Positioning method for concealed copper-gold mine based on CSAMT and TEM joint inversion

Abstract

The present application relates to the technical field of geomagnetic measurement, in particular to a concealed copper-gold deposit positioning method based on CSAMT and TEM joint inversion, which comprises training a deep range identification model and a shallow range identification model, outputting a deep range value and a shallow range value, dividing into a TEM weight area, a joint weight area and a CSAMT weight area, respectively performing differential weighting fusion, obtaining a two-dimensional comprehensive resistivity structure map, and constructing a three-dimensional comprehensive resistivity structure map. The present application establishes a quantitative mapping model between the effective detection depth and the geological parameters of the work area, realizes intelligent and objective definition of the detection range, and adaptively divides the weight area, which not only ensures that the high-resolution method is dominant, but also suppresses the multiple solutions of a single method by fusing two kinds of data, and at the same time improves the reliability and accuracy of resistivity imaging in the full depth section, thereby providing key technical support for accurately delineating concealed ore bodies and reducing drilling risks.

Description

Technical Field

[0001] This invention relates to the field of geomagnetic measurement technology, and more specifically, to a method for locating concealed copper-gold deposits based on joint inversion of CSAMT and TEM. Background Technology

[0002] Concealed copper-gold deposits lack surface outcrops, and their exploration and location rely heavily on geophysical methods. Among these, Controlled-Source Audio-Frequency Magnetotelluric (CSAMT) and Transient Electromagnetic (TEM) are two commonly used electromagnetic exploration methods. CSAMT has a large detection depth and strong anti-interference ability, and can simultaneously conduct depth sounding and profiling studies, but its ability to resolve shallow geological bodies is relatively insufficient. The TEM method has a high lateral and vertical resolution for medium-depth geological bodies, but it has obvious shallow blind zones (usually around 0-20m, affected by the transmission current turn-off time and coil characteristics). At the same time, its effective detection depth is limited by factors such as transmission power and geoelectric conditions. In existing technologies, it is generally believed that "TEM is suitable for shallow areas and CSAMT is suitable for deep areas." However, in actual exploration, the effective detection range of the two methods is not a fixed value, but varies significantly with the geological conditions of the work area (surface resistivity, structural complexity, interference level, etc.). A general qualitative judgment cannot determine the boundary of its own detection capability. As a result, when optimizing the combination of CSAMT and TEM data for specific geological conditions of a work area and simultaneously acquiring CSAMT and TEM data for joint inversion, global unified weights, coupling based on cross gradient constraints, or fusion through external well logging constraints are often used. This results in limited accuracy of geophysical results and makes it difficult to form an accurate comprehensive structural map to guide the location of the mining area.

[0003] In view of this, we propose a method for locating concealed copper-gold deposits that can accurately distinguish the optimal detection depth ranges of TEM and CSAMT, and adopt a differentiated weighted joint inversion strategy for different depth ranges. Summary of the Invention

[0004] The purpose of this invention is to provide a method for locating concealed copper-gold deposits based on the joint inversion of CSAMT and TEM, which utilizes the complementarity of CSAMT and TEM in terms of geoelectric sensitive depth range and data format to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, this invention provides a method for locating concealed copper-gold deposits based on joint inversion of CSAMT and TEM, comprising the following steps: S1. Construct a historical dataset, which includes CSAMT measured data, TEM measured data and corresponding borehole verification data collected under the same geological conditions. Train a deep range identification model based on the CSAMT measured data and borehole verification data, and output the deep range value. Train a shallow range identification model based on the TEM measured data and borehole verification data, and output the deep range value. S2. When the shallow range value is greater than the deep range value, the depth range of the target exploration area is divided into TEM weight zone, joint weight zone, and CSAMT weight zone. Specifically: the intersection of the shallow range value and the deep range value is defined as the joint weight zone; the portion of the difference between the shallow range value and the deep range value that is less than the joint weight zone is defined as the TEM weight zone; and the portion of the difference between the shallow range value and the deep range value that is greater than the joint weight zone is defined as the CSAMT weight zone. S3. Set up survey lines in the area to be surveyed, collect CSAMT and TEM data simultaneously, and perform data registration in a unified coordinate system. Obtain CSAMT resistivity profile and TEM resistivity profile corresponding to the same survey line location through inversion imaging.

[0006] S4. For the CSAMT resistivity profile and TEM resistivity profile, perform differentiated weighted fusion according to the TEM weight region, joint weight region, and CSAMT weight region divided in S2, including: Within the TEM weighting region, the weight of the TEM resistivity profile is made higher than that of the CSAMT resistivity profile. Within the joint weighting region, the weights of the TEM resistivity profile and the CSAMT resistivity profile are balanced. Within the CSAMT weighting region, the weight of the CSAMT resistivity profile is set higher than that of the TEM resistivity profile. The fusion yields a two-dimensional composite resistivity structure map; S5. Construct a three-dimensional comprehensive resistivity structure map from the two-dimensional comprehensive resistivity structure map of multiple survey lines, identify favorable mineralization areas, delineate the spatial distribution range of concealed copper-gold ore bodies based on resistivity anomaly characteristics, and extract the boundary coordinates of the mining area.

