Method for detecting skarn type rich iron ore in deep buried area

Abstract

The invention relates to a deep burial area skarn type rich iron ore detection method, and belongs to the technical field of deep burial area rich iron ore detection, and the method comprises the steps: 1, carrying out the denoising and static correction of wide-area electromagnetic measurement point data of a control key zone, optimizing the data information, and carrying out the inversion modeling, and obtaining an underground electrical structure of which the depth is less than 3000m in an ore concentration area; 2, 2.5 D gravity and magnetic combined inversion is carried out; a gravity and magnetic theoretical curve is calculated through forward modeling by giving gravity and magnetic attribute parameters to an existing theoretical geologic model, then the gravity and magnetic theoretical curve is compared with an actually-measured abnormal curve, the model is continuously adjusted, the theoretical curve and the actually-measured curve are matched, and therefore the purpose of explaining gravity and magnetic anomalies is achieved. According to the method, abnormal values are removed through a manual intervention mode of denoising processing and static correction, 2.5 D gravity-magnetic combined inversion is combined, the detection precision of the skarn type rich iron ore in the deep buried area is greatly improved, and the form and the buried depth of the ore body are predicted more accurately by preposing the theoretical geologic model into the gravity-magnetic combined inversion.

Description

Technical Field

[0001] The invention belongs to the technical field of deep buried area rich iron ore detection, and in particular relates to a deep buried area skarn type rich iron ore detection method. Background Art

[0002] Magnetic exploration is to study geological structures and mineral resource distribution patterns by observing and analyzing magnetic anomalies caused by magnetic differences between rocks and ores. Magnetic anomalies are caused by magnetic differences between rocks and ores within a certain depth range underground and their distribution characteristics. Similar to gravity exploration, magnetic survey data of different scales can reflect geological information at different depths and divide geological structures and rock formations at different depths. Magnetic exploration is the most widely used, most efficient, most direct and effective basic geophysical exploration method and technology in iron ore exploration, and can be directly used for iron ore prospecting. Since the late 1980s, magnetic exploration has entered a new stage of high-precision magnetic exploration technology with the update of instruments and equipment. It has made important progress and remarkable achievements in application fields, theoretical methods and technologies, and geological prospecting, and has made outstanding contributions to magnet prospecting in my country.

[0003] There are two main difficulties in magnetic exploration in deep iron ore exploration: first, it is very difficult to quantify the depth of magnetic exploration, and it is difficult to judge the occurrence of hidden faults and the law of deep changes. Second, magnetic exploration is easily disturbed by the surface magnetic field, which increases the difficulty of qualitative interpretation of deep magnetic geological bodies. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a technical method for detecting skarn-type iron-rich ores in deep buried areas with higher accuracy.

[0005] The present invention is achieved through the following technical solutions: A method for detecting skarn-type iron-rich ore in a deep buried area, the specific steps of the method are as follows: The first step is to optimize the data information and perform inversion modeling on the wide-area electromagnetic measurement point data in the key control zone through denoising and static correction to obtain the underground electrical structure of the mining area shallower than 3000m; ① De-noising to improve the signal-to-noise ratio; remove outliers through human-computer interaction, and use a smoothing method to improve the data according to the change rules of adjacent data, specifically, weighting the data offset to obtain alternative data for interpolation; ② Static correction, used to eliminate static displacement caused by local conductive inhomogeneities or interference sources near the surface; polarization processing and vertical second-order derivatives are added to the conventional data collation and polarization process, and the polarization parameters are selected according to the local magnetic inclination and declination of the region; the vertical second-order derivative of the remaining abnormal polarization magnetic anomaly is taken, and the zero value line of the vertical second-order derivative is used to infer the boundary of the target magnetic body; the Rosenbach II formula is used in the vertical second-order derivative; The second step is 2.5D gravity-magnetic joint inversion. By assigning gravity and magnetic attribute parameters to the existing theoretical geological model, the theoretical gravity-magnetic curve is forward calculated, which is then compared with the measured anomaly curve. The model is continuously adjusted to fit the theoretical curve with the measured curve, thereby achieving the purpose of explaining the gravity and magnetic anomalies.

