Comprehensive prospecting prediction method for pegmatite type lithium ore

Abstract

The invention discloses a comprehensive prospecting prediction method for pegmatite type lithium ore, and relates to the technical field of pegmatite type lithium ore detection. The method comprises the following steps: delineating a large-scale geochemical exploration working area according to small-scale geochemical exploration anomalies; sampling 1: 50,000 water-based sediments; testing the water-based sediment sample, processing data and mapping; interpreting and evaluating geochemical exploration anomaly; geological survey of the prospecting target area; acquiring and processing hyperspectral remote sensing data; rock ground collection and feature analysis; extracting spodumene pegmatite from a hyperspectral remote sensing satellite image; predicting spodumene pegmatite vein in a regional range; and verifying the prediction effect. According to the comprehensive prospecting prediction method for the pegmatite type lithium ore, prospecting prediction is carried out by integrating multiple methods such as geology, geochemical exploration, remote sensing and hyperspectrum, the prediction precision and exploration efficiency of the pegmatite type lithium ore can be remarkably improved, and the increasing lithium requirement is met.

Description

Technical Field

[0001] This invention relates to the field of pegmatite-type lithium deposit exploration technology, and in particular to a comprehensive prospecting and prediction method for pegmatite-type lithium deposits. Background Technology

[0002] Pegmatite-type lithium deposits are one of the world's most important sources of lithium resources, and they are widely used in electric vehicles, battery storage, and electronic devices. With the rapid growth in lithium demand, the search for and development of new lithium resources has become a focus of attention for many countries.

[0003] Traditional lithium exploration methods primarily rely on geological surveys, geophysical exploration, and geochemical exploration. However, these methods are often time-consuming, labor-intensive, and have certain limitations. For example, geological surveys require significant manpower and time, while geophysical and geochemical exploration methods are highly dependent on the environment and perform poorly in complex terrain conditions. Especially in harsh, high-altitude regions like the Qinghai-Tibet Plateau and the Himalayan metallogenic belt, where remoteness and extremely inconvenient transportation make traditional lithium exploration methods extremely slow and costly in terms of manpower and resources.

[0004] Therefore, there is an urgent need for a method that can be used to accurately and quickly identify pegmatite-type lithium deposits and make prospecting predictions in areas where it is difficult to reach by manpower. Summary of the Invention

[0005] The purpose of this invention is to provide a comprehensive prospecting and prediction method for pegmatite-type lithium deposits. By integrating multiple methods such as geology, geochemistry, remote sensing, and hyperspectral analysis for prospecting and prediction, this method can significantly improve the prediction accuracy and exploration efficiency of pegmatite-type lithium deposits, thereby meeting the growing demand for lithium.