[0007] As a further improvement to this technical solution, in step S1, a deep range identification model is trained based on the CSAMT measured data and borehole verification data, and the deep range value is output, including the following steps: For each CSAMT resistivity profile sample in the CSAMT measured data, the entire detection depth range is divided into several depth windows at equal intervals (e.g., Δz=10m). The relative error between the CSAMT inversion interface depth and the actual interface depth corresponding to the borehole is counted window by window from shallow to deep. The first relative error below the preset threshold is marked as the upper limit of the effective CSAMT detection for that profile. Multiple historical CSAMT resistivity profile samples are summarized. Establish a mapping relationship between the effective detection limit of CSAMT and the key geological parameters of the work area, and use this mapping relationship as a deep range identification model; When faced with a new area to be surveyed, simply input the geological parameters of the area into the deep range identification model, and the output CSAMT deep range value is the upper limit of CSAMT effective detection to the maximum depth value.

[0008] As a further improvement to this technical solution, in step S1, a shallow range identification model is trained based on the TEM measured data and borehole verification data, and the shallow range value is output, including the following steps: For each TEM resistivity profile sample in the TEM measured data, first identify and remove the TEM blind zone depth. For the effective detection range of the blind zone, divide the entire remaining depth range into several depth windows at equal intervals. Calculate the relative error between the TEM inversion interface depth and the actual interface depth corresponding to the borehole window by window from the deep part to the TEM blind zone depth. Mark the first relative error below the preset threshold as the lower limit of the effective TEM detection of the profile. By summarizing multiple historical TEM resistivity profile samples, the effective detection limit of TEM and the key geological parameters of the work area corresponding to each sample are obtained. A mapping relationship between the effective detection limit of TEM and the geological parameters of the work area is established, and this mapping relationship is used as a shallow range identification model. When faced with a new area to be surveyed, the geological parameters of the area are input into the shallow range identification model, and the predicted effective detection limit of the TEM is output. The shallow range value is defined as the depth range from the TEM blind zone depth to the effective detection limit of the TEM.

[0009] As a further improvement to this technical solution, the relative error between the CSAMT inversion interface depth and the actual interface depth corresponding to the borehole is statistically analyzed window by window from shallow to deep, and the relative error between the TEM inversion interface depth and the actual interface depth corresponding to the borehole is statistically analyzed window by window from deep to shallow, both including the following steps: The extracted interface depths are summarized to form the inversion interface depth set of the sample. The true depth information of all key geological interfaces is extracted from the corresponding borehole verification data to form the actual interface depth set of the borehole. For each depth window, select all inverted interface depths that fall within the current window range from the inverted interface depth set. For each selected inverted interface depth, find the actual interface depth in the actual interface depth set that is closest in depth and whose absolute distance does not exceed the preset maximum allowable distance for matching. If the actual interface depth exceeds the preset maximum allowable distance, it is considered that there is no matching object and is not included in this statistic. For each successfully matched interface pair, calculate the relative error: relative error = |inverted interface depth - actual interface depth| / actual interface depth. Take the arithmetic mean of the relative errors of all successfully matched interfaces within this window as the average relative error of this depth window, thereby improving the accuracy of relative error calculation. If there are no successfully matched interface pairs within a certain depth window, the average relative error of this depth window is directly determined to be higher than the preset threshold.

[0010] As a further improvement to this technical solution, in step S2, when the shallow range value is greater than the deep range value, the depth range of the target exploration area is divided into a TEM weighted area, a joint weighted area, and a CSAMT weighted area, and the expression includes: Let the shallow region range be [H0, H... TEM The depth range is [H]. CSAMT H max ]; When the shallow range value is greater than the deep range value, [H0, H CSAMT The depth range of [H] is used as the TEM weight region, and [H] is used as the TEM weight region. TEM H max The depth range of [H] is used as the CSAMT weight region, and [H] is used as the weight region. TEM H CSAMT The depth range of ] is used as the joint weight region.

[0011] As a further improvement to this technical solution, within the TEM weighting region, the weight ratio of the TEM resistivity profile to the CSAMT resistivity profile is set to 1.5-3.0. Within the joint weighting region, the weight ratio of the TEM resistivity profile to the CSAMT resistivity profile is set to 0.67-1.5. Within the CSAMT weighted region, the weight ratio of the CSAMT resistivity profile to the TEM resistivity profile is set to 0.3-0.67.

[0012] As a further improvement to this technical solution, within the joint weighting region, the weights of the TEM resistivity profile and the CSAMT resistivity profile are continuously gradient functions along the depth. A cosine smooth transition technique is used to make the ratio smoothly transition from the highest value near the TEM weighting region to the lowest value near the CSAMT weighting region.

