A method for prospecting ion-adsorption type rare earth ore based on spatial heterogeneity modeling
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- YUNNAN UNIV
- Filing Date
- 2026-02-11
- Publication Date
- 2026-06-05
Smart Images
Figure CN122153377A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of mineral resource exploration technology, and more specifically, to an ion adsorption-based rare earth mineral prospecting method based on spatial heterogeneity modeling. Background Technology
[0002] Rare earth minerals are strategic national mineral resources. Among them, ion-adsorption rare earth minerals are an important type, mainly found in the weathered crust of granite. They are characterized by low mining costs and relatively simple beneficiation processes, and occupy an important position in my country's rare earth resource security system. Their formation is closely related to various factors such as the properties of the granite parent rock, the degree of weathering, topographical conditions, hydrological and climatic environment, and material migration processes. Due to the coupling effect of multiple factors, ion-adsorption rare earth minerals exhibit significant spatial heterogeneity and local enrichment characteristics, posing considerable challenges to the study of mineralization mechanisms and mineral exploration prediction.
[0003] In existing research on ion adsorption-type rare earth mineral exploration, prospecting prediction is typically conducted through geological analysis of typical deposits, combined with multi-source data such as geochemical anomalies, topographical conditions, and remote sensing anomalies. Commonly used methods include anomaly overlay analysis, weighted evidence method, analytic hierarchy process (AHP), and comprehensive evaluation methods based on statistical thresholds. While these methods can identify regional anomalies to some extent, they assume that ore-controlling factors have a consistent spatial effect on mineralization enrichment, making it difficult to characterize the spatially non-stationary relationship between ore-controlling factors and rare earth element enrichment. These methods generally suffer from the following shortcomings: they do not fully consider the spatial non-stationarity of the effects of ore-controlling factors, making it difficult to reflect the differences in mineralization mechanisms across different regions; the weights of ore-controlling factors are often globally uniformly set, lacking data-driven local quantitative constraints; and the identification of optimal values for ore-controlling factors relies heavily on empirical judgment, with insufficient quantitative basis.
[0004] In recent years, geographically weighted regression (MGWR) and its improved models have been introduced into the field of geological prospecting to analyze the spatial nonstationarity among variables. However, traditional MGWR models typically employ a uniform spatial scale, making it difficult to reflect the differences in the effects of different ore-forming factors at different spatial scales, thus limiting their explanatory power under complex geological conditions. Multi-scale MGWR models, by introducing independent optimal bandwidth parameters for different factors, distinguish the spatial scales of the effects of each ore-forming factor, providing a new technical means to reveal the spatial heterogeneity in the mineralization process. Nevertheless, existing research based on multi-scale MGWR is mostly focused on qualitative or semi-quantitative analysis, primarily used to identify the spatial influence characteristics of ore-controlling factors, and lacks a quantitative method that can be directly used for mineral prospecting prediction. Furthermore, it ignores the differences in the direction and intensity of the effects of ore-controlling factors in different regions, making it difficult to adapt to the diverse enrichment patterns formed by ion-adsorption rare earth deposits under complex topographic, climatic, and weathering conditions.
[0005] Therefore, there is an urgent need for a prospecting method that can quantitatively screen, extract optimal values, and comprehensively evaluate ore-forming factors based on full consideration of spatial heterogeneity, so as to finely characterize the mineralization and enrichment patterns of ion-adsorption type rare earth minerals and provide reliable technical support for regional prospecting prediction. Summary of the Invention
[0006] The purpose of this invention is to provide an ion adsorption-based rare earth mineral exploration method based on spatial heterogeneity modeling. This method fully considers the spatial heterogeneity characteristics of the mineralization process and adopts a mineral exploration method that quantitatively screens mineralization factors, extracts optimal values, and conducts comprehensive evaluation, which significantly improves the scientificity and accuracy of mineral exploration prediction.