[0006] Furthermore, in the above step ①, when the data increase is significantly increased, it is considered that noise is added. Specifically, i+1 -a i >2(a i -a i-1 ) when it is considered that a i+1 is the starting point of the noise data, that is, the starting point of the outlier. The data at the beginning of this data are all outliers. When a i+1 -a i <2(a i -a i-1 ), the influence of noise is considered to end, that is, the abnormal value sequence ends, and the abnormal value of this segment is processed smoothly by weighted interpolation; i is the normal data number, j is the superposition number of the abnormal value start data, and k is the superposition number of the abnormal value end data; that is, the abnormal value sequence is: a i+j 、a i+j+1 、a i+j+2 、a i+j+3 、a i+j+4 ,......,a i+k ; The sequence before the outlier is: a0, a1, a2, a3, a4, a5, ..., a i ; The sequence after the outlier is: a i+k+1 、a i+k+2 、a i+k+3 、a i+k+4 ,......; Take k-j+2 data before and after the outlier as the weighted data basis, calculate the difference of each adjacent data, take the arithmetic mean of the corresponding values ​​of each sequence of normal values ​​before and after the outlier as the weight of the adjacent data difference for smoothing the outlier, and take a i+k+1 -a i The difference is taken as the total added value of the smoothed data, and the smoothed data of the middle k-j+2 outliers are determined by weight; Weight calculation for outliers: , j=0,1,2,......,J; J is the number of outliers; Weighted calculation of outliers: , P=1,2,3,......,J; J is the number of outliers; Outlier smoothing calculation: .

[0007] The beneficial effects of the present invention compared with the prior art are as follows: Due to the attenuation of geophysical signals by loose layers in deep coverage areas, the intensity of the data transmitted back is greatly reduced, resulting in low resolution. After data processing, the accuracy of information such as the morphology and burial depth of iron ore in deep buried areas is also greatly reduced. The present invention uses denoising and static correction to remove outliers and combines 2.5D gravity and magnetic joint inversion to greatly improve the detection accuracy of skarn-type rich iron ore in deep buried areas. By prepending the theoretical geological model to the gravity and magnetic joint inversion, the morphology and burial depth of the ore body are more accurately predicted.

[0008] In order to narrow the scope of the model and make the results closer to the actual situation, the gravity-magnetic joint profile inversion of the present invention uses the rock mass and stratum interface information inferred and interpreted by wide-area electromagnetic data, and combines and compares it with the prior drilling information, effectively improving the accuracy of the inversion results. The method described in the invention has been applied to deep prospecting in areas with deep iron ore burial, such as Laiwu, Jinling, Qihe-Yucheng, and has achieved good results. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 The actual resistivity map for some of the original data; Figure 2 is the resistivity map after replacing the outliers; Figure 3 This is a model diagram of the detection technology of skarn-type iron-rich ore in deep buried areas; Figure 4 This is the comprehensive inversion inference map of the gravity and magnetic profile of the Qihe-Yucheng ore cluster area; Figure 5 This is a comparison chart of wide-area electromagnetic sounding inversion inference and verification. DETAILED DESCRIPTION

[0010] The present invention firstly denoises and statically corrects the wide-area electromagnetic measurement point data of the key control zone, optimizes the data information and performs inversion modeling, obtains the underground electrical structure of the ore concentration area shallower than 3000m, and then performs 2.5D gravity and magnetic joint inversion, thereby improving the detection accuracy of skarn-type rich iron ore in deep buried areas, and more accurately predicts the shape and burial depth of the ore body. The technical solution of the present invention is further explained below by examples in combination with the accompanying drawings, but the protection scope of the present invention is not limited in any form by the examples.

[0011] Example 1 (1) The wide-area electromagnetic measurement point data in the key control zone is denoised and statically corrected to optimize the data information and perform inversion modeling to obtain the underground electrical structure of the mineral concentration area shallower than 3000m.

[0012] ① Denoising to improve the signal-to-noise ratio of the signal; this is done by removing outliers through human-computer interaction and using a weighted interpolation method based on the adjacent data to smooth the data according to the degree of data offset. When the data increase is significantly increased, it is considered that noise is added, specifically a i+1 -a i >2(a i -a i-1 ) when it is considered that a i+1 is the starting point of the noise data, that is, the starting point of the outlier. The data at the beginning of this data are all outliers. When a i+1 -a i <2(a i -a i-1 ), the noise influence is considered to end, that is, the outlier sequence ends, and the outlier is processed smoothly by weighted interpolation. i is the normal data number, j is the superposition number of the outlier start data, and k is the superposition number of the outlier end data. That is, the outlier sequence is: a i+j 、a i+j+1 、a i+j+2 、a i+j+3 、a i+j+4 ,......,a i+k ; The sequence before the outlier is: a0, a1, a2, a3, a4, a5, ..., a i ; The sequence after the outlier is: a i+k+1 、a i+k+2 、a i+k+3 、a i+k+4 ,......; Take k-j+2 data before and after the outlier as the weighted data basis, calculate the difference of each adjacent data, take the arithmetic mean of the corresponding values ​​of each sequence of normal values ​​before and after the outlier as the weight of the adjacent data difference for smoothing the outlier, and take a i+k+1 -a i The difference is taken as the total added value of the smoothed data, and the smoothed data of the middle k-j+2 outliers are determined by weight.