[0013] As a further improvement to this technical solution, step S4, which involves fusing to obtain a two-dimensional composite resistivity diagram, includes the following steps: Substituting the weights of the CSAMT resistivity profile and the TEM resistivity profile into the fusion formula, the fused resistivity value of this depth node is calculated. The fusion formula is as follows: , To integrate resistivity values, The weights for the TEM resistivity profile are... The weights for the CSAMT resistivity profile are... The CSAMT resistivity of the node at this depth is given. The TEM resistivity of the node at this depth; The above calculations are performed point by point and depth by depth along the survey line to generate and store the fused resistivity values ​​as a two-dimensional composite resistivity data volume for that survey line location. Based on the two-dimensional composite resistivity data volume, a resistivity cross-sectional diagram with coordinate axes is drawn, which is the two-dimensional composite resistivity construction diagram. In this diagram, the same resistivity values ​​are displayed using a unified color code.

[0014] As a further improvement to this technical solution, step S5 specifically includes the following steps: Receive a two-dimensional composite resistivity construction map of each survey line, which includes spatial coordinates in a unified coordinate system and corresponding fused resistivity values. Use the outer rectangle of the survey line coverage area as the planar range to determine the three-dimensional grid range. The grid step size of the three-dimensional grid is set to 10 meters in the plane and 10 meters in the depth. For each unknown node in the three-dimensional grid, the optimal estimated resistivity value of the node is calculated based on the known points in its neighborhood using the Kriging equations. The resistivity values ​​of each grid node obtained by interpolation are stored as a three-dimensional array and rendered to form a three-dimensional integrated resistivity volume model covering the entire survey area. A pre-defined upper limit threshold for resistivity of the concealed copper-gold ore body is used to select voxels with resistivity less than the upper limit threshold from the three-dimensional comprehensive resistivity volume model, forming an initial model of the low-resistivity anomaly. Through three-dimensional connected domain analysis and morphological constraints, each voxel has a continuous set of spatial voxels, so that the internal volume enclosed by the set of spatial voxels is used as the three-dimensional spatial distribution range of the concealed copper-gold mineralization body. The outline is vertically projected, and the outer envelope polygon of the projected outline is extracted. The coordinates of the inflection points of the polygon are the coordinates of the two-dimensional boundary envelope of the mining area on the plane.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: This method for locating concealed copper-gold deposits based on the joint inversion of CSAMT and TEM achieves two key improvements. First, by using deep-range identification models trained on historical borehole verification data and shallow-range identification models, it can automatically and objectively predict the effective detection upper limit of CSAMT and the effective detection lower limit of TEM, realizing intelligent calibration of the reliable detection depth range for different work areas and eliminating reliance on human experience. Second, based on the predicted depth range, it adaptively divides the TEM weight zone, joint weight zone, and CSAMT weight zone, conforming to the respective physical resolution limits of the two methods. This ensures that the detection depth remains consistent across all depth segments. The method is dominated by high signal-to-noise ratio and high resolution detection methods. By setting a balanced weight ratio in the joint weighting area, it effectively integrates the high lateral and vertical resolution of TEM with the deep coverage capability of CSAMT, which significantly suppresses the multiple solutions of single-method inversion. At the same time, the cosine smoothing transition technology makes the weight ratio change continuously and gradually along the depth, which completely eliminates the abrupt change in resistivity value and false structural artifacts at the boundary of the partition. This greatly improves the authenticity and continuity of resistivity imaging in the whole depth range, thus providing a more reliable basis for the accurate delineation and boundary coordinate extraction of concealed copper-gold ore bodies and significantly reducing the uncertainty of drilling verification. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the overall structure of the present invention; Figure 2 This is a partitioning demonstration diagram of the present invention; Figure 3 This is a diagram illustrating the survey line division of the present invention; Figure 4 This is a two-dimensional composite resistivity construction diagram of the present invention; Figure 5 This is a three-dimensional composite resistivity construction diagram of the present invention. Detailed Implementation