[0007] The embodiments of the present invention are implemented as follows: A rare earth mineral exploration method based on spatial heterogeneity modeling using ion adsorption, comprising: S1. Obtain data on mineral exploration factors and rare earth element content in the study area, and use a multi-scale geographical weighted regression model to analyze the spatial heterogeneity of each mineral exploration factor on rare earth element enrichment, and obtain the local regression coefficients and local goodness of fit of the mineral exploration factors at each spatial location. S2. Based on local regression coefficients and local goodness of fit, identify the strong influence range of each prospecting factor on mineralization, and use data-driven inversion to find the optimal mineralization value of each prospecting factor. S3. By integrating local regression coefficients, local goodness of fit, and the degree to which the mineral exploration factor values of each location are close to their optimal mineralization values, a spatial distribution map of mineral exploration potential is generated, and target areas are delineated accordingly.
[0008] Furthermore, in other preferred embodiments of the present invention, in step S1, the multi-scale geographically weighted regression model independently assigns its spatial bandwidth to each mineral exploration factor.
[0009] Furthermore, in other preferred embodiments of the present invention, the optimal bandwidth for each prospecting factor is determined by searching using the golden section method and minimizing the Akaike information criterion.
[0010] Furthermore, in other preferred embodiments of the present invention, in step S1, high-confidence data points are selected based on local goodness of fit, and steps S2 and S3 are performed only on data points whose local goodness of fit is higher than a preset threshold.
[0011] Furthermore, in other preferred embodiments of the present invention, the mineral exploration factors include environmental factors, topographic factors, soil factors, and geological factors; the rare earth element content includes the content of element La and the content of element Y.
[0012] Furthermore, in other preferred embodiments of the present invention, in step S2, the strong influence interval is divided by performing quantile statistics on the local regression coefficients of the mineral exploration factors, including strong positive correlation intervals and strong negative correlation intervals.
[0013] Furthermore, in other preferred embodiments of the present invention, the optimal mineralization value is determined by using the Gaussian kernel density estimation method to determine the value of the prospecting factor corresponding to the peak probability density within the strong influence interval.
[0014] Furthermore, in other preferred embodiments of the present invention, in step S3, the degree to which the value of the mineral exploration factor approaches the optimal mineralization value is calculated using a Gaussian score function: , In the formula, This indicates that the mineral exploration factor takes the value of x The probability density estimate at time , x This indicates the value of the mineral exploration factor to be evaluated. x i Indicates the first i The actual values of the mineral exploration factors at each sampling point. n This indicates the number of sample points involved in kernel density estimation. h This represents the bandwidth parameter of the kernel function.
[0015] Furthermore, in other preferred embodiments of the present invention, in step S3, the spatial distribution map of mineral exploration potential is generated based on a comprehensive mineral exploration potential index, the expression of which is: , In the formula, MPI Indicates the mineral exploration potential index; LocalR j 2 For local goodness of fit; b i ols These are the global regression coefficients; β ji For mineral exploration factors i In position j Local regression coefficients; S i For mineral exploration factors i In position j Gauss score; n The number of mineral exploration factors.
[0016] Furthermore, in other preferred embodiments of the present invention, it also includes: S4. Conduct field verification of the delineated target area.
[0017] The beneficial effects of the embodiments of the present invention are: This invention provides a method for ion-adsorption rare earth mineral exploration based on spatial heterogeneity modeling. The method involves characterizing the spatial nonstationar influence of various exploration factors on rare earth element enrichment using a multi-scale geographically weighted regression model, obtaining local regression coefficients and local goodness of fit. Then, based on these parameters, the strong influence intervals of the mineralization factors are identified, and the optimal mineralization values of each factor are inverted using data-driven methods such as probability density estimation. Finally, the local regression coefficients, goodness of fit, and the degree to which factors approach their optimal values are integrated to construct a mineral potential index, generate a spatial distribution map, and delineate target areas. This method accurately characterizes the complex spatial heterogeneity relationship between mineralization factors and rare earth enrichment, solving the problem that traditional models struggle to reflect differences in local mineralization mechanisms. Furthermore, by quantitatively inverting the optimal mineralization values, it replaces subjective experience judgments and provides an objective basis for assessing mineralization conditions. It significantly improves the accuracy and reliability of mineral exploration prediction and has considerable practical value. Attached Figure Description
[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.
