Geochemical composition data factor analysis noise reduction method

Abstract

The present application provides a geochemical composition data factor analysis noise reduction method, belonging to the field of geological mineral exploration. The method is to collect or collect the relevant geochemical composition data of the study area, to carry out logarithmic ratio transformation and normal distribution test, to apply statistical analysis to the obtained geochemical composition data, to calculate geochemical parameters, to construct noise reduction standard 1, to carry out preliminary noise reduction based on noise reduction standard 1, to carry out first factor analysis, to construct noise reduction standard 2, to carry out secondary noise reduction based on noise reduction standard 2, to carry out second factor analysis, to construct noise reduction standard 3, to carry out third factor analysis based on noise reduction standard 3, to test whether the third factor analysis result meets the noise reduction standard 2 and the noise reduction standard 3, and to continue until the factor analysis result does not meet the noise reduction standard 2 and the noise reduction standard 3, and then the factor analysis result is the final result. The method can be used for regional geochemical exploration data dimensionality reduction and noise reduction analysis, and greatly improves the data utilization efficiency.

Description

Technical Field

[0001] This invention belongs to the field of geological and mineral exploration, and specifically relates to a method for noise reduction of geochemical composition data through factor analysis. Background Technology

[0002] Geochemical data, as a typical type of compositional data, combined with factor analysis, can serve as a direct means of obtaining regional mineralization information and has played an important role in the mineral exploration process. Moreover, long-term geochemical mapping work has accumulated a large amount of high-quality, multi-scale, and multi-media regional geochemical data, providing a rich data foundation for regional geochemical exploration.

[0003] Regional geochemical composition data contains multiple elemental indicators and exhibits equivalence, closed-loop properties, and simplex spatial characteristics, making direct interpretation of geochemical data challenging. Factor analysis, a statistical technique for extracting common factors from a group of variables, is a commonly used method for dimensionality reduction in high-dimensional geochemical composition data. It also serves as a valuable tool for identifying and quantitatively evaluating regional geochemical anomalies, exploring the differences in the ratios of individual components to the overall vector across datasets and uncovering hidden intrinsic connections between data. The principle of geochemical data factor analysis is that once the analytical factors (elemental combinations) are given objective and reasonable geological and mineralization interpretations, the score of each factor for each sample becomes a tracer of the geological and mineralization information of that sample. Therefore, the factor with the highest score can be selected to represent the potential mineralization information of that sample.

[0004] Due to the complexity of regional geology and geochemistry, such as the complexity and heterogeneity of surrounding rocks, the multi-stage nature of mineralization, and surface weathering, elements in geochemical composition data that could originally reflect key information are often obscured by non-critical or secondary elements, thus failing to be reflected in factor analysis results. This necessitates noise reduction processing in factor analysis to ensure that the extracted factors effectively represent the geological and geochemical processes of the study area. Therefore, noise reduction processing in the factor analysis of geochemical composition data is a necessary step to improve data utilization efficiency and effectively extract factors indicative of mineralization.

[0005] In the process of denoising geochemical composition data, there are two significant types of "noise": elements with low variability and minimal impact from different geological and geochemical processes; and elements significantly affected by multiple factors to varying degrees, leading to reduced representativeness and significance. The first type of "noise" needs to be removed based on geochemical statistical indicators during factor analysis, while the second type of "noise" requires further denoising based on the results of factor analysis.

[0006] Current methods for denoising geochemical composition data through factor analysis only address noise reduction at the mathematical and geological level, without a deep understanding of the "noise" mechanism. This results in complex operation and generally poor noise reduction performance. Summary of the Invention

[0007] The purpose of this invention is to provide a noise reduction method for geochemical composition data factor analysis. This method effectively reduces the dimensionality of geochemical composition data and extracts factors that can indicate geological and geochemical processes by performing noise reduction processing on geochemical composition data, thus providing a basis for delineating geochemical anomalies in the study area.

[0008] To achieve the above objectives, the present invention provides a method for noise reduction of geochemical composition data through factor analysis, comprising the following steps:

[0009] Step 1: Collect or gather geochemical composition data related to the study area;

[0010] Step 2: Perform a logarithmic ratio transformation on the geochemical composition data to remove the "closing effect" of the geochemical composition data, and then perform a normal distribution test on the geochemical composition data after the logarithmic ratio transformation.

[0011] Step 3: Apply statistical analysis to the geochemical composition data obtained in Step 1, calculate geochemical parameters, and select specific geochemical parameters to construct noise reduction standard 1; noise reduction standard 1 is to remove elements below the threshold of the selected geochemical parameters.