[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1 Please see Figures 1-5 As shown in the figure, this embodiment provides a method for locating concealed copper-gold deposits based on joint inversion of CSAMT and TEM, specifically as follows: Figure 1 As shown, it includes the following steps: S1. Collect all historical data of existing CSAMT and TEM exploration work within the metallogenic belt where the area to be explored is located, and construct a historical dataset. The historical dataset includes CSAMT measured data, TEM measured data and corresponding borehole verification data collected under the same geological conditions. Among them, the CSAMT measured data includes the original CSAMT time series record (electric field components Ex, Ey, magnetic field components Hx, Hy), measurement point coordinates, electrode distance, transmitter-receiver distance and other observation parameters, as well as the CSAMT resistivity profile map (two-dimensional gridded data, each grid node contains resistivity value) after preprocessing and inversion. The TEM measured data includes the original TEM induced electromotive force decay curve, measurement point coordinates, loop size, transmission current, turn-off time and other observation parameters, as well as the TEM resistivity profile map (two-dimensional gridded data, each grid node contains resistivity value) after preprocessing and inversion. The corresponding borehole verification data includes borehole columnar sections on each CSAMT and TEM profile, and the depths of various geological interfaces accurately marked in the borehole, including but not limited to: the bottom boundary of the overburden (bedrock surface), the main lithological stratification interface, the top and bottom plates of the fracture zone, the top and bottom plates of the mineralization and alteration zone, and the top and bottom plates of the ore body. Firstly, a deep-range identification model is trained based on CSAMT measured data and borehole verification data, and the deep-range value is output, including the following steps: For each CSAMT resistivity profile sample in the CSAMT measured data, the entire detection depth range is divided into several depth windows at equal intervals (e.g., Δz=10m). The relative error between the CSAMT inversion interface depth and the corresponding actual borehole interface depth is statistically analyzed window by window from shallow to deep. Along this direction, the relative error usually shows a gradually decreasing trend. The first relative error below a preset threshold (the preset threshold is 20%, which is the criterion for judging the relative error of the CSAMT inversion interface depth) is marked as the effective detection upper limit of the CSAMT for that profile. Multiple historical CSAMT resistivity profiles are then summarized. The resistivity profile sample can establish a mapping relationship between the effective detection upper limit of CSAMT and key geological parameters of the work area (such as surface resistivity, structural complexity, etc.). This mapping relationship is used as a deep range identification model. When facing a new work area to be explored, only the geological parameters of the work area need to be input into the deep range identification model, and the output CSAMT deep range value is the effective detection upper limit of CSAMT to the maximum depth value. The critical depth value defines the effective detection upper limit of the CSAMT method, which means that from the critical depth value to the maximum detection depth, the CSAMT inversion result reaches the reliable resolution standard.

[0019] Secondly, a shallow depth identification model is trained based on TEM measured data and borehole verification data, and the shallow depth value is output, including the following steps: For each TEM resistivity profile sample in the measured TEM data, first identify and remove the TEM blind zone depth (depth 0~H0, where H0 is determined by factors such as instrument turn-off time, typically 10~30m). For the effective detection range with depths below the blind zone, divide the entire remaining depth range into several depth windows at equal intervals (e.g., Δz=10m). From the depth towards the TEM blind zone depth, window by window, statistically analyze the relative error between the TEM inversion interface depth and the actual interface depth corresponding to the borehole. Along this direction, since the TEM secondary field signal intensity increases with decreasing depth... As the signal-to-noise ratio gradually increases, the inversion accuracy also increases. Therefore, the relative error usually shows a gradual decreasing trend. The first relative error (i.e., the first one encountered in the process of advancing from deep to shallow) is marked as the effective detection limit of the TEM for that profile when it is lower than the preset threshold (the preset threshold is 15%, which is the evaluation standard for the relative error of the TEM inversion interface depth). This means that from the bottom boundary H0 of the blind zone to the effective detection limit of the TEM, the TEM inversion result reaches the reliable resolution standard (error < 15%). Beyond this depth, the signal attenuates to the noise level, and the inversion is no longer reliable. By summarizing multiple historical TEM resistivity profile samples, the effective detection limit of TEM and key geological parameters of the work area (such as surface resistivity, structural complexity, target depth range, interference noise level, etc.) corresponding to each sample are obtained. A mapping relationship between the effective detection limit of TEM and the geological parameters of the work area is established, and this mapping relationship is used as a shallow range identification model. When faced with a new area to be surveyed, the geological parameters of the area are input into the shallow range identification model, and the predicted effective detection limit of TEM is output. The shallow range value is defined as the depth interval from the depth of the TEM blind zone to the effective detection limit of TEM. It means that the depth interval from the bottom boundary H0 of the blind zone to the effective detection limit of TEM is the reliable detection range of TEM. Within this interval, TEM data has high resolution and can dominate fusion.

[0020] Because the resistivity values ​​obtained from TEM and CSAMT inversions are equivalent resistivity over a large volumetric range, they differ fundamentally in measurement scale from borehole resistivity logging (which measures resistivity on a centimeter to decimeter scale near the wellbore). Therefore, the values ​​cannot be directly compared point-to-point. Thus, the relative error between the CSAMT inversion interface depth and the corresponding actual borehole interface depth is statistically analyzed window-by-window from shallow to deep, and the relative error between the TEM inversion interface depth and the corresponding actual borehole interface depth is statistically analyzed window-by-window from deep to shallow. Both of these steps include: The extracted interface depths are summarized to form the inversion interface depth set of the sample. The true depth information of all key geological interfaces is extracted from the corresponding borehole verification data to form the actual interface depth set of the borehole. For each depth window, all inversion interface depths falling within the current window range are selected from the inversion interface depth set (the depth position corresponding to the gradient of the subsurface resistivity with depth exceeding the gradient threshold is called the inversion interface depth). For each selected inversion interface depth, the actual interface depth with the closest depth and an absolute distance not exceeding the preset maximum allowable distance (e.g., 20m) is found in the actual interface depth set for matching. If the actual interface depth exceeds the preset maximum allowable distance, it is considered as no matching object and is not included in this statistic. For each successfully matched interface pair, calculate the relative error: relative error = |inverted interface depth - actual interface depth| / actual interface depth. Take the arithmetic mean of the relative errors of all successfully matched interfaces within this window as the average relative error of this depth window, thereby improving the accuracy of relative error calculation. If there are no successfully matched interface pairs within a certain depth window, it is directly determined that the average relative error of this depth window is higher than the preset threshold, which means that under the current conditions of the work area, the two methods cannot form an effective joint detection range.