[0013] Weight calculation for outliers: , j=0,1,2,......,J; J is the number of outliers; Weighted calculation of outliers: , P=1,2,3,......,J; J is the number of outliers; Outlier smoothing calculation: .

[0014] For ease of understanding, take the case where there are 3 outliers, i.e. k-j+2=3, and use the following calculation formula: Weight calculation for outliers: ; ; ; .

[0015] Weighted calculation of outliers: ; ; .

[0016] Outlier smoothing calculation: ; ; ; ②Static correction: used to eliminate static displacement caused by local conductive inhomogeneities or interference sources near the surface.

[0017] a. First, identify whether the data contains static displacement, and make a judgment based on the geological structure and terrain undulation; b. Distinguish between static displacement and abnormal response, and do not process abnormal response; c. Correct the static displacement data, and add polarization processing and vertical second-order derivatives to the conventional data collation and polarization processing; first use polarization processing, and the polarization parameters select the local magnetic inclination and declination according to the area; then perform vertical second-order derivatives on the remaining abnormal polarization magnetic anomalies, and the zero value line of the vertical second-order derivative is used to infer the boundary of the target magnetic body, and the formula used is Rosenbach's second formula; after static correction processing, the data is significantly improved.

[0018] (2) 2.5D gravity-magnetic joint inversion: By assigning gravity-magnetic attribute parameters to the theoretical model, the gravity-magnetic theoretical curve is forward calculated, and then compared with the measured anomaly curve, the model is continuously adjusted to make the theoretical curve fit the measured curve. Figure 3 ), so as to achieve the purpose of explaining gravity and magnetic anomalies.

[0019] Example 2 The invention is applied in the exploration of rich iron ore.

[0020] (1) The wide-area electromagnetic data of the Qihe-Yucheng mining area were selected, and the wide-area electromagnetic measurement point data of the key control area were denoised and statically corrected to optimize the data information and perform inversion modeling to obtain the underground electrical structure of the mining area shallower than 3000m.

[0021] ① Denoising, improving the signal-to-noise ratio of the signal, eliminating outliers through human-computer interaction and smoothing the data using an interpolation method weighted by the degree of data offset. When the data increase is significantly increased, it is considered that noise is added, specifically a i+1 -a i >2(a i -a i-1 ) when it is considered that a i+1 is the starting point of the noise data, that is, the starting point of the outlier. The data at the beginning of this data are all outliers. When a i+1 -a i <2(a i -a i-1 ), the noise influence is considered to end, that is, the outlier sequence ends, and the outlier is processed smoothly by weighted interpolation. i is the normal data number, j is the superposition number of the outlier start data, and k is the superposition number of the outlier end data. That is, the outlier sequence is: a i+j 、a i+j+1 、a i+j+2 、a i+j+3 、a i+j+4 ,......,a i+k ; The sequence before the outlier is: a0, a1, a2, a3, a4, a5, ..., a i ; The sequence after the outlier is: a i+k+1 、a i+k+2 、a i+k+3 、a i+k+4 ,......; Take k-j+2 data before and after the outlier as the weighted data basis, calculate the difference of each adjacent data, take the arithmetic mean of the corresponding values ​​of each sequence of normal values ​​before and after the outlier as the weight of the adjacent data difference for smoothing the outlier, and take a i+k+1 -a i The difference is taken as the total added value of the smoothed data, and the smoothed data of the middle k-j+2 outliers are determined by weight. The following calculation formula is used: Weight calculation for outliers: , j=0,1,2,......,J; J is the number of outliers; Weighted calculation of outliers: , P=1,2,3,......,J; J is the number of outliers; Outlier smoothing calculation: .

[0022] The actual resistivity data are as follows: 10.43563, 10.14315, 9.85887, 9.5, 9.05295, 8.552612, 8.552612, 9.12, 8.81, 9.34, 10.28, 10.15, 11.36, 13.74, 16.42, 21.45, 25.8, 32.43, 35.4, 41.18, 40.5 2, 47.17, 33.46, 22.48288, 22.48288, 23.13117, 23.79815, 23.13117, 23.79815, 23.79815, 24.48437, 24.48437, 23.5, 23.79815, 21.85275, 20.85, 20.23, 20.06; the curve is shown in Figure 1 It can be found that the abnormal values ​​are: 16.42, 21.45, 25.8, 32.43, 35.4, 41.18, 40.52, 47.17, 33.46; after smoothing according to the above method, the data becomes: 14.13, 14.63, 15.58, 16.33, 17.25, 17.98, 18.61, 19.54, 22.14; after replacing the abnormal values, the curve is formed as follows Figure 2 .

