3, based on the oxide thickness map and the initial demagnetization clue map, introduces a physical constraint that strong oxidation leads to near-surface demagnetization, performs physically guided conditional diffusion, and outputs a demagnetization risk map, including:

[0031] A diffusion coefficient field is generated at each grid point using the oxide thickness map to increase the diffusion coefficient at grid points with large oxide thickness and decrease the diffusion coefficient at grid points with small oxide thickness. A larger directional weight is assigned to the directional direction and a smaller directional weight is assigned to the normal direction to conform to the geometric extension of the contact zone.

[0032] The demagnetization clue initial map generates source terms at each grid point, so that grid points with obvious demagnetization trend are risk supply points, and grid points without trend are non-supply points.

[0033] Zero flux is used as a boundary constraint for the contact zone working area to limit risks from crossing the working area boundary.

[0034] Iterative propagation is performed under the influence of the diffusion coefficient field and the source term until a stable risk distribution field is formed on a unique grid reference, thus obtaining the demagnetization risk map.

[0035] In one embodiment, S4 constructs a location-dependent depth bias function at each grid point based on the demagnetization risk map to achieve adaptive depth weighting, and sets anisotropic focus constraints along the contact zone direction to generate a joint inversion configuration, including:

[0036] The risk value of each grid point is read on a unique grid reference, and the risk value is mapped to a bias coefficient. The mapping relationship is a monotonically increasing mapping, so that high risk corresponds to a stronger deep bias.

[0037] At each grid point, the depth weight is determined based on the bias coefficient and the depth layer. The rate of increase of the depth weight with increasing depth is controlled by the bias coefficient.

[0038] The depth weight field is determined by matching the depth weight with the kernel function depth sensitivity of airborne magnetics and airborne gravity within the working domain of the contact zone.

[0039] Based on the bias coefficient, determine the orientation direction weight and normal direction weight of the anisotropic focus constraint;

[0040] The depth weight field, directional weight, and normal weight are organized into a joint inversion configuration.

[0041] In one embodiment, S5 performs gravity-magnetic joint inversion of airborne gravity and airborne magnetic methods within the contact zone working domain, using a unique grid reference, based on the joint inversion configuration, outputting density volume and magnetic susceptibility volume, and pairing and labeling deep preferential mineralization zones and shallow demagnetization suspected areas under the unique grid reference, including:

[0042] Forward modeling operators for airborne gravity and airborne magnetics are established on a unique grid reference. The depth weight field and anisotropic focus constraints in the joint inversion configuration are written into the inversion target control quantity, so that the orientation constraint with focused orientation and stable normal direction and the position-related depth offset are effective throughout the entire working domain. The sensitivity of airborne gravity to density volume and airborne magnetics to magnetic susceptibility volume are jointly driven by updating at the same grid point, so that the two types of data mutually constrain each other in the same spatial cell and avoid the misleading effect of shallow demagnetization on magnetic susceptibility volume. Under the above configuration and data-driven approach, the solution is iteratively solved until stable, and the density volume and magnetic susceptibility volume expressed on the unique grid reference are output.

[0043] Within the working domain of the contact zone, the depth layer set of each horizontal grid point is scanned. If the deep layer simultaneously satisfies the high density of the density volume and the medium-high magnetic susceptibility of the magnetic susceptibility volume, and is continuous along the strike and stable in the normal direction, then a deep preferential mineralization zone label is written for the deep layer of that grid point. At the same time, in the shallow layer of the same grid point, if the demagnetization risk map is of a high level and the magnetic susceptibility volume is of a low value, then a shallow demagnetization suspected area label is written. The above two types of labels are output in pairs on a unique grid reference.

[0044] In one embodiment, S6 checks the continuity of the deep preferential mineralization zone along the contact zone direction on the paired annotations and verifies the consistency with the spatial variation of skarn-related mineral indicators, outputting the preferential and low-priority zone annotations, including:

[0045] Using paired labels as input on a unique raster datum, select a set of horizontal grid points containing labels for deep preferential mineralization zones, and read the mineral indicator intensity of the corresponding grid points related to skarn.

[0046] Along the contact zone, adjacent horizontal grid points within the horizontal grid point set are grouped to establish a direction segment set;

[0047] Based on the mineral indicator strength associated with skarn, the consistency score of each strike segment in the strike segment set is calculated. The judgment results are divided into consistent and inconsistent based on the consistency score, and the judgment results are written into the consistency verification layer on the unique grid reference.

[0048] Based on the consistency verification layer, priority allocation is performed on the deep preferential mineralization zone and it is recorded in parallel with the shallow demagnetization suspected area, and the priority zone label and low priority zone label are output.

[0049] In one embodiment, the determination result is divided into consistent and inconsistent based on the consistency score, including:

[0050] If the consistency score is greater than or equal to the consistency threshold, the corresponding path segment is judged as consistent; otherwise, it is judged as inconsistent.

[0051] In one embodiment, based on the consistency verification layer, priority allocation is performed on the deep preferential mineralization zone and it is recorded in parallel with the shallow demagnetization suspected area. Priority zone labels and low priority zone labels are output, including:

[0052] Within the working domain of the contact zone, the depth layer set of each horizontal grid point is determined. If the strike segment to which the grid point belongs is determined to be consistent in the consistency verification, a priority zone label is written for the deep layer of the grid point, while the shallow demagnetization suspected area label of the shallow layer of the same grid point is retained. If the strike segment to which the grid point belongs is determined to be inconsistent or does not reach the mineral indicator threshold in the consistency verification, a low priority zone label is written for the deep layer of the grid point, while the shallow demagnetization suspected area label of the shallow layer of the same grid point is retained.

[0053] Based on the priority band label and the low priority band label, determine the priority band annotation and the low priority band annotation respectively, and write the priority band annotation and the low priority band annotation on the unique grid reference into the label layer.

[0054] Secondly, this application provides a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the skarn deep mineralization zone location method as described in the first aspect.

[0055] As can be seen from the technical solutions provided in the embodiments of this specification above, this solution: within the unified grid reference and working domain of the carbonate rock-intermediate-acidic magmatic rock contact zone, explicitly transforms the demagnetization genetic information caused by strong near-surface oxidation into location-related depth dip and orientation constraints along strike / normal. This enables gravity-magnetic joint inversion to suppress shallow bias, focus along strike, and be stable along normal at locations with high demagnetization risk. Furthermore, it pairs deep preferential mineralization zones and shallow suspected demagnetization areas at the same grid point, thereby stably locating skarn mineralization zones and serving borehole deployment. Attached Figure Description

[0056] Figure 1 A flowchart illustrating the method for locating deep mineralized zones in skarn provided in this application. Detailed Implementation

[0057] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings, but it should be understood that the scope of protection of the present invention is not limited to the specific embodiments.

[0058] Unless otherwise expressly stated, throughout the specification and claims, the term "comprising" or its variations such as "including" or "comprises" shall be understood to include the stated elements or components without excluding other elements or other components.

[0059] Reference Figure 1 This document illustrates a flowchart of the method for locating deep skarn mineralization zones applicable to the embodiments of this application. The core technical problem this application aims to solve is: within a unified grid reference and working domain of the carbonate-intermediate-acidic magmatic rock contact zone, how to explicitly transform near-surface strong oxidation-induced demagnetization genetic information into location-related depth dip and strike / normal directional constraints. This allows gravity-magnetic joint inversion to suppress shallow bias, focus along the strike, and maintain normal stability at high demagnetization risk locations. Furthermore, it pairs deep-preferred mineralization zones and shallow-demagnetized suspected areas at the same grid point, thereby stably locating skarn mineralization zones and facilitating borehole deployment.

[0060] like Figure 1 As shown, the method for locating deep mineralization zones in skarn may include:

[0061] S1. Based on the geological boundary and tectonic line of the contact zone between carbonate rocks and intermediate-acidic intrusive bodies, and the combined coverage of airborne magnetic, airborne gravity, and hyperspectral data, a working domain and a unique raster reference for the contact zone are generated.

[0062] Specifically, the common coverage area is the spatial intersection area of ​​the three types of data—airborne magnetics, airborne gravity, and hyperspectral—in a unified coordinate system. This is used to eliminate missing and non-overlapping areas of the three types of data, so that the contact zone working domain and the unique grid reference are established on a consistent spatial boundary, and serve as the boundary constraint between the contact zone working domain and the unique grid reference.