[0021] S2, such as Figure 2 As shown, when the shallow range value is greater than the deep range value, the depth range of the target exploration area is divided into the TEM weight zone, the joint weight zone, and the CSAMT weight zone. Specifically: the intersection of the shallow range value and the deep range value is defined as the joint weight zone; the portion of the difference between the shallow range value and the deep range value that is less than the joint weight zone is defined as the TEM weight zone; and the portion of the difference between the shallow range value and the deep range value that is greater than the joint weight zone is defined as the CSAMT weight zone.

[0022] Specifically, in S2, when the shallow range value is greater than the deep range value, the depth range of the target exploration area is divided into the TEM weighted area, the joint weighted area, and the CSAMT weighted area. The expression includes: Let the shallow region range be [H0, H... TEM The depth range is [H]. CSAMT H max ]; When the shallow range value is greater than the deep range value, [H0, H CSAMT The depth range of [H] is used as the TEM weight region, and [H] is used as the TEM weight region. TEM H max The depth range of [H] is used as the CSAMT weight region, and [H] is used as the weight region. TEM H CSAMTThe depth range is used as the joint weight region. The partitioning is directly mapped to the detection physical limits of each method (TEM maximum depth, CSAMT shallowest reliable depth). The shallow range value is greater than the deep range value. As the resistivity, structural complexity, noise level and other factors of the work area change, the above partitioning strategy can automatically adjust the partitioning according to the characteristics of the work area, always maintain the optimal configuration, and improve the accuracy of subsequent joint analysis.

[0023] S3, such as Figure 3 As shown, survey lines are laid out in the area to be surveyed. The survey lines are arranged in parallel or intersecting patterns in space, with a line spacing of 100 to 400 m. Survey points are set along each survey line at a certain point spacing (e.g., 20 to 50 m). CSAMT data and TEM data are collected simultaneously, and data registration is performed in a unified coordinate system. CSAMT resistivity profile and TEM resistivity profile corresponding to the same survey line location are obtained through inversion imaging.

[0024] The technical methods involved in S3 are all common knowledge to those skilled in the art, and will not be elaborated upon here. Specifically: CSAMT data acquisition uses a frequency-domain controllable source electromagnetic method exploration system. Currently, mainstream equipment at home and abroad includes the V8 electrical method workstation and the GDP-32II multi-functional electrical method instrument. TEM data acquisition uses a time-domain transient electromagnetic method exploration system. Mainstream equipment at home and abroad includes the PROTEM series transient electromagnetic instrument and the TerraTEM transient electromagnetic instrument. Since CSAMT and TEM measuring points may use different measurement markers and coordinate systems, the first step is to unify the spatial coordinates: use GPS-RTK or a total station to measure the precise coordinates (latitude, longitude, and elevation) of each measuring point, and convert the latitude and longitude of all measuring points to the same projected coordinate system (such as Gauss-Kruger 3-degree zone coordinates) to obtain the plane (X,Y) and elevation (Z) coordinates of each measuring point. The CSAMT measured data processing software has a point deviation correction function, which can replace the theoretical coordinates of each measuring point with the measured coordinates and attach the elevation value. Then, Kriging interpolation or cubic spline interpolation is used for spatial interpolation processing to make the spatial positions of CSAMT and TEM measuring points coincide on the measuring line and the point distance is the same. Then, the nonlinear conjugate gradient method (NLCG) or Occam method is used to perform two-dimensional inversion and generate a CSAMT resistivity profile. The TEM two-dimensional inversion is realized by the pseudo-MT two-dimensional inversion technique based on time-frequency conversion and a TEM resistivity profile is generated.