[0012] Step 4: Perform preliminary denoising on the geochemical composition data that conforms to a normal distribution based on the denoising standard 1 constructed in Step 3; perform the first factor analysis on the geochemical composition data after preliminary denoising; and construct denoising standard 2 based on the results of the first factor analysis.

[0013] Step 5: Perform secondary noise reduction based on the noise reduction standard 2 constructed in Step 4, conduct a second factor analysis on the geochemical composition data after secondary noise reduction, and construct noise reduction standard 3 based on the results of the second factor analysis.

[0014] Step 6: Based on the noise reduction standard 3 constructed in Step 5, perform noise reduction again, and conduct a third factor analysis on the geochemical composition data after noise reduction.

[0015] Step 7: Check whether the result of the third factor analysis in step 6 meets noise reduction criteria 2 and 3. If it does, iterate steps 5 and 6 until the factor analysis results do not meet noise reduction criteria 2 and 3. Then the factor analysis result is the final result.

[0016] Furthermore, in step 1, the geochemical composition data includes geochemical data of regional soil, rocks, stream sediments, geothermal data, and vegetation, as well as their component data. The component data includes soil fine particles and active phase extraction.

[0017] Furthermore, in step 1, the consistency of sample types and data quality of the collected geochemical composition data should be checked. For the collected samples, it should be ensured that the sample types and representativeness are consistent. When analyzing and testing geochemical composition data, efforts should be made to ensure that the analysis is carried out in the same laboratory to avoid systematic errors caused by different laboratories, so as to ensure the reliability of the analysis results.

[0018] Because geochemical composition data exhibits a "closing effect," a logarithmic ratio transformation is performed on the geochemical composition data to remove this effect before applying statistical analysis. Furthermore, since factor analysis of geochemical composition data requires the assumption that the data conforms to a normal distribution, a normality test is conducted on the data.

[0019] The closure effect means that the total amount of all geochemical components (elemental content) equals 1 (100%), and the components are mutually restrictive, exhibiting negative or positive correlations. These correlations are spurious and have no geological significance. This is a common problem encountered in the analysis of geochemical composition data.

[0020] Furthermore, in step 3, the selected geochemical parameters are the enrichment coefficient and the coefficient of variation.

[0021] Furthermore, in step 3, the specific steps for constructing noise reduction standard 1 include:

[0022] (1) Calculate the coefficient of variation and enrichment coefficient of each element in the study area;

[0023] (2) Based on the coefficient of variation and enrichment coefficient of each element in the study area, a noise reduction standard 1 is constructed, that is, elements below a certain coefficient of variation and / or enrichment coefficient threshold are removed.

[0024] Enrichment coefficient and coefficient of variation are two commonly used parameters for measuring the degree of enrichment and variation of a measured element in a region. They are affected by a variety of factors, including bedrock, weathering, mineralization, and sampling particle size. However, if an element has both a large enrichment coefficient and a large coefficient of variation, it may be closely related to mineralization.

[0025] Elements with high variability typically represent specific geological and geochemical processes and are therefore key research subjects. Elements with high enrichment levels often indicate potential involvement in mineralization and are also key research subjects. Elements with variability and / or enrichment coefficients below specific thresholds—that is, elements influenced by multiple geological and geochemical factors and lacking representativeness and significance—cannot indicate specific geological and geochemical processes and are considered "noise." These elements need to be discarded during factor analysis to ensure that the extracted factors have clear geological and geochemical meanings.

[0026] Furthermore, in step 4, the noise reduction standard 2 is to remove factors composed of single elements, that is, to remove elements that are significantly affected by a certain geological and geochemical process.

[0027] Furthermore, in step 5, the noise reduction standard 3 is to remove elements that have high loads in two or more factors but whose loads are all below the determined tuple combination threshold, and to extract element combinations in the factors whose loads are greater than the determined tuple combination threshold. That is, to remove elements that are affected by multiple geological and geochemical processes and whose influence is comparable, and to extract representative factors that are indicative of the geology, geochemistry and mineralization processes of the study area.

[0028] An element that is influenced by multiple geological and geochemical processes to a similar degree indicates that it does not dominate any single geological and geochemical process and therefore cannot significantly indicate the geological and geochemical processes implied in the geochemical data.