[0063] The contact zone working domain is a calculation and interpretation area constructed along the geological boundary and tectonic line of the contact zone between the carbonate host rock and the intermediate-acidic intrusive body. The generation method is as follows: obtain the geological boundary and tectonic line of the contact zone; select a common coverage area; use the geological boundary of the contact zone as the center line, and determine the strike and normal directions based on the tectonic line; set buffer zone widths along the strike and normal directions to form an initial domain; trim the initial domain according to the common coverage area to obtain the contact zone working domain. The contact zone working domain consists of boundary curve set strike parameters, normal parameters, buffer zone width, and a trimming mask. It spatially limits the generation range of the hyperspectral extraction oxidation thickness estimation and demagnetization clue initial map, ensuring that subsequent demagnetization risk field and depth weighting are implemented along the predetermined strike and normal directions. The output of the contact zone working domain serves as the input of S2, enabling S2 to perform feature extraction and clue construction within a region consistent with the mineralization control geometry. From the establishment stage, spatial geometric constraints are applied to address the shallow bias caused by near-surface demagnetization.

[0064] The unique grid reference is a regular grid generated within the contact zone working domain, used to uniformly represent airborne magnetics, airborne gravity, and hyperspectral data, as well as subsequent results. The generation method is as follows: the outer boundary of the common coverage area is used as the grid outline; the specification resolution is determined according to the spatial sampling interval of the three types of data (airborne magnetics, airborne gravity, and hyperspectral); based on the grid outline and specification resolution, a regular grid with a unified origin, unified grid resolution, and unified coordinate system is generated. Within the contact zone working domain, airborne magnetics, airborne gravity, and hyperspectral data are indexed and mapped using the grid, ensuring that the three types of data have consistent spatial positioning and recording structure at the same grid point. The unique grid reference consists of the grid origin, grid resolution, grid coordinate system, and grid mask. Its output serves as the input for S2, ensuring that the oxidation thickness map and the initial demagnetization clue map are expressed at the same grid point, and providing a unified spatial reference for the inputs and outputs of S3 to S7, allowing deep preferential mineralization zones and shallow suspected demagnetization areas to be paired and compared under the same grid reference.

[0065] For unified modeling and data organization of the contact strip working domain, the contact strip working domain is first determined in S1 and a unique grid reference is established to ensure that the results of subsequent steps can be directly superimposed and called under the same spatial reference.

[0066] The following explains the characters that will be used subsequently:

[0067] Define the working domain of the contact strip as , This represents a three-dimensional spatial subdomain centered on the contact zone, covering the area affected by the target mineralization and demagnetization.

[0068] Define a unique raster reference as , Indicates in An internal set of raster coordinates used for unified storage and computation;

[0069] Introducing spatial position vector as , Indicates in A three-dimensional coordinate point on the surface consists of horizontal coordinates and depth coordinates;

[0070] Define the horizontal index as With depth index , , , They represent Grid point numbering in three directions;

[0071] Define the grid spacing as , , , They represent Grid spacing in three directions;

[0072] Define the boundary mask as , express The internal and external binary constraints, with a value of 1 indicating that the grid point belongs to... A value of 0 indicates that the grid point is located at... outside;

[0073] Define the depth layer set as , Indicates in The set of discrete depth level indexes used for hierarchical representation.

[0074] After establishing a unified spatial reference, S1 combines airborne magnetic and airborne gravity observation data, contact zone geometric constraints, topographic data, and hyperspectral data. The system is organized to ensure that subsequent physical constraints are complete with the input of the joint inversion and aligned at the same grid point.

[0075] Define airborne magnetic data as , Indicates in Airborne magnetic observations or derived anomalies at each grid point;

[0076] Define airborne gravity data as , Indicates in Airborne gravity observations or derived anomalies at each grid point;

[0077] Define the contact zone geometric constraints as follows , This represents a set of geometric information about the contact strip, used in... Internally generated orientation direction and normal direction field;

[0078] Define terrain data as , Indicates in The terrain elevation and terrain derivation of each grid point;

[0079] Define hyperspectral data as , Indicates in The hyperspectral reflectance information of each grid point as a function of wavelength, wherein This represents the wavelength variable in the hyperspectral spectrum.

[0080] To serve the subsequent anisotropic constraints of focused direction and stable normal direction, S1 in Internal contact zone geometric constraints Generate a direction field and write it using a unique raster reference. Specifically, construct the orientation direction field as follows: , The contact zone direction at each grid point is represented by a unit vector; the normal direction field is... , Indicates the relationship between each grid point and The corresponding normal direction unit vector, and at each grid point, the two are orthogonal to each other; the above and exist Complete record on the grid points.

[0081] To address the sources of information regarding the causes of near-surface demagnetization due to strong oxidation, a weathering zone information framework was constructed. , This represents the weathering-related parameter vector at each grid point; In the middle, the thickness level of the weathered layer is constructed as follows: , This represents the level of shallow weathering thickness at each grid point; the terrain slope is defined as... , This represents the slope index at each grid point; the erosion morphology is defined as... , This indicates the erosion morphology classification or indicator at each grid point; the surface water system indicator is defined as... , This indicates the presence and impact of surface water systems at each grid point. (The above...) Information about the style of life Internal and hyperspectral data Together, they are used by S2 to extract iron oxide and hydrous mineral indicators from hyperspectral data and combine them with topographic and weathering zone information to output an oxide thickness map and a preliminary map of demagnetization clues.

[0082] S1 in The above clearly adopts As a boundary mask, to contact the working domain For a unique range, and in depth layer set The expression of the shallow and deep layers is divided into layers. Used to implement boundary constraints (including zero flux constraints in S3) in subsequent steps. Used to unify the weight allocation and anomaly interpretation of depth layers in the adaptive depth weighting of S4 and the gravity-magnetic joint inversion of S5.

[0083] It can be understood that demagnetization caused by strong near-surface oxidation can be misleading in the shallow expression of the magnetization volume, in S1 and above. , , , , and Based on this, it ensures that subsequent physical constraints based on the oxide thickness map and the initial demagnetization clue map can act within the same grid point, thereby providing a consistent spatial and directional control framework for the physical guidance condition diffusion of S3, the adaptive depth weighting and anisotropic focus constraint of S4, and the gravity-magnetic joint inversion of S5.

[0084] S2. Within the working domain of the contact zone, using a unique grid reference, extract iron oxide and hydrous mineral indicators based on hyperspectral data, and combine topographic and weathering zone information to generate an oxide thickness map and a preliminary map of demagnetization clues.

[0085] Specifically, the indicators of iron oxides and hydrous minerals are the mineral absorption features and band ratios that can be identified in hyperspectral imaging, including the absorption features of hematite, limonite, goethite, and hydrous silicates such as chlorite, clay, and sericite, which are used to characterize the spatial variation of oxidation and hydrous states within the working domain of the contact zone.

[0086] Weathering zone information refers to the spatial distribution and intensity parameters of the weathered layer within the working area of ​​the contact zone, including weathered layer thickness grade, topographic slope, erosion morphology and surface water system indication, which are used to constrain the spatial range and intensity of the oxidation process.

[0087] The oxide thickness map is a grid map expressed on a unique grid reference, recording the thickness or thickness level of the shallow oxide layer by grid point, and is used to quantify the spatial distribution of oxide thickness within the working domain of the contact zone.

[0088] The initial demagnetization clue map reflecting the weakening trend of near-surface magnetization is a raster map expressed on a unique raster reference. The shallow demagnetization trend value or trend level is recorded by grid point to indicate the spatial trend of the decrease in near-surface magnetic susceptibility and to be used as input for S3.

[0089] Based on the extraction of iron oxide and hydrous mineral indicators from hyperspectral data, and combined with topographic and weathering zone information, an integrated construction process is constructed. Specifically, with the contact zone working domain and a unique grid reference as the boundary, the ratio of iron oxide absorption depth indicator to hydrous mineral absorption depth indicator is calculated at each grid point within the contact zone working domain to form a mineral indicator grid layer; at the same grid point within the contact zone working domain, the topographic slope, erosion morphology, and weathering layer thickness grade in the weathering zone information are read to form a weathering constraint grid layer; the mineral indicator grid layer and the weathering constraint grid layer are superimposed under orientation and normal constraints to output a grid point feature set.