[0019] Figure 1 A flowchart of an ion adsorption rare earth mineral exploration method based on spatial heterogeneity modeling provided as an application example of the present invention. Figure 2 Spatial distribution map of mineral exploration potential provided as an application example of the present invention; Figure 3 The weathering crust layering characteristics and rapid field test verification diagram of verification point No. 2 provided as an application example of the present invention; Figure 4 The graph shows the total rare earth content test results of sample No. 2 at verification point, which is provided as an application example of the present invention. Detailed Implementation
[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. 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. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to represent selected embodiments of the invention. 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.
[0021] The abbreviations and key terms used in the following embodiments are defined as follows: Example
[0022] This embodiment provides an ion adsorption-based rare earth mineral exploration method based on spatial heterogeneity modeling, which includes: S1. Obtain data on mineral exploration factors and rare earth element content in the study area, and use a multi-scale geographically weighted regression model to analyze the spatial heterogeneity of each mineral exploration factor on rare earth element enrichment, and obtain the local regression coefficients and local goodness of fit of the mineral exploration factors at each spatial location.
[0023] Among these, prospecting factors include environmental factors, topographic factors, soil factors, and geological factors. Optionally, environmental factors include, but are not limited to, average annual rainfall, average annual temperature, and vegetation cover index; topographic factors include, but are not limited to, altitude, slope, topographic relief, topographic roughness, and surface incision depth; soil factors include, but are not limited to, chemical alteration index and silica-alumina-iron ratio; and geological factors include, but are not limited to, tectonic density and river density. These four categories of factors, from the aspects of weathering and leaching conditions, weathering crust development and preservation conditions, rare earth adsorption and enrichment material basis, and element migration and redistribution channels, jointly characterize the mineralization control mechanism of ion-adsorption type rare earth deposits.
[0024] The rare earth element content includes the content of element La and element Y. Element La represents light rare earth elements, while element Y represents heavy rare earth elements. The data comes from regional-scale geochemical data.
[0025] Furthermore, a multi-scale geographically weighted regression model uses the contents of rare earth elements La and Y as dependent variables and mineral exploration factors as independent variables to perform regression analysis on the spatial response relationship between mineral exploration factors and rare earth elements. This multi-scale geographically weighted regression model independently assigns its spatial bandwidth to each mineral exploration factor. Optionally, the kernel function of the multi-scale geographically weighted regression model is an adaptive bisquared Gaussian function, and the spatial bandwidth of each mineral exploration factor is searched using the golden section method and automatically determined by minimizing the Akaike Information Criterion (AIC) correction value.
[0026] Optionally, high-confidence data points are selected based on local goodness of fit, and steps S2 and S3 are performed only on data points with a local goodness of fit higher than a preset threshold. The preset threshold is 0.8, and sampling points with a local goodness of fit greater than 0.8 are considered high-confidence mineralization response points.
[0027] Furthermore, the ion adsorption-based rare earth mineral exploration method based on spatial heterogeneity modeling provided in this embodiment also includes: S2. Based on local regression coefficients and local goodness of fit, identify the strong influence range of each prospecting factor on mineralization, and use data-driven inversion to find the optimal mineralization value of each prospecting factor.
[0028] Furthermore, the strong influence interval is defined by performing quantile statistics on the local regression coefficients of the mineral exploration factors, including strong positive correlation intervals and strong negative correlation intervals.
[0029] Within the strongly positively correlated region, sample points with larger factor values were selected, while within the strongly negatively correlated region, sample points with smaller factor values were selected. Gaussian kernel density estimation was used to model the probability distribution of the factor values for both types of sample points. The optimal mineralization value within the corresponding region was determined by the position of the main peak of the kernel density function, and the probability density value at the main peak was used as the distribution intensity index.
[0030] Optionally, the strong negative correlation range is from the minimum value of the local regression coefficient to the 25th percentile, and the strong positive correlation range is from the 75th percentile to the maximum value of the local regression coefficient.