[0029] The beneficial technical effects of this invention are as follows:

[0030] 1. The method of this invention constructs a denoising method for geochemical composition data by factor analysis, and completes factor analysis denoising processing for different types of geochemical composition data, providing a basis for extracting mineralization-related element combinations from geochemical composition data by factor analysis.

[0031] 2. The method of the present invention can be used for dimensionality reduction and noise reduction analysis of regional geochemical data, so that the extracted factors have more explicit geological and geochemical significance, and greatly improve the data utilization efficiency.

[0032] 3. The method of this invention is mainly based on a combination of data-driven and knowledge-driven approaches to perform multiple noise reduction processes in the geochemical composition data factor analysis process, which is highly operable. Currently, this method has been used to perform noise reduction processing on geochemical composition data factor analysis of soil fine particles, core geochemical data of Qujia gold mine, and soil fine particle geochemical data of Yueyang Basin surrounding Zijinshan mining area. The extracted factors have clear geological and geochemical significance and have a great promoting effect on mineral exploration. Attached Figure Description

[0033] Figure 1 The flowchart illustrates a method for noise reduction in geochemical composition data factor analysis provided by this invention. Detailed Implementation

[0034] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0035] Example

[0036] In this embodiment, taking the Qujia hidden gold mine research area in Laizhou, Jiaodong Peninsula, my country as an example, the method for denoising geochemical composition data factor analysis provided by the present invention is further described in detail.

[0037] The principle for judging the noise reduction effect is that the element combination extracted by factor analysis has a clearer geological and geochemical significance, that is, it can effectively indicate the bedrock and mineralization processes in the study area. For example, the element combination that indicates mineralization after noise reduction in the Qujia concealed gold deposit area needs to have a high degree of consistency with the known mineralization and associated elements (Au, Ag, Hg, Bi, Cu, Pb, Zn, As, Sb).

[0038] Combination Figure 1 This embodiment provides a method for denoising geochemical composition data through factor analysis, and the specific steps are as follows:

[0039] Step 1: In the study area of ​​the Qujia hidden gold deposit in Laizhou, Jiaodong, 1:50,000 soil fine-particle geochemical data were collected. The sample types, representativeness and analytical tests showed high consistency.

[0040] Step 2: The 1:50000 soil fine-particle geochemical data were subjected to central log ratio (clr) transformation using CoDaPack software to open the "closing effect". The normality test was performed on the 1:50000 soil fine-particle geochemical data after the central log ratio transformation to confirm that the 1:50000 soil fine-particle geochemical data conforms to a normal distribution.

[0041] Step 3 involves descriptive applied statistical analysis of the 1:50,000 soil fine-grained geochemical data, and calculation of geochemical indices reflecting data variability and enrichment levels—the coefficient of variation and enrichment coefficient. The coefficient of variation is calculated as the standard deviation divided by the mean; the enrichment coefficient is calculated as the mean divided by the abundance of that element in Chinese soils. Noise reduction criterion 1 is defined as follows: if an element shows neither variability (coefficient of variation < 0.75) nor enrichment (enrichment coefficient < 1.11), then that element is considered "noise" and removed. Noise reduction criterion 1 can be adjusted appropriately based on the research objectives, with the coefficient of variation adjusted within the range of 0.9–1.25 and the enrichment coefficient within the range of 0.45–1.11.

[0042] Step 4: Preliminary noise reduction is performed on the 1:50000 soil fine-particle geochemical data that conforms to a normal distribution, based on the noise reduction standard 1 constructed in Step 3. The elements belonging to noise include As, Cr, Co, Ni, Mn, Mo, Sb, Se, Sn, V, Ti, W, SiO2, Al2O3, Fe2O3, MgO, and F. The remaining elements (after clr transformation) are subjected to the first factor analysis. The results of the first factor analysis are shown in Table 1. The threshold used to extract the element combinations is 0.6.

[0043] Table 1 Results of the first factor analysis

[0044] element F1 F2 F3 Au 0.53 0.20 -0.55 Ag 0.78 0.27 -0.38 Ba -0.39 -0.81 -0.14 Bi 0.68 0.52 -0.11 Cd 0.83 0.26 0.17 Cu 0.69 0.64 -0.06 Hg 0.75 0.23 -0.12 Pb 0.63 -0.59 -0.17 Zn 0.81 0.24 0.10 S 0.69 0.48 0.16 Sr -0.27 -0.75 0.15 Cl 0.37 0.56 0.05 Br 0.55 0.65 0.19 I 0.55 0.55 0.19 CaO 0.12 0.25 0.79 <![CDATA[Na2O]]> -0.20 -0.82 -0.11 <![CDATA[K2O]]> -0.15 -0.93 -0.04 Eigenvalues 5.66 5.41 1.33 Variance contribution rate % 33.29 31.82 7.81 Cumulative variance contribution rate % 33.29 65.10 72.91

[0045] In Table 1, F1-F3 represent three different factors. The first column lists the different variables (elements in this case), and the values ​​in the table represent the loadings of the elements in the corresponding factors. A threshold (absolute value 0.6) is typically set to determine the element composition of each factor. If the absolute value of an element's loading in a factor is greater than 0.6, then that element belongs to that factor.