[0090] The oxide thickness map is a raster result generated based on the grid feature set and the geometric constraints of the contact zone working domain. Specifically, it includes: estimating the oxide layer thickness at each grid point in the grid feature set based on the correspondence between iron oxide and water-bearing mineral indicators and weathering zone information; smoothing the thickness variation along the strike and normal directions to conform to the geometric continuity of the contact zone; writing the thickness value or thickness level on a unique raster reference; and outputting the oxide thickness map.

[0091] The initial demagnetization clue map reflecting the near-surface magnetization weakening trend is a raster result generated based solely on the oxidation thickness map and combined with the physical relationship between the skarn mineralization lithology and the influence of oxidation on magnetic susceptibility. Specifically, it includes: calculating the demagnetization trend value at each grid point in the grid feature set based on the mapping relationship between oxidation thickness and magnetic susceptibility attenuation, maintaining trend continuity along the contact zone direction, suppressing non-genetic abrupt changes in the normal direction, writing the trend value or trend level on the unique raster reference, and outputting the initial demagnetization clue map.

[0092] This step S2 follows the definition of S1 in terms of space and indexing system.

[0093] exist Within the grid range, taking the typical spectral band sets of iron oxides and hydrated silicates as the extraction objects, the ternary absorption band set of iron oxides is defined as... Each element Corresponding shoulder wavelength Center wavelength The other shoulder wavelength Weights of linear continuous terms .

[0094] Define the set of iron oxide band ratio pairs as Each element Corresponding ratio molecular wavelength Wavelength of the ratio denominator Accordingly, the set of ternary absorption bands of hydrous minerals is defined as follows: The set of spectral band ratio pairs is At each grid point Normalized absorption depth Ratio to spectral band These serve as basic indicators for iron oxides and hydrous minerals, respectively. This indicates the magnitude of the absorption dip relative to the center wavelength. This indicates the relative intensity relationship between two spectral bands.

[0095] The above indicators are combined with weathering zone information to generate weathering constraint weights. It is used to limit oxidation to occur within the weathering control range.

[0096] To further demonstrate the geometric continuity of the contact zone and suppress non-genetic sharpness in the normal direction, S2 in Anisotropic neighborhood sets using orientation and normal. and along the direction distance and distance along the normal direction Measuring the relative directional position within the neighborhood, where , .

[0097] Based on this, the oxide thickness map Calculate and write the unique raster reference at each grid point using the following formula:

[0098]

[0099] in, Indicates at grid points The thickness or thickness grade of the oxide layer at that location; Represents scalar The function that performs upper and lower limit clipping returns... , and These represent the lower and upper limits of the oxide thickness, respectively; A reference term indicating the oxide thickness; , , , This represents the global weights of the four types of indicators; and These represent the band coefficients of the iron oxide absorption depth term and the band ratio term, respectively. and These represent the spectral coefficients of the absorption depth term and the spectral band ratio term for hydrous minerals, respectively. Spectral bands At grid points Normalized absorption depth at the location; For spectral band ratio pairs At grid points The ratio at the point; As the weathering constraint weight, it is composed of Generate and use it to modulate indicator pairs within the weathering control range Contributions; Integrate weights for the neighborhood; and These are the positive and negative modulation coefficients in the directional and normal directions, respectively, used to enhance continuity in the directional direction and suppress non-genetic sharpness in the normal direction; For distance measurement along the direction; This is a distance metric along the normal direction; and These are the correlation scales for the direction and the normal direction, respectively; Represents an exponential function; For Centered on, along and The set of neighborhood grid points for directional sampling.

[0100] In the above calculation framework, iron oxide indicators and hydrous mineral indicators are obtained through... The absorption depth term and the band ratio term are directly converted into a quantitative contribution to the oxide thickness, and by Classification of weathering layer thickness Topographic slope erosion morphology With surface water system indicators Integrate into location-related constraint weights to ensure It only enhances or suppresses weathering within the controlled area. The anisotropic neighborhood term is... and For directional reference, via , and , Quantitative control is exercised over the continuity of the directional direction and the constraint of the normal direction to ensure that The contact zone is continuous in a band-like pattern in the direction of contact and does not exhibit non-causal sharp changes in the normal direction.

[0101] The above formula can solve the following problem: Under the background of demagnetization caused by strong near-surface oxidation, it is necessary to quantitatively estimate the oxide layer thickness using mineral spectral indicators and weathering constraints within the same grid point, and ensure the rationality of the thickness field's origin under the geometric constraints of the contact zone, thereby providing a reliable thickness input for subsequent physical constraints and joint inversion.

[0102] In generating an oxide thickness map Afterwards, S2 with Based solely on and combined with the physical relationship between skarn mineralization lithology and the effect of oxidation on magnetic susceptibility, a preliminary demagnetization clue map reflecting the weakening trend of near-surface magnetization was generated. .in, Indicates at grid points The demagnetization trend value or trend level at the location; the skarn mineralization lithology sensitivity coefficient is defined as... , Indicates at grid points The sensitivity of magnetic susceptibility to oxidation corresponding to the mineralizing lithology of skarn; a monotonically increasing mapping relationship. Will according to Weighted conversion This results in a more pronounced demagnetization trend at lattice points with thicker oxide layers.

[0103] S3. Based on the oxide thickness map and the initial demagnetization clue map, introduce the physical constraint that strong oxidation leads to near-surface demagnetization, perform physical-guided diffusion, and output the demagnetization risk map.

[0104] Specifically, the physical constraint introduced for strong oxidation leading to near-surface demagnetization is to use a set of rules to limit diffusion by establishing the causal relationship between the oxidation thickness map and the initial demagnetization clue map. This includes a monotonic mapping relationship between oxidation thickness and demagnetization intensity, upper and lower limits of demagnetization intensity, and zero flux constraint at the boundary of the contact zone working domain, to ensure that the risk only propagates within the oxidation control area.

[0105] Physically guided conditional diffusion, within a unique grid reference and contact zone working domain, uses the oxide thickness map and initial demagnetization clue map output by S2 as input. It constructs a diffusion coefficient field and source term according to the physical constraint that strong oxidation leads to near-surface demagnetization, and performs conditional diffusion with variable coefficients at various locations to output a demagnetization risk map. Specifically:

[0106] Source terms are generated at each grid point using the initial demagnetization criterion map, with grid points showing a clear demagnetization trend designated as risk supply points and those without a trend designated as non-supply points; zero flux is used as the boundary constraint of the contact zone working domain to prevent risks from crossing the working domain boundary.

[0107] S31. Read the oxide thickness map and the initial demagnetization clue map, and load the orientation, normal direction field and boundary mask;

[0108] S32. Normalize the oxidation thickness at each grid point to obtain the oxidation control indicator, which is used to limit the diffusion range and amplitude;

[0109] S33. Modulate the diffusion coefficient with oxidation control indicator in the directional and normal directions, that is, construct an anisotropic diffusion coefficient field with enhanced directional diffusion, restricted normal diffusion, and conforming to the contact zone geometry.

[0110] S34. Generate source terms at each grid point using the initial demagnetization clue map, so that grid points with obvious demagnetization trends are risk supply points and grid points without trends are non-supply points; use zero flux as the boundary constraint of the contact zone working domain to restrict risks from crossing the working domain boundary.

[0111] Iterative propagation is performed under the influence of the diffusion coefficient field and the source term until a stable risk distribution field is formed on a unique grid reference, resulting in a demagnetization risk map:

[0112] S35. Conditional diffusion iterative propagation of risk distribution, updating demagnetization risk value on the grid, and then performing convergence judgment: compare the risk change before and after, if the threshold and stability requirements are not met, then iterate; otherwise, converge and execute S36.

[0113] S36. Output a clearly tiered demagnetization risk map, preserving continuous directional movement and restricted normal direction within the working domain.