[0006] The technical solution of this invention to solve the above-mentioned technical problems is: a comprehensive prospecting and prediction method for pegmatite-type lithium deposits, comprising the following steps: Step 1: Delineate the large-scale geochemical exploration work area based on small-scale geochemical anomalies; The scope of geochemical work at the standard map sheet of 1:50,000 was delineated based on the metallogenic belt and the 1:200,000 geochemical anomaly. Among them, the 1:200,000 scale geochemical survey is a geochemical survey that has been basically fully covered by the state. The lithium element anomaly at the 1:200,000 scale can provide a direct basis for selecting areas for carrying out larger scale 1:50,000 stream sediment surveys. Step 2: 1:50,000 scale sediment sampling of the river system; 1:50,000 stream sediment measurements were conducted to further delineate geochemical anomaly zones of different levels. The quality control methods and indicators for the 1:50,000 stream sediment measurements were carried out in accordance with the requirements of the China Geological Survey's "Geochemical Survey Specification (1:50000)" (DZ / T 0011-2015) and "Geological and Mineral Laboratory Testing Quality Management Specification" (DZ / T 0130-2006), as well as other quality technical standards, specifications, and systems. According to the "Geochemical Survey Specification" (1:50000) and the landscape geochemical conditions within the survey area, the survey area was divided into three main zones: an intensive sampling zone, a basic sampling density zone, and a non-sampling zone. The basic sampling density for stream sediment measurements was 4-8 points / km², with actual sampling based on the principle of effectively controlling the anomaly range and avoiding continuous blank areas in the survey area. The sampling techniques included: GPS verification, determination of sampling grain size, and determination of sampling points. Step 3: Testing of river sediment samples and data processing and mapping; The testing employed a comprehensive analytical approach combining multiple methods, including inductively coupled plasma mass spectrometry (ICP-MS), X-ray fluorescence spectrometry (XRF), foam adsorption-inductively coupled plasma mass spectrometry (ICP-MS), atomic fluorescence spectrometry (AFS), and emission spectrometry (ES). The tested elements included Au, Be, Co, Cr, Cu, Li, Mo, Nb, Ni, Sb, Sn, U, W, Zn, Pb, Ta, Rb, As, Hg, Ag, Al, Ba, Fe, Cd, and Bi, totaling 25 elements. Geochemical data processing and the compilation of a series of maps were carried out based on 25 elemental indicators analyzed and tested, and were compiled according to standard map sheets. A total of one set of geochemical maps was compiled, including original point data maps, geochemical maps, single-element anomaly maps, combined anomaly maps, comprehensive anomaly maps, mineral exploration prediction maps, and major comprehensive anomaly analysis maps. The coordinate system used for mapping is Gaussian 6-degree zoning with 15 zones, 2000 National Geodetic Coordinate System, and 1:50,000 standard map sheets. Mapping uses MAPGIS as the basic platform, data processing and encryption are implemented in GeoExpl using index weighting, and parameter statistics, histograms, etc. are completed using self-developed programs. Step 4: Geochemical anomaly interpretation and evaluation, which specifically includes the following: (i) Delineation of single-element anomalies; (ii) Delineation of combined anomalies; (iii) Delineation of prospective mineral exploration areas; (iv) Delineation of prospective mineral target areas. Step 5: Geological survey of the mineral exploration target area; Prioritize high-level prospecting target areas and conduct 1:10,000 specialized geological mapping to locate spodumene ore bodies. Deploy large-scale 1:2000 geological profiles and petrogeochemical profiles at the locations of discovered ore bodies. Step Six: Hyperspectral Remote Sensing Data Acquisition and Processing; Using satellites to identify lithium deposits in pegmatites; The main processing steps include: data preparation and import, radiometric calibration, atmospheric correction, bad band removal, and geometric correction; Step 7: Rock surface sampling and feature analysis; Ground spectral information of spodumene pegmatite and ore-free rock bodies in the visible to short-wave infrared range was collected using an ASD spectrometer. The collected results were analyzed and compared with the standard spectra of single minerals in commonly used spectral databases. The spectral characteristics of spodumene pegmatite and related ore-free rock bodies were analyzed to find differences in spectral characteristics, in order to prepare for the extraction of spodumene-bearing pegmatite. Step 8: Extraction of spodumene pegmatite from hyperspectral remote sensing satellite imagery; By analyzing the relevant rock surface spectra, the extraction process of spodumene pegmatite from hyperspectral remote sensing satellite images was determined. First, the range of pegmatite veins was determined using high spatial resolution remote sensing images. Then, the overall waveform characteristics of the hyperspectral remote sensing images were matched with the image spectra of known spodumene pegmatite locations through spectral angle matching. Finally, the extraction accuracy of spodumene pegmatite was further improved by calculating the slope of the image in the 2370nm-2410nm spectral range. Step Nine: Prediction of spodumene pegmatite veins within a specific area; Satellite remote sensing spectral characteristics of spodumene-bearing pegmatites are used as positive samples, while spectral characteristics of spodumene-free pegmatites and spodumene-free granites are used as negative samples. The possible locations of ore bodies are interpreted from satellite remote sensing images based on the spectral characteristics of known ore bodies. Step 10: Verify the prediction results; The location of the lithium-bearing pegmatite vein was predicted by combining steps five, six, seven, eight, and nine, and then verified by field surveys.

[0007] As a further improvement of the present invention, the single-element anomaly identification in step four is as follows: Based on the geological structure characteristics of the work area, the geological background of the map sheets is relatively stable. Therefore, a unified lower limit for anomaly delineation is adopted for each map sheet. Among them, three methods are used to calculate the lower limit for anomaly: First, the lower limit for anomaly is calculated directly from the original data (encrypted data) according to X±2S; Second, the lower limit for anomaly is calculated by performing a logarithmic transformation on the original data (encrypted data) and then calculating according to X±2S, and converting the result into a true value to obtain the lower limit for anomaly; Third, the data with a cumulative frequency of 90% of the original data (encrypted data) is directly used as the lower limit for anomaly. Six sets of abnormal lower limits were obtained using the above three methods. These six sets of abnormal lower limits were used to delineate anomalies. The average of these six sets of abnormal lower limits was then used to delineate anomalies, which yielded good delineation results. After rounding and adjustment, the abnormal lower limit value was finally determined. Based on the range of 1, 2, and 4 times the abnormal lower limit value, the three-level concentration zones of outer, middle, and inner anomalies were determined.