[0046] Eigenvalues ​​refer to the eigenvalues ​​of a matrix and the factor loadings in the corresponding eigenvectors. In factor analysis, in order to determine the number and properties of latent factors, it is necessary to perform eigenvalue decomposition on the original variables and determine the final number of factors and the meaning of the factors based on the magnitude of the eigenvalues ​​and the factor loadings of the eigenvectors. Factor eigenvalues ​​can be used to calculate the variance contribution rate and cumulative variance contribution rate of factors, as well as to perform further factor analysis operations such as factor rotation and factor score estimation.

[0047] The variance contribution rate represents the sum of the variance contributions of the same common factor F to all variables, and is used to measure the relative importance of each common factor. It is the proportion of variation caused by a single common factor to the total variation.

[0048] The cumulative variance contribution rate is the proportion of the total variance caused by all common factors, indicating the combined influence of all common factors on the dependent variable.

[0049] The results of the first factor analysis in Table 1 show that the F1 element combination extracted in step 4 that can indicate mineralization is Ag-Bi-Cd-Cu-Hg-Pb-Zn-S. However, the main mineralizing elements are displayed in any factor due to interference from other noise elements.

[0050] Based on the results of the first factor analysis, noise reduction standard 2 was constructed. Among them, factor F3 consists only of CaO, which has weak indicative significance for geological and geochemical processes and is therefore noise, so it was removed.

[0051] Step 5: Perform secondary noise reduction based on the noise reduction standard 2 constructed in Step 4, and perform a second factor analysis on the remaining elements. The results of the second factor analysis are shown in Table 2.

[0052] Table 2 Results of the second factor analysis

[0053] element F1 F2 F3 Au 0.08 0.11 0.92 Ag 0.15 0.51 0.74 B -0.78 -0.49 -0.09 Bi 0.45 0.49 0.57 Cd 0.22 0.83 0.25 Cu 0.57 0.59 0.47 Hg 0.15 0.68 0.39 Pb -0.66 0.49 0.32 Zn 0.18 0.83 0.23 S 0.47 0.54 0.45 Sr -0.67 -0.44 -0.02 Cl 0.55 0.27 0.31 Br 0.64 0.59 0.17 I 0.53 0.65 0.06 <![CDATA[Na2O]]> -0.81 -0.16 -0.22 <![CDATA[K2O]]> -0.91 -0.12 -0.23 Eigenvalues 4.84 4.55 2.76 Variance contribution rate % 30.22 28.45 17.24 Cumulative variance contribution rate % 30.22 58.67 75.91

[0054] The results of the second factor analysis in Table 2 show that the elemental combination of the factors extracted in step 5 that indicate mineralization is Au-Ag. However, due to the high activity of Au and Ag under surface conditions, the correspondence between Au-Ag anomalies and deep concealed minerals is poor.

[0055] Based on the results of the second factor analysis, noise reduction standard 3 was constructed. The loadings of Bi, Cu, S, and Cl in any one factor are all less than the threshold, but they have high loadings in multiple factors. This reflects that these elements are affected by a variety of geological and geochemical processes and the degree of influence is quite similar, which also constitutes noise.

[0056] Step 6: Perform a third factor analysis on the remaining elements. The results of the third factor analysis are shown in Table 3. The results of the third factor analysis do not meet either noise reduction criteria 2 or 3. Therefore, the results of the third factor analysis are the final results (Table 3).