[0023] ②Static correction: used to eliminate static displacement caused by local conductive inhomogeneities or interference sources near the surface.

[0024] a. First, identify whether the data contains static displacement, and make a judgment based on the geological structure and terrain undulation; b. Distinguish between static displacement and abnormal response, and do not process abnormal response; c. Correct the static displacement data. First, polarization processing is used. The polarization parameters are magnetic inclination 55.41° and magnetic declination -6.28°. Then the vertical second-order derivative of the remaining abnormal polarization magnetic anomaly is taken. The zero value line of the vertical second-order derivative is the boundary of the inferred target magnetic body. The formula used is Rosenbach's second formula. After static correction processing, the data is significantly improved.

[0025] (2) 2.5D gravity and magnetic joint inversion. By assigning gravity and magnetic attribute parameters to the theoretical model, the gravity and magnetic theoretical curve is forward calculated, and then compared with the measured anomaly curve, the model is continuously adjusted to make the theoretical curve fit the measured curve ( Figure 4), so as to achieve the purpose of explaining the gravity and magnetic anomalies. It is inferred that the Pandian magnetic anomaly is mainly caused by the superposition of the diorite body and the iron ore body anomaly, and the local gravity anomaly is mainly caused by the diorite body and the Cambrian Ordovician limestone. The shallow low-resistance electrical layer corresponds to the low-resistance Cenozoic and Carboniferous Permian, and the deep high-resistance area is the reflection of the Ordovician limestone and diorite body. There is a relatively obvious electrical boundary between high and low resistance. The same conclusion as the inferred result was obtained through drilling verification ( Figure 5 ).

Claims

1. A method for detecting skarn-type iron-rich ore in deep buried areas, characterized in that: The specific steps of the method are as follows: The first step is to optimize the data information and perform inversion modeling on the wide-area electromagnetic measurement point data in the key control zone through denoising and static correction to obtain the underground electrical structure of the mining area shallower than 3000m; ① De-noising to improve the signal-to-noise ratio; remove outliers through human-computer interaction, and use a smoothing method to improve the data according to the change rules of adjacent data, specifically, weighting the data offset to obtain alternative data for interpolation; ② Static correction, used to eliminate static displacement caused by local conductive inhomogeneities or interference sources near the surface; polarization processing and vertical second-order derivatives are added to the conventional data collation and polarization process, and the polarization parameters are selected according to the local magnetic inclination and declination of the region; the vertical second-order derivative of the remaining abnormal polarization magnetic anomaly is taken, and the zero value line of the vertical second-order derivative is used to infer the boundary of the target magnetic body; the Rosenbach II formula is used in the vertical second-order derivative; The second step is 2.5D gravity-magnetic joint inversion. By assigning gravity and magnetic attribute parameters to the existing theoretical geological model, the gravity-magnetic theoretical curve is forward calculated, and then compared with the measured anomaly curve. The model is continuously adjusted to make the theoretical curve fit the measured curve.

2. The method for detecting skarn-type iron-rich ore in deep buried areas according to claim 1, characterized in that: In the above step ①, outliers are removed in the denoising process. When the data increase is significantly increased, it is considered that noise is added. Specifically, i+1 -a i >2(a i -a i-1 ) when it is considered that a i+1 is the starting point of the noise data, that is, the starting point of the outlier. The data at the beginning of this data are all outliers. When a i+1 -a i <2(a i -a i-1 ), the influence of noise is considered to end, that is, the abnormal value sequence ends, and the abnormal value of this segment is processed smoothly by weighted interpolation; i is the normal data number, j is the superposition number of the abnormal value start data, and k is the superposition number of the abnormal value end data; that is, the abnormal value sequence is: a i+j 、a i+j+1 、a i+j+2 、a i+j+3 、a i+j+4 ,......,a i+k ; The sequence before the outlier is: a0, a1, a2, a3, a4, a5, ..., a i ; The sequence after the outlier is: a i+k+1 、a i+k+2 、a i+k+3 、a i+k+4 ,......; Take k-j+2 data before and after the outlier as the weighted data basis, calculate the difference of each adjacent data, take the arithmetic mean of the corresponding values ​​of each sequence of normal values ​​before and after the outlier as the weight of the adjacent data difference for smoothing the outlier, and take a i+k+1 -a i The difference is taken as the total added value of the smoothed data, and the smoothed data of the middle k-j+2 outliers are determined by weight; Weight calculation for outliers: , j=0,1,2,......,J; J is the number of outliers; Weighted calculation of outliers: , P=1,2,3,......,J; J is the number of outliers; Outlier smoothing calculation: .

Images