[0114] The clearly defined demagnetization risk map is a raster result expressed on a unique raster reference, including a risk value layer and a risk level layer. The risk value layer records the demagnetization risk value of each grid point, and the risk level layer classifies the risk value into low, medium and high levels according to a preset threshold. The demagnetization risk map is directly derived from the output of the diffusion of physical guidance conditions, and retains the directional continuity and normal constraint characteristics in the working domain of the contact zone. It is used as the input of S4 to modulate adaptive depth weighting and anisotropic focus constraints.

[0115] This step S3 follows the definitions of S1 and S2 in terms of space and indexing system.

[0116] S3 introduces a physical constraint to explicitly implement strong oxidation leading to near-surface demagnetization, at each grid point... Monotonous return to one Obtain the oxidation control indicator. ,in and These represent the lower and upper limits of the oxide thickness, respectively; in the orientation and normal directions... Modulation diffusion coefficient, defined as the directional diffusion coefficient is With normal diffusion coefficient ,in and These are the reference diffusion coefficients for the directional and normal directions, respectively. and These are the gain coefficients of oxidation control on bidirectional diffusion; with Construct risk source items ,in To establish a source term strength benchmark, the risk is supplied only within the oxidation control range; to limit the ineffective diffusion of risk in non-oxidizing regions, a dissipation coefficient is introduced. Risk is mitigated in non-oxidized regions.

[0117] Based on the above coefficients and source terms, The steady-state solution for conditional diffusion within the lattice range using the anisotropic diffusion tensor is calculated as follows:

[0118]

[0119] in, Indicates at grid points Demagnetization risk value at the location; express Spatial gradient; Represents the divergence operator; and Let represent the second-order tensor bases for the direction of travel and the direction of normality, respectively. and These represent the position-related diffusion coefficients in the azimuth and normal directions, respectively. This indicates a boundary mask used in the contact zone working area. It takes effect within the working domain and applies zero-flux constraints to the work domain boundaries; This represents the risk dissipation coefficient in the non-oxidized region, used to suppress risk propagation across regions; This represents the risk source term determined jointly by the initial demagnetization clue diagram and oxidation control.

[0120] Based on the above equations, an iterative solution is used to reach a steady state, yielding the following results. through Write the unique raster reference after trimming. ,in and These represent the lower and upper limits of the demagnetization risk value, respectively, ensuring that the risk value remains within the effective range.

[0121] Based on the steady-state risk distribution, a demagnetization risk map is output in a clearly tiered manner, comprising two parts: a risk value layer and a risk level layer. The risk value layer directly records... Risk level tiers are based on preset thresholds. and right Divide, when When marked as low level, When marked as medium level, It is marked as high level.

[0122] The two layers mentioned above are Within the grid range, a unique raster reference is used. Record and preserve the continuity along the directional direction and the restricted characteristics along the normal direction.

[0123] The physical constraint of near-surface demagnetization caused by strong oxidation is introduced into the risk propagation model through a composite mechanism of monotonic mapping, anisotropic diffusion, and zero flux boundary, thus making physically guided diffusion subject to... and Control and generate a clearly defined demagnetization risk map that can be directly used for S4.

[0124] S4. Based on the demagnetization risk map, construct a position-related depth bias function at each grid point to achieve adaptive depth weighting, and set anisotropic focus constraints along the contact zone direction to generate a joint inversion configuration.

[0125] Specifically, the location-related depth bias function is constructed by taking the grid risk value of the demagnetization risk map as the independent variable within the unique grid reference and the working domain of the contact zone, and outputting a set of bias coefficients for depth weight calculation. The bias coefficients change monotonically with the risk value, with larger bias coefficients for high risk and reference bias coefficients for low risk.

[0126] Adaptive depth weighting is a depth weight field generated at each grid point and each depth layer based on a location-related depth bias function. The depth weight quantifies the relative weight allocation between shallow and deep layers, which is used to suppress shallow response and improve deep response at high-risk locations, while maintaining normal weight allocation at low-risk locations.

[0127] The joint inversion configuration is a set of inversion control parameters organized on a unique grid reference, including a depth weight field generated by a position-dependent depth bias function, anisotropic focus constraints set along the contact zone direction, and proportional parameters of data weights and structural constraints. The joint inversion configuration is used as input to S5 to perform gravity-magnetic joint inversion.

[0128] Anisotropic focus constraints are set along the contact zone to focus in the direction of the contact zone and stabilize in the normal direction. The focus in the direction of the contact zone and the stability in the normal direction are the target behaviors of the anisotropic focus constraints. That is, focusing constraints are used in the direction of the contact zone to enhance the continuity of the strip, and stabilizing constraints are used in the normal direction of the contact zone to limit non-causal abrupt changes.

[0129] To achieve adaptive depth weighting, a location-dependent depth bias function is constructed to directly transform the demagnetization risk map into a depth tendency control quantity. Specifically, the risk value of each grid point is read on a unique grid reference, and the risk value is mapped to a bias coefficient. The mapping relationship is a monotonically increasing mapping, so that high risk corresponds to a stronger depth bias. At each grid point, the depth weight is determined jointly by the bias coefficient and the depth layer. The rate of increase of the depth weight with increasing depth is controlled by the bias coefficient. At high-risk grid points, the shallow weight is suppressed and the deep weight is increased, while at low-risk grid points, the baseline growth rate is adopted. The above depth weights are matched with the depth sensitivity of the kernel function of airborne magnetism and airborne gravity in the contact zone working domain, and a unified depth weight field is output and written into the joint inversion configuration, so that S5 prioritizes the interpretation of deep information and reduces shallow demagnetization interference at high-risk locations.

[0130] Anisotropic focus constraints are set along the contact zone to ensure stability in the strike direction and focused normal direction. The contact zone geometry is then directly used for inversion regularization orientation control. Specifically, strike and normal direction fields are extracted within the contact zone working domain, and a direction vector is established for each grid point on a unique grid reference. Focusing regularization is introduced in the strike direction to make the anomaly present a compact and continuous shape in the strike direction, highlighting the deep volume distributed along the contact zone. Stability regularization is introduced in the normal direction to limit the violent fluctuations in the normal direction with first-order or second-order smoothing penalties, preventing shallow defects caused by demagnetization from being misinterpreted as causal interfaces in the normal direction. The strike and normal regularization weights are written into the joint inversion configuration at a fixed ratio, and together with the depth weight field generated by the position-related depth bias function, they constitute anisotropic and depth-adaptive composite constraints. The output is used as the input of S5 to guide the gravity-magnetic joint inversion to achieve deep-priority interpretation while focusing in the strike direction and stabilizing in the normal direction.

[0131] The space and indexing system for the S4 parameter in this step follows the definitions of S1, S2, and S3 above.

[0132] This step mainly involves the demagnetization risk map. As the core control quantity, a position-dependent depth bias function is derived and a unified depth weight field is generated. At the same time, anisotropic constraint weights that are focused along the orientation and stationary along the normal are generated. Both are written into the joint inversion configuration for subsequent gravity-magnetic joint inversion.

[0133] To ensure that depth weights explicitly favor deeper interpretations at high-risk locations, in The above constructs a unified depth weight field for each grid point and each depth layer. Simultaneously, it generates the orientation weights of the anisotropic focus constraints. With normal weight The calculation is as follows:

[0134]

[0135] in, Indicates at grid points With depth layer index Uniform depth weight value at the location; This indicates a boundary mask used in the contact zone working area. It takes effect within the domain and restricts weights from being written outside the domain; A baseline term representing depth weights, used to maintain a normal depth-to-shallow distribution in low-risk locations; The depth bias coefficient, representing the location-dependent parameters, is derived from the demagnetization risk map. Obtained by monotonic mapping; Indicates the index of depth layers An increasing linear growth rate coefficient is used to accelerate the deep weighting at high-risk locations. The rate of increase; and Let these represent the sensitivity ratio coefficients of airborne magnetics and airborne gravity in a unified depth weighted field, respectively, satisfying... ; and These represent the indexes at the depth level. The depth-sensitivity normalized spectrum of the kernel function of airborne magnetics and airborne gravity, and The range of values ​​is It also reflects the relative strength of the two types of data response to different depth levels; and These represent the lower and upper limits of the demagnetization risk value, respectively, used to determine... Perform linear normalization; and These represent the lower and upper limits of the location-dependent depth offset coefficient, respectively. and These represent the directional weights and normal weights of the anisotropic focus constraint, respectively. The baseline weights representing anisotropic constraints; This represents a fixed proportionality coefficient between the directional and normal constraints, used to ensure that the focusing penalty in the directional direction is stronger than the smooth penalty in the normal direction.