[0057] Table 3 Results of the third factor analysis

[0058] element F1 F2 Au -0.09 0.64 Ag -0.25 0.83 Ba 0.87 -0.32 Cd -0.39 0.76 Hg -0.29 0.74 Pb 0.54 0.69 Zn -0.35 0.76 Sr 0.78 -0.27 Br -0.74 0.45 I -0.66 0.46 <![CDATA[Na2O]]> 0.83 -0.11 <![CDATA[K2O]]> 0.92 -0.08 Eigenvalues 4.59 3.89 Variance contribution rate % 38.22 32.38 Cumulative variance contribution rate % 38.22 70.60

[0059] The results of the third factor analysis in Table 3 show that two factors were extracted, with a cumulative variance contribution rate of 70.60%. Among them, factor F1 is composed of positive load K2O-Na2O-Ba-Sr and negative load Br-I, with a variance contribution rate of 30.22%. The K2O-Na2O-Ba-Sr combination indicates areas with low-lying terrain and thick alluvial-diluvial deposits, while the Br-I combination is composed of elements from the leading edge halo, indicating ore-forming hydrothermal activity. Factor F2 is composed of Au-Ag-Hg-Cd-Pb-Zn, with a variance contribution rate of 32.38%, and is composed of ore-forming and near-ore-forming elements, indicating mineralization.

[0060] Comparative Example

[0061] Unlike the previous example, this example performed factor analysis without noise reduction. The results are shown in Table 4. A total of 6 factors were extracted, with a cumulative variance contribution rate of 84.12%. Factor F5 is composed of Au-Ag elemental combinations, indicating mineralization.

[0062] Table 4 shows the results of factor analysis without noise reduction.

[0063] element F1 F2 F3 F4 F5 F6 Au 0.25 0.04 -0.02 0.08 0.84 -0.04 Ag 0.19 0.22 0.30 0.20 0.81 -0.05 Hg 0.12 0.15 0.31 0.62 0.45 -0.13 Ba -0.52 -0.17 -0.73 -0.17 -0.11 -0.16 Pb -0.46 0.45 -0.22 0.26 0.45 -0.19 Sr -0.54 -0.16 -0.66 -0.08 -0.05 0.26 Cd 0.18 0.51 0.40 0.39 0.37 0.09 Zn 0.14 0.67 0.43 0.25 0.36 0.09 As 0.55 0.21 -0.05 0.66 -0.01 0.06 Bi 0.48 0.21 0.27 0.37 0.53 0.08 Co 0.72 0.42 0.29 0.29 0.26 0.10 Cr 0.73 0.27 0.42 0.32 0.23 0.15 Cu 0.58 0.34 0.38 0.33 0.46 0.11 Mn 0.42 0.73 0.12 0.17 0.08 -0.10 Mo 0.49 0.48 -0.16 0.38 0.30 -0.10 Ni 0.72 0.37 0.35 0.34 0.21 0.14 Sb 0.59 0.31 0.50 0.42 0.19 0.09 Se 0.13 0.15 0.13 0.81 0.25 0.09 Sn 0.22 0.46 0.60 0.46 0.11 -0.05 Ti 0.83 -0.14 0.08 0.07 0.15 0.01 V 0.83 0.25 0.24 0.23 0.22 0.15 W 0.58 0.25 0.39 0.34 0.41 0.09 Cl 0.29 0.06 0.52 0.14 0.29 0.30 Br 0.40 0.14 0.54 0.59 0.12 0.18 F 0.23 0.64 0.42 0.34 0.12 0.34 I 0.41 0.31 0.42 0.59 0.01 0.06 S 0.33 0.05 0.32 0.58 0.45 0.32 <![CDATA[Na2O]]> -0.81 0.00 -0.18 -0.28 -0.06 -0.20 <![CDATA[K2O]]> -0.80 0.07 -0.45 -0.03 -0.15 -0.21 <![CDATA[SiO2]]> -0.24 -0.71 -0.15 -0.11 -0.02 -0.58 <![CDATA[Al2O3]]> -0.16 0.89 0.05 0.02 0.06 0.05 <![CDATA[TFe2O3]]> 0.66 0.46 0.40 0.27 0.25 0.15 MgO 0.65 0.26 0.46 0.37 0.22 0.25 CaO 0.20 0.08 0.04 0.07 -0.10 0.87 Eigenvalues 8.81 4.97 4.69 4.59 3.62 1.92 Variance contribution rate % 25.90 14.63 13.81 13.49 10.64 5.66 Cumulative variance contribution rate % 25.90 40.53 54.33 67.83 78.46 84.12

[0064] Comparing the factor analysis results before and after noise reduction, we can find that: (1) Before noise reduction, a total of 6 factors were extracted, with a cumulative variance contribution rate of 84.12%. After noise reduction, a total of 2 factors were extracted, with a cumulative variance contribution rate of 70.60%. The factors extracted after noise reduction are more representative, that is, they can indicate the significant geological and geochemical processes in the study area, such as weathering and mineralization. (2) Before noise reduction, some potential elements related to mineralization were not included in the factors indicating mineralization due to noise interference. After noise reduction, the elements related to mineralization gradually appeared and constituted mineralization-related factors, which can indicate the most significant and important geological and geochemical processes in the study area. For example, in the Qujia concealed gold mine, it is mineralization and weathering (alluvial-diluvial deposits).