[0031] Furthermore, the optimal mineralization value is determined by using the Gaussian kernel density estimation method to identify the mineralization factor corresponding to the peak probability density within the strong influence interval. This is performed using the following formula: , In the formula, This indicates that the mineral exploration factor takes the value of x The probability density estimate at time , x This indicates the value of the mineral exploration factor to be evaluated. x i Indicates the first iThe actual values of the mineral exploration factors at each sampling point. n This indicates the number of sample points involved in kernel density estimation. h This represents the bandwidth parameter of the kernel function.
[0032] The optimal mineralization value within the corresponding interval is determined by the position of the main peak of the kernel density function. That is, the mineralization factor value corresponding to the position of the maximum probability density is the optimal mineralization value within the corresponding interval.
[0033] The comprehensive optimal mineralization value for each prospecting factor is calculated by weighted fusion of the optimal mineralization values in strongly positively correlated and strongly negatively correlated intervals, based on the following formula: , In the formula, β ( x ) represents mineral exploration factors x The comprehensive optimal mineralization value; x + and x - These represent the optimal mineralization values for the strongly positively correlated interval and the strongly negatively correlated interval, respectively. ϵ + and ϵ - These represent the distribution intensity of optimal mineralization values in the strongly positively correlated interval and the strongly negatively correlated interval, respectively.
[0034] Furthermore, the ion adsorption-based rare earth mineral exploration method based on spatial heterogeneity modeling provided in this embodiment also includes: S3. By integrating local regression coefficients, local goodness of fit, and the degree to which the mineral exploration factor values of each location are close to their optimal mineralization values, a spatial distribution map of mineral exploration potential is generated, and target areas are delineated accordingly.
[0035] The spatial distribution map of mineral exploration potential is generated based on a comprehensive mineral exploration potential index, which is expressed as follows: , In the formula, MPI Indicates the mineral exploration potential index; LocalR j 2 For local goodness of fit; b i ols For mineral exploration factors i The ordinary least squares global regression coefficients; β ji For mineral exploration factors i In position j Local regression coefficients; S i For mineral exploration factors iIn position j Gauss score; n The number of mineral exploration factors.
[0036] in, S i The formula for calculation is: , In the formula, S i Indicating mineral exploration factors i Gauss score; x i These are the actual mineral exploration factor values; u i This represents the optimal mineralization value for the prospecting factor. σ i The standard deviation of the factor is 1.
[0037] Furthermore, the ion adsorption-based rare earth mineral exploration method based on spatial heterogeneity modeling provided in this embodiment also includes: S4. Conduct field verification of the delineated target area.
[0038] Optionally, the verification method includes: In the anomaly area, samples were subjected to leaching-titration analysis using ammonium sulfate and oxalic acid solutions. The mineralization intensity was initially and rapidly determined based on solution discoloration and turbidity. Further drilling and sampling analysis were then conducted, and the layering characteristics and mineralization status were recorded on-site. Finally, the samples were sent to a laboratory for testing, and the final grade of the ore body was determined based on the test results.
[0039] Application examples This application example uses the ion adsorption-based rare earth mineral exploration method based on spatial heterogeneity modeling provided in Example 1, and implements it in a region of western Yunnan. The process is as follows. Figure 1 As shown, it specifically includes: (1) Based on the metallogenic geological background, the original multi-source data of the region were preprocessed to extract multiple mineralization factors related to the mineralization of ion adsorption type rare earth minerals, and a mineral exploration factor system was constructed, including environmental factors, topographic factors, soil weathering factors and geological factors. The content data of rare earth elements La and Y were obtained, wherein the rare earth element content was obtained from 1:200,000 stream sediment geochemical data. (2) Using the contents of rare earth elements La (light rare earth element) and Y (heavy rare earth element) as dependent variables and the mineral exploration factors as independent variables, a multi-scale geographical weighted regression model is established to perform regression analysis on the spatial response relationship between mineral exploration factors and rare earth elements. The local regression coefficients and local goodness of fit of each mineral exploration factor are obtained through adaptive bandwidth optimization.
[0040] Among them, the La element regression model results are: AIC value of 17937.53; R 2 The value is 0.80. The exploration factors involved and their optimal bandwidths are shown in Table 1.