[0136] The construction of the above weights directly uses As a bias source, high-risk grid points are placed in To gain a stronger deep inclination, while in and Maintaining a fixed proportional relationship between the upward direction and the stable normal direction helps to resolve the misleading interpretation of shallow layers caused by near-surface demagnetization in contact zone scenarios.

[0137] After the above calculations are completed, , and The organization is configured for joint inversion and based on a unique raster benchmark. Write.

[0138] In the joint inversion configuration, the location-dependent depth bias function is determined by... Clear representation, adaptive depth weighting A uniform weight value is given at each grid point and at each depth level; the anisotropic focus constraint is determined by... and With a fixed ratio Record the orientation constraint strength of the direction and normal; simultaneously record the ratio parameters of the data and structure for use in the inversion objective function, specifically the data weights. and (Weights corresponding to the data fitting terms for airborne magnetics and airborne gravity, respectively) and structural weights (Total weights corresponding to anisotropic focus constraints), where , and exist Write uniformly within the grid range and use This is a unique range.

[0139] The aforementioned joint inversion configuration is generated under the same spatial reference and indexing system, with the core being the demagnetization risk map. The depth bias function related to the driving position is oriented and bound to the anisotropic focus constraint, so that the gravity-magnetic joint inversion in the subsequent step S5 can focus on the direction and stabilize the normal direction while implementing a deep-first interpretation of high-risk locations. This effectively reduces the misleading effect of near-surface demagnetization on the shallow expression of the magnetization volume in the contact zone scenario.

[0140] S5. Within the working domain of the contact zone, using a unique grid reference, perform gravity-magnetic joint inversion of airborne gravity and airborne magnetic methods according to the joint inversion configuration, output density volume and magnetic susceptibility volume, and pair and label deep preferential mineralization zones and shallow demagnetization suspected areas under the unique grid reference.

[0141] Specifically, the density volume is a three-dimensional density distribution volume expressed on a unique grid reference. Density values ​​are recorded by grid points and depth layers and are constrained by the boundary mask of the contact zone working domain. It is used to characterize deep density anomalies under the same spatial reference.

[0142] The magnetic susceptibility volume is a three-dimensional magnetic susceptibility distribution expressed on a unique grid reference. Magnetic susceptibility values ​​are recorded by grid points and depth layers and are constrained by the boundary mask of the contact zone working domain. It is used to characterize deep magnetization anomalies under the same spatial reference.

[0143] Paired labeling is a method of simultaneously recording the labeling of deep preferential mineralization zones and shallow suspected demagnetization zones in the set of depth layers corresponding to the same horizontal grid point of a unique raster reference, so that deep and shallow information with the same spatial location can be compared in pairs.

[0144] The deep preferential mineralization zone is a deep anomalous combination zone that is continuous along the strike within the working domain of the contact zone and restricted in the normal direction by density volume and magnetic susceptibility volume.

[0145] The suspected shallow demagnetization zone is the area in the near-surface depth layer of the contact zone working area where the high level of the demagnetization risk map is consistent with the low value of the shallow magnetic susceptibility.

[0146] The gravity-magnetic joint inversion takes the joint inversion configuration output by S4 as input and simultaneously solves for the density volume and magnetic susceptibility volume within a unique grid datum and the contact zone working domain. Specifically, forward modeling operators for airborne gravity and airborne magnetics are established on the unique grid datum. The depth weight field and anisotropic focus constraints from the joint inversion configuration are written into the inversion target control quantity, ensuring that orientation constraints with focused orientation and stable normal orientation, along with position-related depth offsets, take effect throughout the working domain. The sensitivity of the density volume corresponding to airborne gravity and the sensitivity of the magnetic susceptibility volume corresponding to airborne magnetics are jointly driven by updating at the same grid point, so that the two types of data mutually constrain each other within the same spatial cell and avoid the misleading effect of shallow demagnetization on the magnetic susceptibility volume. Under the above configuration and data-driven approach, the solution is iteratively solved until stable, and the density volume and magnetic susceptibility volume expressed on the unique grid datum are output as direct input for paired annotation.

[0147] The pairing of deep preferential mineralization zones and shallow demagnetization suspected areas is achieved by recording the combined anomalies of density volume and magnetic susceptibility volume, along with the shallow trend of the demagnetization risk map, in pairs at the same spatial grid point. Specifically, within the contact zone working domain, the depth layer set of each horizontal grid point is scanned. If the deep layer simultaneously satisfies the high density of the density volume and the medium-to-high magnetic susceptibility of the magnetic susceptibility volume, and is continuous along the strike and stable in the normal direction, then a deep preferential mineralization zone label is written for that grid point. Simultaneously, at the same grid point in the shallow layer, if the demagnetization risk map is of a high level and the magnetic susceptibility volume is of a low value, then a shallow demagnetization suspected area label is written. The above two types of labels are output in pairs on a unique raster reference, forming the input for S6, so that subsequent continuity checks and consistency verifications can directly call the deep and shallow genetic relationship under the same spatial reference.

[0148] The space and indexing system of parameters in step S5 follows the definitions of S1, S2, S3, and S4 above.

[0149] Density volume and magnetic susceptibility volume respectively in The upper part is organized into a three-dimensional field according to grid points and depth layers, which are denoted as follows: and ,in Indicates the density volume at the lattice point With depth layer The value of , Indicates the magnetic susceptibility volume at the lattice point With depth layer The values ​​of both are within Solve and save within the specified range.

[0150] To jointly interpret airborne gravity and airborne magnetism under the same spatial reference and explicitly perform adaptive depth weighting with anisotropic focus constraints that are focused in the direction of orientation and stable in the normal direction, and position-related depth bias, this embodiment solves the following objective function as the core:

[0151]

[0152] in, Represent the objective function; and Represent the forward modeling operators for airborne gravity and airborne magnetism, respectively, with the input being... and That is, only in The density volume and magnetic susceptibility volume within the cell participate in forward modeling, and the output is related to... and exist Top aligned; Represents the L2 norm; The directional derivative in the direction of travel ( , For the horizontal gradient operator, dot product " (where "is the vector dot product") It is the absolute value; This represents the directional derivative in the normal direction. It is a norm 2; Indicates at grid points With depth layer The unified depth weighting at the location has been implemented in step S4 to demagnetize the risk map. Mapped to a position-dependent depth bias function and then... , according to , The result is after combination; and These are the weights of the anisotropic focus constraints in the directional and normal directions, respectively. They are derived from step S4 and satisfy a fixed proportional relationship of focusing in the directional direction and stabilizing in the normal direction. , and These are the proportional parameters for fitting airborne gravity data, airborne magnetic data, and structural constraints, respectively.

[0153] The first and second terms of the above objective function are fitted respectively under the condition that the density volume and the magnetic susceptibility volume are aligned at the same grid point. and The third item is Adaptive weighting of depth layers, in order to and By applying a focusing penalty in the directional direction and a stabilization penalty in the normal direction, the misleading interpretation of shallow magnetic susceptibility by shallow demagnetization is suppressed within the contact zone working domain, while enhancing stable imaging of deep stripes. The optimization problem is solved iteratively until stability is achieved, yielding the optimal solution. and In respectively The above is written as a density volume and a magnetic susceptibility volume, in which This represents the optimized solution.

[0154] After obtaining the density volume and magnetic susceptibility volume, step S5 performs paired labeling under a unique grid reference, recording the deep preferential mineralization zone and the shallow suspected demagnetization zone in pairs, using the same horizontal grid point as the unit.

[0155] Define the deep layer set as With the shallow layer as ,in This represents the set of deep stratigraphic indexes. Represents the set of shallow stratigraphic indices; the density threshold is defined as... The magnetic susceptibility threshold is The demagnetization risk threshold is The shallow magnetic susceptibility threshold is For each horizontal grid point within the contact band working domain. If it exists Make and Then, the deep strata at that grid point are marked with the deep preferential mineralization zone; at the same time, if And exist Make Then, the shallow demagnetization suspected area is marked at the shallow layer level at the same grid point. Both types of labels are linked by the same horizontal grid point. The above is written in pairs to ensure that step S6 can check the strike continuity of the deep preferential mineralization zone under the same spatial reference and verify the consistency with the spatial variation of mineral indicators related to skarn.