[0065] For any points not covered above, existing technologies shall apply.

[0066] Although specific embodiments of the present invention have been described in detail by way of examples, those skilled in the art should understand that the above examples are for illustrative purposes only and are not intended to limit the scope of the invention. Those skilled in the art can make various modifications or additions to the described specific embodiments or use similar methods to replace them, without departing from the direction of the invention or exceeding the scope defined by the appended claims. Those skilled in the art should understand that any modifications, equivalent substitutions, improvements, etc., made to the above embodiments based on the technical essence of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for noise reduction in geochemical composition data through factor analysis, characterized in that, Includes the following steps: Step 1: Collect or gather geochemical composition data related to the study area; Step 2: Perform a logarithmic ratio transformation on the geochemical composition data to open the "closing effect" of the geochemical composition data, and then perform a normal distribution test on the geochemical composition data after the logarithmic ratio transformation. The closure effect refers to the fact that the total amount of all geochemical components is equal to 1, and the geochemical components are mutually restrictive, showing negative or positive correlations. These correlations are spurious correlations and have no geological significance. Step 3: Apply statistical analysis to the geochemical composition data obtained in Step 1, calculate geochemical parameters, and select specific geochemical parameters to construct noise reduction standard 1; Noise reduction standard 1 is to remove elements that are below the threshold of the selected geochemical parameter; Step 4: Perform preliminary denoising on the geochemical composition data that conforms to a normal distribution based on the denoising standard 1 constructed in Step 3; perform the first factor analysis on the geochemical composition data after preliminary denoising; and construct denoising standard 2 based on the results of the first factor analysis. Noise reduction standard 2 is to remove factors composed of single elements, that is, to remove elements that are significantly affected by a certain geological and geochemical process. Step 5: Perform secondary denoising based on the denoising standard 2 constructed in Step 4. Perform a second factor analysis on the geochemical composition data after secondary denoising, and construct denoising standard 3 based on the results of the second factor analysis. The denoising standard 3 is to remove elements that have high loadings in two or more factors but whose loadings are all below the determined tuple combination threshold, and extract element combinations in the factors whose loadings are greater than the determined tuple combination threshold. That is, remove elements that are affected by multiple geological and geochemical processes and whose influence is comparable, and extract representative factors that are indicative of the geology, geochemistry and mineralization processes of the study area. Step 6: Based on the noise reduction standard 3 constructed in Step 5, perform noise reduction again, and conduct a third factor analysis on the geochemical composition data after noise reduction. Step 7: Check whether the result of the third factor analysis in step 6 meets noise reduction criteria 2 and 3. If it does, iterate steps 5 and 6 until the factor analysis results do not meet noise reduction criteria 2 and 3. Then the factor analysis result is the final result.

2. The method for denoising geochemical composition data factor analysis according to claim 1, characterized in that: In step 1, the geochemical composition data includes geochemical data of regional soil, rocks, stream sediments, geothermal data, and vegetation, as well as their component data. The component data includes soil fine particles and active phase extraction.

3. The method for denoising geochemical composition data factor analysis according to claim 1, characterized in that: In step 1, the consistency of sample types and data quality of the collected geochemical composition data should be checked. For the collected samples, it should be ensured that the sample types and representativeness are consistent. When analyzing and testing geochemical composition data, efforts should be made to ensure that the analysis is carried out in the same laboratory to avoid systematic errors caused by different laboratories.

4. The method for denoising geochemical composition data factor analysis according to claim 1, characterized in that: In step 3, the selected geochemical parameters are the enrichment coefficient and the coefficient of variation.

5. The method for denoising geochemical composition data factor analysis according to claim 4, characterized in that, Step 3, the specific steps for constructing noise reduction standard 1 include: (1) Calculate the coefficient of variation and enrichment coefficient of each element in the study area; (2) Based on the coefficient of variation and enrichment coefficient of each element in the study area, a noise reduction standard 1 is constructed, that is, elements below a certain coefficient of variation and / or enrichment coefficient threshold are removed.

Images