[0008] As a further improvement of the present invention, the specific steps of step four (ii) for identifying combined anomalies are as follows: Based on the single-element anomaly map, and taking into account factors such as cluster analysis, the metallogenic relationship of the elements, and the geological and metallogenic background of the area, the elements are divided into several groups. The main metallogenic elements are represented by surface colors in three zones: inner, middle, and outer. Other elements are represented by line colors, which only indicate the range of the outer anomaly zone.

[0009] As a further improvement to the present invention, step (iv) of delineating the mineral exploration target area in step four is as follows: Based on the indicators of prospecting target areas in the "Technical Requirements for Mineral Prospect Survey", the determination of the category of prospecting target areas takes into account factors such as the degree of mineralization favorability, engineering control, mineralization intensity, richness and intensity of comprehensive prospecting information (physical, chemical and remote sensing), sufficiency of prediction basis, and size of resource potential. Prospecting target areas are divided into three levels: A, B, and C. Through comprehensive analysis of the geological and geochemical characteristics of the work area, and based on mineralization regularity and prospecting indicators, three prospecting target areas are delineated.

[0010] As a further improvement of the present invention, in step eight, for the spectral matching of spodumene pegmatite, the spectral angle matching algorithm of ENVI software is selected to initially extract the range of spodumene pegmatite.

[0011] As a further improvement of the present invention, in step eight, for the calculation of the slope of the image spectral interval, the core algorithm for calculating the slope of the image spectral interval is obtained by Python programming: using polynomial fitting to calculate the spectral slope within a specified wavelength range, and saving the fitting result to the output image. The polynomial fitting part in the spectral slope() function is specifically implemented using the numpy.polyfit() function.

[0012] As a further improvement of the present invention, in step nine, the visual interpretation results, spectral matching, and slope calculation results of the pegmatite veins are superimposed to further improve the accuracy of the location of the spodumene pegmatite; the specific process is as follows: The classification raster file obtained after spectral matching and oblique intersection calculation is converted into a vector file. The intersection is then taken in ArcGIS software to obtain the spodumene pegmatite region extracted from the hyperspectral remote sensing data. The visual interpretation results of the pegmatite veins are then superimposed to identify the veins falling within the spodumene pegmatite region as spodumene pegmatite veins.

[0013] Beneficial effects Compared with existing technologies, the advantages of the comprehensive prospecting and prediction method for pegmatite-type lithium deposits of the present invention are as follows: This method integrates multiple approaches, including geological, geochemical, and hyperspectral remote sensing, for mineral exploration prediction. This organic combination of methods makes the predicted mineral types more targeted and the predicted area locations more accurate, thus significantly improving the prediction accuracy and exploration efficiency of pegmatite-type lithium deposits to meet the growing demand for lithium. In particular, the method of extracting hyperspectral remote sensing data allows for difference analysis between the extracted spodumene pegmatite spectral characteristics and lithium-free pegmatite spectral characteristics, enabling the determination of an effective prediction range. Meanwhile, compared with traditional lithium exploration methods, this method can first accurately and quickly identify pegmatite-type lithium deposits and make prospecting predictions in areas where it is difficult for humans to reach, thus saving a lot of manpower and material resources.