[0156] This embodiment uses the joint inversion configuration as the sole control source to generate the position-related depth bias function. With anisotropic focus constraints , By directly inputting the objective function of the inversion, the gravity-magnetic joint inversion achieves focused orientation and stable normal orientation, while prioritizing deep interpretation of high-risk locations indicated by the demagnetization risk map. Density volumes and magnetic susceptibility volumes in... The data is stored within a unique raster reference range and used as input for step S6 together with the paired labels.

[0157] S6. Check the continuity of the deep priority mineralization zone along the contact zone direction on the paired labels, and verify the consistency with the spatial variation of skarn-related mineral indicators, and output the priority zone label and the low priority zone label.

[0158] Specifically, the spatial variation of mineral indicators related to skarn is the spatial distribution of mineral indicator intensity and indicator gradation recorded within the unique grid reference and contact zone working domain. This includes gridded expressions of the absorption characteristics and band ratios of typical skarn minerals such as garnet, diopside, epidote, and pyroxene in hyperspectral data, as well as their variations along strike and normal, to provide mineral clues related to mineralization.

[0159] Consistency verification is a comparative test between the spatial variation of paired labels and skarn-related mineral indicators. It includes strike continuity comparison, normal constraint comparison and threshold consistency determination. It is used to determine whether the deep preferential mineralization zone and mineral indicators are consistent on the same strike strip on a unique raster reference.

[0160] Priority zones and low priority zones are labeled as tag layers expressed on a unique raster reference, containing priority zone tags and low priority zone tags. The tags are recorded on the same horizontal grid point and its depth layer set, on the one hand marking the priority of deep priority mineralization zones, and on the other hand preserving parallel information of shallow demagnetization suspected areas.

[0161] To verify the consistency of spatial variations in skarn-related mineral indicators, the genetic consistency between paired labels and mineral indicators was determined along the strike within the contact zone working domain. Specifically, on a unique raster datum, using paired labels as input, a set of horizontal grid points containing labels for deep-preferred mineralization zones was selected, and the intensity and grading of skarn-related mineral indicators for the corresponding grid points were read. A set of strike segments was established along the contact zone strike, and within each strike segment, the continuity of deep-preferred mineralization zone labels was checked. Within the same strike segment, the mineral indicators were compared to see if they formed a high-intensity zone consistent with the strike, and whether this zone was consistent with the normal direction. The direction is limited by the width of the buffer zone; the threshold consistency judgment constraint requires that the mineral indicator intensity reaches the preset indicator threshold and its peak position coincides or nearly coincides with the center line of the deep priority mineralization zone in the same strike segment, while not crossing the boundary mask of the contact zone working domain in the normal direction; if the above strike continuity, normal constraint and threshold consistency are all satisfied, the strike segment is judged as consistent; if any condition is not satisfied, the strike segment is judged as inconsistent; the consistency and inconsistency judgment results are written into the consistency verification layer on the unique grid reference to support priority division.

[0162] The output of priority and low priority zones is based on the consistency verification results, prioritizing deep priority mineralization zones and recording them in parallel with shallow demagnetization suspected areas. Specifically, within the contact zone working domain, the depth layer set of each horizontal grid point is determined. If the strike segment to which the grid point belongs is determined to be consistent in the consistency verification, a priority zone label is written to the deep layer of the grid point, while the shallow demagnetization suspected area label of the same grid point is retained. If the strike segment to which the grid point belongs is determined to be inconsistent in the consistency verification or does not reach the mineral indicator threshold, a low priority zone label is written to the deep layer of the grid point, while the shallow demagnetization suspected area label of the same grid point is retained to indicate the coexistence of near-surface demagnetization. The above priority and low priority zones are marked on a unique raster reference and output in the form of a label layer, so that the priority zones can be deployed and the low priority zones can be verified under the same spatial reference.

[0163] The S6 space and index system follows the definitions of S1, S2, S3, S4, and S5 mentioned above.

[0164] The input pairing label is specifically defined as follows: Define the label for the deep preferential mineralization zone as... , Indicates at grid points The deep strata are marked with deep preferential mineralization zones. This indicates that it does not exist; the label for the suspected shallow demagnetization area is defined as... , Indicates at grid points The shallow layers show suspected areas of shallow demagnetization. This indicates that it does not exist.

[0165] To further facilitate consistency verification, The above-introduced mineral indicator strength related to skarn is: , Indicates at grid points The mineral indices are obtained by integrating the spectral characteristics of typical skarn minerals (such as garnet, diopside, epidote, and pyroxene). The larger the value, the stronger the indices.

[0166] To measure the neighborhood relationship between the orientation and normal directions, the oriented neighborhood and distance metrics from step S2 are used as follows:

[0167] Define the neighborhood as , Indicated by Centered on, along The set of neighborhood grid points for directional sampling; the normal neighborhood is defined as... , Indicated by Centered on, along The set of neighborhood grid points for directional sampling; the distance metric along the direction is defined as... The distance metric along the normal direction is defined as ,in" "" indicates the vector dot product.

[0168] To match the contact zone geometry, the orientation correlation scale is defined as follows: , The correlation attenuation scale along the direction of orientation is represented; the normal correlation scale is defined as... , This indicates the associated attenuation scale along the normal direction.

[0169] Subsequently Establish a set of directional segments within the grid range , Indicates by along Adjacent directions and A set of directional segments composed of groups of lattice points; each directional segment is denoted as . , It represents a set of lattice points that are continuous in the direction of travel. Represents a set Number of cells.

[0170] The core of consistency verification in this application lies in simultaneously quantifying the continuity of orientation, the constraint of normal direction, and the consistency of mineral indicators.

[0171] Based on this, in each directional segment Calculate the consistency score Its expression is as follows:

[0172]

[0173] in, Indicates the direction of the segment The consistency score indicates that the strike segment meets the genetic requirements in terms of strike continuity, normal constraint, and mineral indication consistency. Indicates the weight of the term moving towards continuity; Indicates the weight of the mineral indicator consistency item; The weight of the normal instability penalty term is indicated; The weight of the conflict penalty term indicating the coexistence of shallow demagnetization and weak mineral indication; Represents an exponential function; This indicates an indicator function that returns 1 if the condition within the brackets is true, and 0 otherwise. The threshold for determining the intensity of mineral indicators. This indicates that the mineral indicator strength has met the preset requirements; Indicates the direction of the segment The offset of the centroid of the internal mineral indicator from the centroid of the deep preferential mineralization zone along the normal direction is defined as follows: ,in Indicates the direction of the segment Inner For the spatial centroid of power, Indicates the direction of the segment Inner For the spatial centroid of power, Indicates the direction of the segment The average unit vector of the inward normal direction; The buffer width scale representing normal uniformity is used for modulation. Impact on ratings.

[0174] In the above formula Going to neighboring areas The Inner Canon Attenuation accumulation quantifies the continuity of deep preferential mineralization zones in the direction of strike; In the normal neighborhood The Inner Canon Attenuation accumulation quantifies the fluctuations of deep preferential mineralization zones in the normal direction and imposes penalties; In the direction of the section Internal average, and with normal offset through Consistency modulation is performed to quantify the spatial overlap between mineral indicators and deep preferential mineralization zones; conflict penalty terms are applied through shallow demagnetization suspected area tags. With mineral indicator threshold This is a combined trigger, used to indicate the coexistence of weakened shallow magnetization caused by strong near-surface oxidation and insufficient mineral indication.

[0175] To further prioritize the implementation, in each directional segment The above determination is based on a consistency score and a threshold, with the consistency threshold defined as follows: ,when If the condition is met, the directional segment is considered consistent; otherwise, it is considered inconsistent.

[0176] Subsequently Within the grid range, label layers are generated in units of directional segments: Preferred labels are defined as... , Representing grid points The strike segments are determined to be consistent and their deep strata exist. ,otherwise Define low-priority tagged characters as , Representing grid points The strike segment is determined to be inconsistent and its deep strata exist. ,otherwise At the same time, retain the label of the suspected shallow demagnetization area. Provide shallow information in parallel. and Based on a unique raster benchmark Write to the tag layer, and with Recording pairs of data at the same horizontal grid point enables subsequent step S7 to directly deploy priority bands and verify low-priority bands under the same spatial reference.