[0025] S4. Differential weighted fusion is performed on the CSAMT resistivity profile and TEM resistivity profile according to the TEM weighted region, joint weighted region, and CSAMT weighted region divided in S2. This facilitates the use of the optimal resolution detection method at each depth level, including: Within the TEM weighted region, the TEM inversion interface depth error is below 15%, indicating highly reliable data. However, CSAMT, being above its shallowest effective depth, suffers from insufficient shallow resolution, resulting in an inversion interface error exceeding 20%, making the data unreliable. Giving CSAMT excessively high weight in this region would introduce low-quality information and reduce the accuracy of the fusion results. Therefore, the weight of the TEM resistivity profile is set higher than that of the CSAMT resistivity profile, with TEM taking a dominant position. The weight ratio of the TEM resistivity profile to the CSAMT resistivity profile is set between 1.5 and 3.0, ensuring the dominance of TEM while not completely ignoring CSAMT (setting it to zero would result in the loss of any potentially useful background information). Within the joint weighting region, the inversion interface depth errors of both methods are below their respective thresholds (TEM < 15%, CSAMT < 20%), indicating reliable resolution for both. Therefore, the weights of the TEM and CSAMT resistivity profiles are balanced, with the weight ratio set at 0.67-1.5. TEM provides higher lateral and vertical resolution, while CSAMT offers broader depth coverage and deeper trend constraints. This balanced fusion effectively suppresses the ambiguity of individual methods while preserving the advantages of both, achieving optimal overall imaging results. Within the joint weight region, the weights of the TEM resistivity profile and the CSAMT resistivity profile are continuous gradient functions along the depth. A cosine smoothing transition technique is used to make the ratio smoothly transition from the highest value (e.g., 1.5) near the TEM weight region to the lowest value (e.g., 0.67) near the CSAMT weight region, with the rate of change at both ends approaching zero, thus avoiding any abrupt changes in the weight derivative at the partition boundary. Within the CSAMT weighted region, the maximum effective depth of the TEM is exceeded, and the TEM signal attenuates below the noise level. The inversion interface error exceeds 15%, and the data is no longer reliable. However, CSAMT remains within the reliable resolution range at this depth. Therefore, the weight of the CSAMT resistivity profile is higher than that of the TEM resistivity profile. The weight ratio of the CSAMT resistivity profile to the TEM resistivity profile is set to 0.3-0.67. If too high a weight is given to the TEM in this range, it will introduce severely distorted information and contaminate the final structural profile. Therefore, CSAMT should dominate, while not completely ignoring the TEM (e.g., retaining a small amount of TEM weight to maintain a smooth transition of the longitudinal continuity of the profile). The process of fusing the resistivity data to obtain a two-dimensional composite resistivity map includes the following steps: Substituting the weights of the CSAMT resistivity profile and the TEM resistivity profile into the fusion formula, the fused resistivity value of this depth node is calculated. The fusion formula is as follows: , To integrate resistivity values, The weights for the TEM resistivity profile are... The weights for the CSAMT resistivity profile are... The CSAMT resistivity of the node at this depth is given. The TEM resistivity of the node at this depth; The above calculations are performed point by point and depth by depth along the survey line to generate and store the fused resistivity values ​​as a two-dimensional comprehensive resistivity data volume for that survey line location. Based on the two-dimensional comprehensive resistivity data volume, a resistivity cross-sectional diagram with coordinate axes is drawn, i.e., a two-dimensional comprehensive resistivity construction diagram. In this diagram, the same resistivity values ​​are displayed using a unified color code, which helps to clearly present the electrical structure distribution characteristics of the underground medium.

[0026] S5, such as Figure 4 As shown, a three-dimensional comprehensive resistivity structure map is constructed from the two-dimensional comprehensive resistivity structure map of multiple survey lines to identify favorable mineralization areas, delineate the spatial distribution range of concealed copper-gold ore bodies based on resistivity anomaly characteristics, and extract the boundary coordinates of the mining area. The specific steps include: Receive a two-dimensional composite resistivity construction map of each survey line, which includes spatial coordinates in a unified coordinate system and corresponding fused resistivity values. Use the outer rectangle of the survey line coverage area as the planar range to determine the three-dimensional grid range. The grid step size of the three-dimensional grid is set to 10 meters in the plane and 10 meters in the depth. For each unknown node (i.e. a point not located at a known survey line) in the three-dimensional grid, the optimal estimated resistivity value of the node is calculated based on the known points (merged resistivity values) within its neighborhood (horizontal radius 200m, depth 50m). The resistivity values ​​of each grid node obtained by interpolation are stored as a three-dimensional array and rendered to form a three-dimensional integrated resistivity volume model covering the entire survey area. Since concealed copper-gold ore bodies and their surrounding rock alteration zones typically exhibit low resistivity anomalies, a pre-defined upper limit threshold for resistivity of the concealed copper-gold ore body is used. Voxels with resistivity values ​​below this threshold are selected from the three-dimensional integrated resistivity volume model to form an initial model of the low-resistivity anomaly. Through three-dimensional connected domain analysis and morphological constraints, each voxel possesses a continuous set of spatial voxels. The internal volume enclosed by this set of spatial voxels serves as the three-dimensional spatial distribution range of the concealed copper-gold mineralization. The outline is then vertically projected, and the outer envelope polygon of the projected outline is extracted. The coordinates of the polygon's inflection points are the coordinates of the two-dimensional boundary envelope of the mining area on the plane. This facilitates direct guidance of drilling operations and significantly reduces positioning uncertainty.