[0014] The invention will become clearer from the following description, taken in conjunction with the accompanying drawings, which are used to explain embodiments of the invention. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 GPS track map; Figure 2 A table outlining the delineation of prospective mineral exploration areas and target areas within the working area; Figure 3 Table of spectral absorption characteristic vectors for tourmaline granite; Figure 4 Spectral reflectance and outer envelope diagram of tourmaline granite; Figure 5 A diagram showing the convex hull ratio after removing the envelope of tourmaline granite. Figure 6 Table of spectral absorption characteristic vectors for spodumene pegmatite (weathered surface); Figure 7 The spectral reflectance curve of spodumene pegmatite (weathered surface); Figure 8 A diagram showing the convexity ratio of spodumene pegmatite (weathered surface); Figure 9 Table of spectral absorption characteristic vectors for spodumene pegmatite (fresh surface); Figure 10The spectral reflectance curve of spodumene pegmatite (fresh surface); Figure 11 A convexity hull ratio diagram for spodumene pegmatite (fresh face); Figure 12 Table of spectral absorption characteristic vectors for albite granite; Figure 13 This is a graph showing the spectral reflectance of albite granite. Figure 14 A diagram showing the convexity ratio of albite granite; Figure 15 Table of spectral absorption characteristic vectors for mineral-free granite pegmatite; Figure 16 This is a spectral reflectance curve of mineral-free granite pegmatite; Figure 17 A bulge ratio diagram for mineral-free granite pegmatite; Figure 18 The spectral differences of different rocks in the 2370nm-2410nm range; Figure 19 The spectral slope values ​​of different rocks in the 2370nm-2410nm range; Figure 20 Features of high-resolution remote sensing image of granite pegmatite (light-colored area); Figure 21 This is pegmatite vein 1 in the Shenglong target area (red band); Figure 22 This is Shenglong pegmatite vein 2 (red band); Figure 23 This is Shenglong pegmatite vein 3 (red band); Figure 24 This is Shenglong pegmatite vein 4 (red band); Figure 25 A comparison of the image spectrum of spodumene pegmatite with the ground spectrum; Figure 26 This is a depth map of the 2200nm absorption peak in a hyperspectral remote sensing image (green area). Figure 27 The image spectrum is plotted as slope * 10000 for the 2370-2410 nm range. Figure 28 This is a graph showing the results of spectral interval slope threshold segmentation. Figure 29 This is a partial result image of a spodumene pegmatite vein; Figure 30 This is a flowchart of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0018] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; of course, they can also refer to a mechanical connection or an electrical connection; furthermore, they can refer to a direct connection, an indirect connection through an intermediate medium, or a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0019] Embodiments of the present invention will now be described with reference to the accompanying drawings.

[0020] Example Specific embodiments of the present invention are as follows: Figure 30 As shown, a comprehensive prospecting and prediction method for pegmatite-type lithium deposits includes the following steps: Step 1: Delineate the large-scale geochemical exploration work area based on small-scale geochemical anomalies; In this step, technicians delineate the geochemical exploration area at a standard map scale of 1:50,000 based on the metallogenic belt and 1:200,000 geochemical anomalies. In this embodiment, the target area selected by the inventors is located in the Himalayan metallogenic belt, where several pegmatite-type lithium deposits have been discovered, including the Gabo, Kuju, and Qiongjiagang lithium deposits, showing favorable conditions for pegmatite-type lithium mineralization. The 1:200,000 scale geochemical exploration is a nationally completed, essentially comprehensive geochemical survey. The lithium anomalies at this scale provide direct evidence for selecting areas for larger-scale 1:50,000 stream sediment surveys. This step determines the scope of 1:50,000 stream sediment surveys conducted on the Gunda and Gyirong map sheets.

[0021] Step 2: 1:50,000 scale sediment sampling of the river system; In this step, 1:50,000 stream sediment measurements are conducted to further delineate geochemical anomaly zones of different levels. The quality control methods and control indicators for the 1:50,000 stream sediment measurements are carried out in accordance with the quality technical standards, specifications and systems of the China Geological Survey, such as the "Geochemical Survey Specification (1:50000)" (DZ / T 0011-2015) and the "Geological and Mineral Laboratory Testing Quality Management Specification" (DZ / T 0130-2006).

[0022] According to the "Geochemical Survey Specification" (1:50000) and considering the landscape geochemical conditions within the survey area, the survey area was divided into three main zones: an intensive sampling zone, a basic sampling density zone, and a non-sampling zone. The basic sampling density for stream sediment measurements was 4-8 points / km², with actual sampling based on the principle of effectively controlling anomalies and avoiding continuous blank areas. In this embodiment, the non-sampling zone mainly consisted of high-altitude perennial snowfields, steep cliffs, and areas inhabited by predatory wild animals such as brown bears and wolves. Except for the non-sampling zone and the intensive sampling zone, all other areas were considered basic sampling density zones, with a sampling density of not less than 4 points / km². Sampling techniques included: GPS verification, determination of sampling grain size, and determination of sampling points.

[0023] During GPS verification, the work area sampling uses planar coordinates (Cartesian coordinates). To reduce errors, parameter calibration must be performed on each GPS device. The specific steps for GPS verification are as follows: (a) Correcting GPS errors using known coordinate points: 1. Before fieldwork begins, use each GPS to measure the coordinates of the known triangulation points to obtain the coordinates XGPS and YGPS, and then calculate the difference between the two (△X = XGPS - X (known), △Y = YGPS - Y (known)).

[0024] 2. Calculate FALSE′E′ (east-west deviation) and FALSE′N′ (north-south deviation). East-west deviation = 500000 - △X, North-south deviation = 0 - △Y.