[0177] Furthermore, to provide supplementary explanation, the above implementation is in and The process unfolds within a unified spatial reference, with both the directional and normal directions originating from... and By definition, neighborhood and distance metrics are adopted and The correlation scale adopts and ; Paired labeling and Definition; mineral indicators associated with skarn It expresses and constrains the consistency score.

[0178] This implementation method uses the strike segment as the core calculation point and quantitatively integrates strike continuity, normal constraint, and mineral indicator consistency to solve the pain point of "whether the deep priority mineralization zone and the mineral indicator are consistent on the same strike strip" in the contact zone scenario, thereby stably outputting priority zone and low priority zone labels under the same grid reference.

[0179] S7. Within the contact zone working domain, using a unique grid reference, overlay the paired labels of S5 with the priority zone labels and low priority zone labels of S6 to form a borehole layout reference map for drilling deployment. Within the contact zone working domain, provide suggested borehole locations and depth ranges, and output the results for drilling deployment.

[0180] The S7 space and index system follows the definitions of S1, S2, S3, S4, S5, and S6 mentioned above.

[0181] The following priority rules are recommended for selecting drilling locations:

[0182] exist Within the scope, priority will be given to meeting the requirements. and The horizontal grid points are marked as suggested drilling locations; when but and At that time, mark the suboptimal recommended borehole locations for verification and deployment; label the suspected shallow demagnetization areas. Displayed in parallel on the borehole reference map, this information serves to highlight near-surface demagnetization background and assist in risk management during on-site layout. All recommended borehole locations are based on a unique grid reference. Output the layer above, maintaining the same spatial alignment as the aforementioned layers.

[0183] It is recommended that the drilling depth range be determined based on the deep anomaly of the density volume and the magnetic susceptibility volume. At each candidate drilling point... At this point, select the deep layer set. And use the threshold from step S5. and ,exist Search in the middle satisfies and A continuous segment, denoted by the index of the shallowest layer of that continuous segment. The deepest index is The corresponding physical depth range is ,in , If multiple consecutive segments satisfy the conditions exist at the same grid point, the segment length takes precedence (i.e., ...). (The larger one is given priority), and the secondary layer segment is reserved in the form of a note in the hole layout reference diagram for future scheme adjustments.

[0184] The final output consists of three parts: a perforation reference diagram (in...) Overlay display , , and The list of recommended borehole locations (a set of horizontal grid points selected based on the above priority rules) and the recommended drilling depth interval (at each recommended borehole location) are also provided. ).

[0185] This application proposes an improved gravity-magnetic joint inversion constraint and weighting method. It monotonically maps the risk map based on strong oxidation demagnetization to a location-dependent depth bias function, forming an adaptive depth weighting. This weighting is combined with anisotropic focus constraints that focus along the contact zone and are stable in the normal direction, while simultaneously matching the depth sensitivity spectra of gravity and magnetic kernel functions. Unlike existing algorithms that commonly use fixed depth weights and isotropic smoothing, this method suppresses shallow responses and enhances deep responses at high-risk grid points. It maintains strip continuity in the strike direction and constrains non-causal abrupt changes in the normal direction, thereby reducing the bias of near-surface demagnetization on the shallow interpretation of magnetic susceptibility, improving the stable imaging and strike-alignment consistency of deep anomalies, and enhancing the directional positioning capability of deep preferential mineralization zones.

[0186] This application proposes a method for constructing a demagnetizing risk field using physical-guided conditional diffusion. It quantifies oxidation thickness using hyperspectral extracted iron oxides and hydrous mineral indicators combined with weathering / topographic parameters, and maps the demagnetizing trend based on skarn lithology sensitivity. Subsequently, it constructs anisotropic diffusion coefficients and source terms varying with oxidation thickness within the contact zone working domain, applies zero-flux boundaries and dissipation in non-oxidized areas, and iteratively calculates a risk distribution with clear gradations. Unlike existing methods that generate masks / priority regions using empirical thresholds or morphological filtering, this method explicitly encodes the monotonic causal relationship and boundary constraints of oxidation thickness-demagnetization intensity. This ensures that risk propagates reasonably along the orientation within the oxidation control range and is restricted in the normal direction, effectively reducing risk transgression and pseudo-anomaly expansion, and providing physically consistent control quantities for subsequent depth-weighted calculations.

[0187] This application proposes a holistic method for unified contact zone modeling and genetic consistency verification. A unique grid reference and strike / normal direction field are established based on a common coverage area, ensuring that airborne magnetics, airborne gravity, hyperspectral, and inversion results are aligned at the same grid point. Based on this, deep-seated preferential mineralization zones and shallow-seated demagnetized suspected areas are paired and labeled at the same horizontal grid point. Segments are constructed along the strike, and the continuity, normal constraint, and spatial overlap of skarn mineral indicators of the deep-seated preferential mineralization zones are scored and judged for consistency. Finally, borehole placement suggestions and depth ranges are output. Unlike existing processes that separate deep and shallow information and lack a unified reference and systematic comparison, this method achieves parallel comparison and priority division of deep and shallow genesis under a unified spatial reference, reducing spatial mismatch and interpretation bias, and improving the targeting and verifiability of borehole deployment.

[0188] In another aspect, this application also provides a readable storage medium, which may be the storage medium included in the aforementioned apparatus in the above embodiments; or it may be a standalone storage medium not assembled into the device. The storage medium stores one or more programs, which are used by one or more processors to execute the skarn deep mineralization zone location method described in this application.

[0189] Storage media, including both permanent and non-permanent, removable and non-removable media, can be used to store information by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0190] The foregoing description of specific exemplary embodiments of the invention is for illustrative and explanatory purposes. These descriptions are not intended to limit the invention to the precise forms disclosed, and it will be apparent that many changes and variations can be made in accordance with the foregoing teachings. The exemplary embodiments were chosen and described in order to explain the specific principles of the invention and its practical application, thereby enabling those skilled in the art to implement and utilize various different exemplary embodiments of the invention, as well as various different choices and variations. The scope of the invention is intended to be defined by the claims and their equivalents.

Claims

1. A method for locating a deep mineralization zone of skarns, characterized in that, The method comprises the following steps: S1, based on the contact zone geological boundary and structural line of carbonate rock and intermediate-acid intrusive body, and the common coverage range of airborne magnetic method, airborne gravity and hyperspectral data, a contact zone working area and a unique grid reference are generated; S2, within the contact zone working area, based on the unique grid reference, iron oxide and hydrous mineral indicators are extracted from the hyperspectral data, and an oxidation thickness map and a demagnetization clue preliminary map are generated in combination with topographic and weathering zone information; S3, based on the oxidation thickness map and the demagnetization clue preliminary map, the physical constraint of strong oxidation leading to near-surface demagnetization is introduced, and the physical guided condition diffusion is performed to output a demagnetization risk map; S4, according to the demagnetization risk map, a position-related depth bias function is constructed at each grid point to realize adaptive depth weighting, and anisotropic focus constraints are set along the strike of the contact zone to generate a joint inversion configuration; S5, within the contact zone working area, based on the unique grid reference, the joint inversion configuration is used to implement gravity-magnetic joint inversion of airborne gravity and airborne magnetic method to output density and magnetic susceptibility bodies, and the unique grid reference is used to pair and label deep preferential mineralization zone and shallow demagnetization suspected area; S6, along the strike of the contact zone, the continuity of the deep preferential mineralization zone is checked on the paired label, and the spatial variation of skarn-related mineral indicators is verified for consistency to output preferential zone label and low-priority zone label; S7, the paired label is superimposed with the preferential zone label and the low-priority zone label to form a hole layout reference map, and the recommended drilling location and depth range are given in the contact zone working area to output the results for drilling deployment.