[0041] Table 1. Mineral exploration factors and their optimal bandwidth in the La element regression model variable Rainfall Temperature NDVI altitude slope Topographic relief Surface cutting depth River density Elevation variation coefficient Chemical alteration index silicon-aluminum-iron ratio bandwidth 44 354 484 355 520 449 326 701 305 424 421 Y-element regression model results: AIC value is 16826.02; R 2 The value is 0.80. The optimal bandwidth for the mineral exploration factor is shown in Table 2.
[0042] Table 2. Mineral exploration factors and their optimal bandwidth in the La element regression model. variable Rainfall Temperature NDVI altitude slope Topographic relief Surface cutting depth River density Elevation variation coefficient Chemical alteration index silicon-aluminum-iron ratio bandwidth 44 2145 464 537 1006 475 712 891 652 673 1238 (3) The reliability of the regression results is screened based on the local goodness of fit. The sampling points with local goodness of fit greater than the preset threshold are selected as high reliability mineralization response points. For the high reliability mineralization response points, the local regression coefficients of each mineral exploration factor are statistically analyzed to divide the strong positive correlation range and the strong negative correlation range.
[0043] The strong positive and strong negative correlation intervals of the La element prospecting factors are shown in Table 3.
[0044] Table 3. Strong positive and negative correlation intervals of La mineral exploration factors Mineral exploration factors negative correlation interval Positive correlation interval Elevation variation coefficient [0.001, 0.004] [0.008, 0.022] silicon-aluminum-iron ratio [1.47, 2.43] [2.90, 5.30] Chemical alteration index [39, 76] [80, 85] Altitude (m) [849.95, 1376.58] [1747.47, 2278.88] NDVI [0.23, 0.63] [0.70, 0.83] Rainfall (mm / yr) [933.92, 971.11] [1556.35, 1598.58] Slope (°) [11.15, 16.32] [20.67, 27.88] Temperature (°C) [11.46, 14.87] [20.28, 22.37] River density (m / km 2 ))]]> [171.89, 985.30] [1909.25, 5991.54] Surface cutting depth (m) [14.75, 28.11] [43.94, 63.70] Topographic relief (m) [40.75, 65.65] [85.73, 140.30] The strong positive and strong negative correlation intervals of the Y element prospecting factors are shown in Table 4.
[0045] Table 4. Strong positive and negative correlation intervals of Y element prospecting factors factor negative correlation interval Positive correlation interval Elevation variation coefficient [0.0003,0.0040] [0.008,0.019] silicon-aluminum-iron ratio [1.55,2.21] [3.04,4.00） Chemical alteration index [68,76] [79,85] Altitude (m) [831.58,1371.75] [2189.06,2891.63] NDVI [0.32,0.65] [0.69,0.84] Rainfall (mm / yr) [938.45,1512.43] [1345.43,1471.21] Slope (°) [5.71,13.53] [22.57,28.17] Temperature (°C) [15.84,19.26] [18.89,21.65] <![CDATA[River density (m / km 2 )]]> [80.22,825.54] [2091.08,5555.19] Surface cutting depth (m) [16.11,32.33] [34.09,52.38] Topographic relief (m) [17.79,51.92] [83.24,113.53] (4) Within the strongly positively correlated range, sample points with larger factor values are selected, and within the strongly negatively correlated range, sample points with smaller factor values are selected. The Gaussian kernel density estimation method is used to model the probability distribution of the factor values for both types of sample points. The optimal mineralization value within the corresponding range is determined by the position of the main peak of the kernel density function, and the probability density value at the main peak is used as the distribution intensity index. Based on this, the comprehensive optimal mineralization value for each prospecting factor is calculated.
[0046] The optimal comprehensive mineralization value of the La element prospecting factor is shown in Table 5.
[0047] Table 5. Comprehensive optimal mineralization value of La mineral exploration factors factor Optimal mineralization value Elevation variation coefficient 0.01 silicon-aluminum-iron ratio 2.65 Chemical alteration index 77 Altitude (m) 1498.78 NDVI 0.66 Rainfall (mm / yr) 1265.24 Slope (°) 19.36 Temperature (°C) 17.51 <![CDATA[River density (m / km 2 )]]> 1395.69 Surface cutting depth (m) 37.69 Topographic relief (m) 75.71 The optimal comprehensive mineralization values of Y element prospecting factors are shown in Table 6.