[0027] Example 2 Set the shallow range value to [H0, H TEM The depth range is [H].CSAMT H max When the shallow range value is not greater than the deep range value, [H0, H] TEM [H] is used as the TEM weight region. CSAMT H max As the CSAMT weight region, there is no joint weight region, and [H] is not related to the weight region. TEM H CSAMT The intervals are assigned weights according to the principle of proximity. The intervals closer to TEM are temporarily assigned to the TEM weight zone (but given a rapidly decaying weight), and the intervals closer to CSAMT are assigned to the CSAMT weight zone.

[0028] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for locating concealed copper-gold deposits based on joint inversion of CSAMT and TEM, characterized in that, Includes the following steps: S1. Construct a historical dataset, which includes CSAMT measured data, TEM measured data and corresponding borehole verification data collected under the same geological conditions. Train a deep range identification model based on the CSAMT measured data and borehole verification data, and output the deep range value. Train a shallow range identification model based on the TEM measured data and borehole verification data, and output the deep range value. S2. When the shallow range value is greater than the deep range value, the depth range of the target exploration area is divided into TEM weight zone, joint weight zone, and CSAMT weight zone. Specifically: the intersection of the shallow range value and the deep range value is defined as the joint weight zone; the portion of the difference between the shallow range value and the deep range value that is less than the joint weight zone is defined as the TEM weight zone; and the portion of the difference between the shallow range value and the deep range value that is greater than the joint weight zone is defined as the CSAMT weight zone. S3. Set up survey lines in the area to be surveyed, collect CSAMT and TEM data simultaneously, and perform data registration in a unified coordinate system. Obtain CSAMT resistivity profile and TEM resistivity profile corresponding to the same survey line location through inversion imaging. S4. For the CSAMT resistivity profile and TEM resistivity profile, perform differentiated weighted fusion according to the TEM weight region, joint weight region, and CSAMT weight region divided in S2, including: Within the TEM weighting region, the weight of the TEM resistivity profile is made higher than that of the CSAMT resistivity profile. Within the joint weighting region, the weights of the TEM resistivity profile and the CSAMT resistivity profile are balanced. Within the CSAMT weighting region, the weight of the CSAMT resistivity profile is set higher than that of the TEM resistivity profile. The fusion yields a two-dimensional composite resistivity structure map; S5. Construct a three-dimensional comprehensive resistivity structure map from the two-dimensional comprehensive resistivity structure map of multiple survey lines, identify favorable mineralization areas, delineate the spatial distribution range of concealed copper-gold ore bodies based on resistivity anomaly characteristics, and extract the boundary coordinates of the mining area.

2. The method for locating concealed copper-gold deposits based on joint inversion of CSAMT and TEM according to claim 1, characterized in that: In step S1, a deep range identification model is trained based on the CSAMT measured data and borehole verification data, and the deep range value is output, including the following steps: For each CSAMT resistivity profile sample in the CSAMT measured data, the entire detection depth range is divided into several depth windows at equal intervals (e.g., Δz=10m). The relative error between the CSAMT inversion interface depth and the actual interface depth corresponding to the borehole is counted window by window from shallow to deep. The first relative error below the preset threshold is marked as the upper limit of the effective CSAMT detection for that profile. Multiple historical CSAMT resistivity profile samples are summarized. Establish a mapping relationship between the effective detection limit of CSAMT and the key geological parameters of the work area, and use this mapping relationship as a deep range identification model; When faced with a new area to be surveyed, simply input the geological parameters of the area into the deep range identification model, and the output CSAMT deep range value is the upper limit of CSAMT effective detection to the maximum depth value.

3. The method for locating concealed copper-gold deposits based on joint inversion of CSAMT and TEM according to claim 2, characterized in that: In step S1, a shallow range identification model is trained based on the TEM measured data and borehole verification data, and the shallow range value is output, including the following steps: For each TEM resistivity profile sample in the TEM measured data, first identify and remove the TEM blind zone depth. For the effective detection range of the blind zone, divide the entire remaining depth range into several depth windows at equal intervals. Calculate the relative error between the TEM inversion interface depth and the actual interface depth corresponding to the borehole window by window from the deep part to the TEM blind zone depth. Mark the first relative error below the preset threshold as the lower limit of the effective TEM detection of the profile. By summarizing multiple historical TEM resistivity profile samples, the effective detection limit of TEM and the key geological parameters of the work area corresponding to each sample are obtained. A mapping relationship between the effective detection limit of TEM and the geological parameters of the work area is established, and this mapping relationship is used as a shallow range identification model. When faced with a new area to be surveyed, the geological parameters of the area are input into the shallow range identification model, and the predicted effective detection limit of the TEM is output. The shallow range value is defined as the depth range from the TEM blind zone depth to the effective detection limit of the TEM.