2. The method of claim 1, wherein, The S1 based on the contact zone geological boundary and structural line of carbonate rock and intermediate-acid intrusive body, and the common coverage range of airborne magnetic method, airborne gravity and hyperspectral data, generates a contact zone working area and a unique grid reference, comprising: The contact zone geological boundary and structural line are obtained; The common coverage range is selected; The contact zone geological boundary is taken as the center line, and the strike and normal directions are determined according to the structural line; Buffer zone widths are set along the strike and normal directions respectively to form an initial area; The initial area is cropped according to the common coverage range to obtain the contact zone working area; The outer boundary of the common coverage range is taken as the grid outer frame, and the specification resolution is determined according to the spatial sampling interval of the airborne magnetic method, airborne gravity and hyperspectral data; According to the grid outer frame and the specification resolution, the unique grid reference is generated.

3. The method of claim 1, wherein, The S2, within the contact zone working area, based on the unique grid reference, iron oxide and hydrous mineral indicators are extracted from the hyperspectral data, and an oxidation thickness map and a demagnetization clue preliminary map are generated in combination with topographic and weathering zone information, comprising: Within the contact zone working area and the unique grid reference, the iron oxide absorption depth indicator ratio and the hydrous mineral absorption depth indicator ratio are calculated at each grid point to form a mineral indicator grid layer; In the same grid point in the contact zone working area, the topographic slope, denudation form and weathering layer thickness grade in the weathering zone information are read to form a weathering constraint grid layer; Superimpose the mineral indicator grid layer and the weathering constraint grid layer under strike and normal constraints, and output a grid point feature set; At each grid point in the grid point feature set, estimate the oxide layer thickness according to the correspondence between the iron oxide and hydrous mineral indicators and the weathering zone information, smooth the thickness variation along the strike and normal to meet the geometric continuity of the contact zone, and write the thickness value or thickness level on the unique grid reference to output the oxide thickness map; At each grid point in the grid point feature set, calculate the demagnetization trend value according to the mapping relationship between the oxide thickness and the magnetic susceptibility attenuation, maintain the trend continuity along the strike of the contact zone, and suppress non-causal mutations in the normal direction, and write the trend value or trend level on the unique grid reference to output the demagnetization clue preliminary map.

4. The method of claim 1, wherein, The S3 introduces the physical constraint of strong oxidation leading to near-surface demagnetization based on the oxide thickness map and the demagnetization clue preliminary map, performs physical guided condition diffusion, and outputs a demagnetization risk map, including: Generate a diffusion coefficient field at each grid point with the oxide thickness map, so that the diffusion coefficient of the grid point with large oxide thickness increases, and the diffusion coefficient of the grid point with small oxide thickness decreases, and a larger directional weight is given in the strike direction and a smaller directional weight is given in the normal direction to meet the geometric extension of the contact zone; Generate a source term at each grid point with the demagnetization clue preliminary map, so that the grid point with obvious demagnetization trend is a risk supply point, and the grid point without trend is a non-supply point; Use zero flux as the contact zone working domain boundary constraint to limit the risk from crossing the working domain boundary; Iterative propagation is performed under the action of the diffusion coefficient field and the source term until a stable risk distribution field is formed on the unique grid reference to obtain the demagnetization risk map.

5. The method of claim 1, wherein, The S4 constructs a position-dependent depth bias function at each grid point according to the demagnetization risk map to realize adaptive depth weighting, and sets anisotropic focus point constraints along the strike of the contact zone to generate a joint inversion configuration, including: Read the risk value of each grid point on the unique grid reference, map the risk value to a bias coefficient, and the mapping relationship is a monotonic increasing mapping, so that high risk corresponds to stronger deep bias; At each grid point, determine the depth weight according to the bias coefficient and the depth horizon, and the increasing speed of the depth weight increasing with the depth is controlled by the bias coefficient; Match the depth weight in the contact zone working domain with the depth sensitivity of the kernel function of the airborne magnetic method and the airborne gravity to determine the depth weight field; Determine the strike direction weight and the normal direction weight of the anisotropic focus point constraint according to the bias coefficient; Organize the depth weight field, the strike direction weight, and the normal direction weight into the joint inversion configuration.

6. The method of claim 1, wherein, The S5 refers to the unique grid reference in the contact zone working domain, performs gravity-magnetic joint inversion of the airborne gravity and the airborne magnetic method according to the joint inversion configuration, outputs the density body and the magnetic susceptibility body, and pairs and labels the deep preferential mineralization zone and the shallow demagnetization suspected area under the unique grid reference, including: The forward operator of the airborne gravity and the airborne magnetic method is established on the unique grid reference, the depth weight field and the anisotropic focus constraint in the joint inversion configuration are written into the inversion target control quantity, the directional constraint of the directional focusing and the normal direction stability is effective in the whole working domain, the position related depth bias is effective in the whole working domain, the directional constraint of the directional focusing and the normal direction stability is effective in the whole working domain, the position related depth bias is effective in the whole working domain; the sensitivity of the airborne gravity to the density body and the sensitivity of the airborne magnetic method to the magnetization body are jointly driven to update in the same grid point, so that the two types of data are mutually restricted in the same spatial unit and the demagnetization of the shallow layer is avoided; the above configuration and data driving are iteratively solved until stable, and the density body and the magnetization body expressed on the unique grid reference are output; In the contact zone working domain, the depth layer set of each horizontal grid point is scanned, if the high density of the density body and the medium-high magnetization of the magnetization body are met at the deep layer position, and the continuity along the strike and the normal stability are met, the deep preferential mineralization zone label is written in the deep layer position of the grid point; at the same time, in the shallow layer position of the same grid point, if the demagnetization risk map is high level and the magnetization body is low value, the shallow demagnetization suspected area label is written; the above two types of labels are output in the form of pairs on the unique grid reference.

7. The method of claim 1, wherein, The S6 checks the continuity of the deep preferential mineralization zone along the strike of the contact zone on the paired label, and verifies the consistency with the spatial variation of the skarn related mineral indication, and outputs the preferential zone and the low priority zone label, including: On the unique grid reference, the paired label is input, the horizontal grid point set containing the deep preferential mineralization zone label is selected, and the corresponding grid point and the skarn related mineral indication intensity are read; Along the strike of the contact zone, the horizontal grid point set in the horizontal grid point set is grouped to establish a strike section set; According to the indication intensity of the skarn related mineral, the consistency score of each strike section in the strike section set is calculated, the determination result is divided into consistent and inconsistent according to the consistency score, and the determination result is written in the consistency verification layer on the unique grid reference; On the basis of the consistency verification layer, the priority of the deep preferential mineralization zone is allocated and recorded in parallel with the shallow demagnetization suspected area, and the preferential zone label and the low priority zone label are output.

8. The method of claim 7, wherein, According to the consistency score, the determination result is divided into consistent and inconsistent, including: If the consistency score is greater than or equal to the consistency threshold, the corresponding strike section is determined to be consistent; otherwise, it is determined to be inconsistent.

9. The method of claim 7, wherein, On the basis of the consistency verification layer, the priority of the deep preferential mineralization zone is allocated and recorded in parallel with the shallow demagnetization suspected area, and the preferential zone label and the low priority zone label are output, including: In the contact zone working domain, the depth layer set of each horizontal grid point is scanned, if the high density of the density body and the medium-high magnetization of the magnetization body are met at the deep layer position, and the continuity along the strike and the normal stability are met, the deep preferential mineralization zone label is written in the deep layer position of the grid point; at the same time, in the shallow layer position of the same grid point, if the demagnetization risk map is high level and the magnetization body is low value, the shallow demagnetization suspected area label is written; the above two types of labels are output in the form of pairs on the unique grid reference. In the contact zone working domain, the depth layer set of each horizontal grid point is scanned, if the high density of the density body and the medium-high magnetization of the magnetization body are met at the deep layer position, and the continuity along the strike and the normal stability are met, the deep preferential mineralization zone label is written in the deep layer position of the grid point; at the same time, in the shallow layer position of the same grid point, if the demagnetization risk map is high level and the magnetization body is low value, the shallow demagnetization suspected area label is written; the above two types of labels are output in the form of pairs on the unique grid reference. A priority band annotation and a low priority band annotation are determined based on the priority band label and the low priority band label, respectively, and the priority band annotation and the low priority band annotation are written on the label layer at the unique grid reference.

10. A readable storage medium, having stored thereon a computer program, characterized in that, The program is executed by the processor to realize the method for locating the deep mineralization zone of skarn as claimed in any one of claims 1-9.

Images