[0048] Table 6. Comprehensive Optimal Metallogenic Values of Y Element Prospecting Factors factor Optimal mineralization value Elevation variation coefficient 0.01 silicon-aluminum-iron ratio 2.63 Chemical alteration index 0.78 Altitude (m) 1788.88 NDVI 0.67 Rainfall (mm / yr) 1412.73 Slope (°) 18.04 Temperature (°C) 19.89 <![CDATA[River density (m / km 2 )]]> 1621.44 Surface cutting depth (m) 32.68 Topographic relief (m) 67.04 (5) Based on the optimal mineralization value, the Gaussian score of each mineral exploration factor in the study area is calculated using a Gaussian function. The mineral exploration potential index is constructed by combining the local goodness of fit, local regression coefficient, global regression coefficient, and factor Gaussian score, and a spatial distribution map of the mineral exploration potential is generated (e.g., ...). Figure 2 (As shown).
[0049] (6) Based on the spatial distribution map of the prospecting potential, spatial prediction of the prospecting potential of ion adsorption type rare earth minerals in the study area is carried out, anomaly areas are delineated, and verification work is carried out.
[0050] The validation work included leaching and titration analysis of samples in the anomaly area using ammonium sulfate and oxalic acid solutions. Based on the solution color change and turbidity, a preliminary and rapid assessment of mineralization intensity was made. Further drilling and sampling analysis were conducted, and the layering characteristics and mineralization status were recorded on-site. Figure 3 For example, drilling and sampling were carried out at a depth of 24m in the No. 2 borehole. The section from 6 to 20m was completely weathered. The soil samples from 13.2 to 20m were subjected to leaching titration. The solution was snowflake-like and flocculent. After standing, there was obvious precipitate, indicating the presence of rare earth elements.
[0051] In this study, four sites were selected for borehole sampling verification, denoted as ZK1 to ZK4. Multiple samples were taken from each sampling point in order from shallow to deep. The samples were sent to the laboratory for testing, and the final grade of the ore body was determined based on the test results. The sampling verification results are shown in Tables 7 to 10.
[0052] Table 7. Sampling verification results of ZK1 sampling points
[0053] Table 8. Sampling verification results of ZK2 sampling points
[0054] Table 9. Sampling verification results of ZK3 sampling points
[0055] Table 10. Sampling verification results of ZK4 sampling points
[0056] As shown in Tables 7-10, the prospecting potential index of boreholes No. 2 and No. 4 both reached above 0.85, indicating relatively rich rare earth content in the samples collected in this area. The mineable thickness of borehole No. 2 reached 8.8m, and that of borehole No. 4 reached 4.6m. The prospecting potential index of borehole No. 3 was 0.75, with relatively low rare earth content and a mineable thickness of only 0.8m. Borehole No. 1 had the lowest prospecting potential index at only 0.68. Although it contained some rare earth content, none of it reached above 0.08%, and the mineable thickness was 0m. Therefore, the prospecting potential index of this embodiment basically matches the laboratory verification results, and the ion adsorption rare earth prospecting method of this embodiment has good accuracy.
[0057] Furthermore, to more intuitively analyze the rare earth content in the sample, a histogram can be created from the sample data of well number two, such as... Figure 4 As shown in the figure, the total rare earth content in samples ZK2-10 to ZK2-20 all exceeded 0.08%, meeting the industrial grade requirements and possessing mining value. Among them, sample ZK2-18 had the highest total rare earth content at 0.165%.