4. The method for locating concealed copper-gold deposits based on joint inversion of CSAMT and TEM according to claim 3, characterized in that: The relative error between the CSAMT inversion interface depth and the actual interface depth corresponding to the borehole is statistically analyzed window by window from shallow to deep, and the relative error between the TEM inversion interface depth and the actual interface depth corresponding to the borehole is statistically analyzed window by window from deep to shallow, both including the following steps: The extracted interface depths are summarized to form the inversion interface depth set of the sample. The true depth information of all key geological interfaces is extracted from the corresponding borehole verification data to form the actual interface depth set of the borehole. For each depth window, select all inverted interface depths that fall within the current window range from the inverted interface depth set. For each selected inverted interface depth, find the actual interface depth in the actual interface depth set that is closest in depth and whose absolute distance does not exceed the preset maximum allowable distance for matching. If the actual interface depth exceeds the preset maximum allowable distance, it is considered that there is no matching object and is not included in this statistic. For each successfully matched interface pair, calculate the relative error: relative error = |inverted interface depth - actual interface depth| / actual interface depth. Take the arithmetic mean of the relative errors of all successfully matched interfaces within this window as the average relative error of this depth window, thereby improving the accuracy of relative error calculation. If there are no successfully matched interface pairs within a certain depth window, the average relative error of this depth window is directly determined to be higher than the preset threshold.

5. The method for locating concealed copper-gold deposits based on joint inversion of CSAMT and TEM according to claim 4, characterized in that: In step S2, when the shallow range value is greater than the deep range value, the depth range of the target exploration area is divided into a TEM weighted area, a combined weighted area, and a CSAMT weighted area, with the following expression: Let the shallow region range be [H0, H... TEM The depth range is [H]. CSAMT H max ]; When the shallow range value is greater than the deep range value, [H0, H CSAMT The depth range of [H] is used as the TEM weight region, and [H] is used as the TEM weight region. TEM H max The depth range of [H] is used as the CSAMT weight region, and [H] is used as the weight region. TEM H CSAMT The depth range of ] is used as the joint weight region.

6. The method for locating concealed copper-gold deposits based on joint inversion of CSAMT and TEM according to claim 5, characterized in that: Within the TEM weighting region, the weight ratio of the TEM resistivity profile to the CSAMT resistivity profile is set to 1.5-3.

0. Within the joint weighting region, the weight ratio of the TEM resistivity profile to the CSAMT resistivity profile is set to 0.67-1.

5. Within the CSAMT weighted region, the weight ratio of the CSAMT resistivity profile to the TEM resistivity profile is set to 0.3-0.

67.

7. The method for locating concealed copper-gold deposits based on joint inversion of CSAMT and TEM according to claim 6, characterized in that: Within the joint weighting region, the weights of the TEM resistivity profile and the CSAMT resistivity profile are continuously gradient functions along the depth. A cosine smoothing transition technique is used to smoothly transition the ratio from the highest value near the TEM weighting region to the lowest value near the CSAMT weighting region.

8. The method for locating concealed copper-gold deposits based on joint inversion of CSAMT and TEM according to claim 7, characterized in that: In step S4, the two-dimensional composite resistivity structure map is obtained by fusion, including the following steps: Substituting the weights of the CSAMT resistivity profile and the TEM resistivity profile into the fusion formula, the fused resistivity value of this depth node is calculated. The fusion formula is as follows: , To integrate resistivity values, The weights for the TEM resistivity profile are... The weights for the CSAMT resistivity profile are... The CSAMT resistivity of the node at this depth is given. The TEM resistivity of the node at this depth; The above calculations are performed point by point and depth by depth along the survey line to generate and store the fused resistivity values ​​as a two-dimensional composite resistivity data volume for that survey line location. Based on the two-dimensional composite resistivity data volume, a resistivity cross-sectional diagram with coordinate axes is drawn, which is the two-dimensional composite resistivity construction diagram. In this diagram, the same resistivity values ​​are displayed using a unified color code.

9. The method for locating concealed copper-gold deposits based on joint inversion of CSAMT and TEM according to claim 8, characterized in that: S5 specifically includes the following steps: Receive a two-dimensional composite resistivity construction map of each survey line, which includes spatial coordinates in a unified coordinate system and corresponding fused resistivity values. Use the outer rectangle of the survey line coverage area as the planar range to determine the three-dimensional grid range. The grid step size of the three-dimensional grid is set to 10 meters in the plane and 10 meters in the depth. For each unknown node in the three-dimensional grid, the optimal estimated resistivity value of the node is calculated based on the known points in its neighborhood using the Kriging equations. The resistivity values ​​of each grid node obtained by interpolation are stored as a three-dimensional array and rendered to form a three-dimensional integrated resistivity volume model covering the entire survey area. A pre-defined upper limit threshold for resistivity of the concealed copper-gold ore body is used to select voxels with resistivity less than the upper limit threshold from the three-dimensional comprehensive resistivity volume model, forming an initial model of the low-resistivity anomaly. Through three-dimensional connected domain analysis and morphological constraints, each voxel has a continuous set of spatial voxels, so that the internal volume enclosed by the set of spatial voxels is used as the three-dimensional spatial distribution range of the concealed copper-gold mineralization body. The outline is vertically projected, and the outer envelope polygon of the projected outline is extracted. The coordinates of the inflection points of the polygon are the coordinates of the two-dimensional boundary envelope of the mining area on the plane.

Images