[0058] In summary, this invention provides a method for ion-adsorption rare earth mineral exploration based on spatial heterogeneity modeling. This method involves characterizing the spatial nonstationar influence of various exploration factors on rare earth element enrichment using a multi-scale geographically weighted regression model, obtaining local regression coefficients and local goodness of fit; then, identifying the strong influence intervals of ore-forming factors based on these parameters, and inverting the optimal ore-forming values of each factor using data-driven methods such as probability density estimation; finally, integrating the local regression coefficients, goodness of fit, and the degree to which factors approach their optimal values to construct a mineral potential index, generate a spatial distribution map, and delineate target areas. This method accurately characterizes the complex spatial heterogeneity relationship between ore-forming factors and rare earth enrichment, solving the problem that traditional models struggle to reflect differences in local mineralization mechanisms; and by quantitatively inverting the optimal ore-forming values, it replaces subjective experience judgment, providing an objective basis for assessing mineralization conditions. It significantly improves the accuracy and reliability of mineral exploration prediction and has considerable practical value.
[0059] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A rare earth mineral exploration method based on spatial heterogeneity modeling using ion adsorption, characterized in that, include: S1. Obtain data on mineral exploration factors and rare earth element content in the study area, and use a multi-scale geographical weighted regression model to analyze the influence of each mineral exploration factor on the spatial heterogeneity of rare earth element enrichment, and obtain the local regression coefficient and local goodness of fit of the mineral exploration factor at each spatial location. S2. Based on the local regression coefficients and the local goodness of fit, identify the strong influence range of each of the prospecting factors on mineralization, and invert the optimal mineralization value of each of the prospecting factors in a data-driven manner. S3. By integrating the local regression coefficients, the local goodness of fit, and the degree to which the mineral exploration factor values of each location are close to their optimal mineralization values, a spatial distribution map of mineral exploration potential is generated, and the target area is delineated accordingly.
2. The ion adsorption-based rare earth mineral exploration method according to claim 1, characterized in that, In step S1, the multi-scale geographically weighted regression model independently assigns its spatial bandwidth to each of the mineral exploration factors.
3. The ion adsorption-based rare earth mineral exploration method according to claim 2, characterized in that, The optimal bandwidth for each prospecting factor is determined by searching using the golden section method and minimizing the Akaike information criterion.
4. The ion adsorption-based rare earth mineral exploration method according to claim 3, characterized in that, In step S1, high-confidence data points are selected based on the local goodness of fit, and steps S2 and S3 are performed only on data points whose local goodness of fit is higher than a preset threshold.
5. The ion adsorption-based rare earth mineral exploration method according to claim 1, characterized in that, The prospecting factors include environmental factors, topographic factors, soil factors, and geological factors; the rare earth element content includes the content of element La and element Y.
6. The ion adsorption-based rare earth mineral exploration method according to claim 1, characterized in that, In step S2, the strong influence interval is divided by performing quantile statistics on the local regression coefficients of the mineral exploration factors, including strong positive correlation intervals and strong negative correlation intervals.
7. The ion adsorption-based rare earth mineral exploration method according to claim 1, characterized in that, The optimal mineralization value is obtained by using the Gaussian kernel density estimation method to determine the value of the prospecting factor corresponding to the peak probability density within the strong influence range.
8. The ion adsorption-based rare earth mineral exploration method according to claim 7, characterized in that, In step S3, the degree to which the value of the mineral exploration factor approaches the optimal mineralization value is calculated using a Gaussian score function: ， In the formula, This indicates that the value of the mineral exploration factor is... x The probability density estimate at time , x This indicates the value of the mineral exploration factor to be evaluated. x i Indicates the first i The actual values of the prospecting factors at each sampling point. n This indicates the number of sample points involved in kernel density estimation. h This represents the bandwidth parameter of the kernel function.
9. The ion adsorption-based rare earth mineral exploration method according to claim 1, characterized in that, In step S3, the spatial distribution map of mineral exploration potential is generated based on a comprehensive mineral exploration potential index, the expression of which is: ， In the formula, MPI This indicates the mineral exploration potential index; LocalR j 2 The local fit goodness; b i ols These are the global regression coefficients; β ji For the mineral exploration factor i In position j The local regression coefficients; S i For the mineral exploration factor i In position j Gauss score; n The number of the prospecting factors.
10. The ion adsorption-based rare earth mineral exploration method according to claim 1, characterized in that, Also includes: S4. Conduct field verification of the delineated target area.