[0025] (ii) Determining the custom coordinate format (User UTM Grid); 1. Determine the longitude of the central meridian of the working area according to the projection zone number (select the 6° zone number).

[0026] 2. Other relevant parameters are set as follows: the central meridian longitude should be set to E870000, the projection scale parameter is 1, the east-west deviation is 500000, and the north-south deviation is 0, with the unit set to meters. Generally, these parameters should be kept at their default settings.

[0027] (III) Determining the projection parameters of the user-defined coordinate system; 1. The GPS custom coordinate system (User) projection parameter setting interface provides five variables (△X, △Y, △Z, △A, △F) that need to be set. However, in actual work, the latter two parameters (△A, △F) are fixed parameters for a certain coordinate system (National 2000 coordinate system △A=0, △F=0.0000005) and do not need to be changed. The parameters that need to be calculated by yourself are mainly the first three (△X, △Y, △Z), which are generally three parameters.

[0028] 2. Only GPS devices that have been calibrated to meet the required error level can be used in fieldwork.

[0029] The sampling granularity is determined as follows: The sampling medium is mainly composed of sandy components, avoiding the collection of silt and organic matter. Based on the characteristics of the arid and semi-arid alpine mountainous natural landscape of the work area, and the results of previous stream sediment measurement work in the surrounding Shannan and Shigatse regions (the "Shannan Region 1:50,000 Stream Sediment Measurement Sampling" project and the "Gangdise Xiongba Region-Eastern Tibet Geyingduo Region 1:50,000 Stream Sediment Sampling" project), the sampling particle size is set at -10 to +60 mesh.

[0030] Furthermore, in order to improve the representativeness of the samples, samples should be collected at 3-5 locations within a 50-meter range above and below the designed sampling points (non-active flow line locations), and combined into one sample. The sample weight should be at least 300g after sieving.

[0031] The specific steps for determining the sampling points are as follows: (i) Use a 1:50,000 topographic map as the working map. Use GPS in conjunction with the topographic map to locate sampling points. Navigate to the vicinity of the design point using GPS, and select and confirm the point location based on the topography. Sampling can only proceed after the point location is confirmed to be correct. Record and save the coordinate information of each actual sampling point on the GPS, and record the flight track for future reference. Figure 1 As shown.

[0032] (ii) The location error or offset of the field sampling point shall not exceed 50m (not more than 1.0mm on the map). Each sampling point shall be marked with red paint, and the marking content shall include the large grid number + small grid number + serial number.

[0033] (iii) If, during the sampling process, it is found that qualified samples cannot be collected or sampling is not suitable at the designed sampling points, the sampling points shall be moved as appropriate, and the actual sampling location shall be marked on the topographic map and the reason for the relocation shall be noted in the record table. For example, after on-site verification, it was found that the sampling points JLQ284A1 and JLQ305B1 in the Jilong area were located above a cliff and there was no road access to the points. Therefore, the points were moved within the scope required by the specifications.

[0034] (iv) All sampling sites and markings shall be photographed for future reference. The photos and GPS records shall be entered into the database for unified management after the team returns each day. The photos shall include a distant view of the sampling site and 2-4 photos of the collected samples.

[0035] (v) If the wrong water system point, the point location is out of tolerance, or the sample is unqualified is found during the sampling process, resampling will be arranged. For example, the points JLQ392D1 and JLQ392B1 in the Jilong area were resampled because the wrong water system was sampled, and the points GD063A1, GD038D1, and GD038A1 in the Gunda area were resampled because the sample quantity was insufficient for screening.

[0036] (vi) After each day's fieldwork is completed, the GPS information is promptly imported into the sampling information database and directly projected onto the calibrated topographic map to compile a map of the actual sampled materials.

[0037] Step 3: Testing of river sediment samples and data processing and mapping; In this step, the testing employed a comprehensive analytical approach combining multiple methods, including inductively coupled plasma mass spectrometry (ICP-MS), X-ray fluorescence spectrometry (XRF), foam adsorption-inductively coupled plasma mass spectrometry (ICP-MS), atomic fluorescence spectrometry (AFS), and emission spectrometry (ES). The analyzed elements were Au, Be, Co, Cr, Cu, Li, Mo, Nb, Ni, Sb, Sn, U, W, Zn, Pb, Ta, Rb, As, Hg, Ag, Al, Ba, Fe, Cd, and Bi, totaling 25 elements. The Li element content was determined using inductively coupled plasma mass spectrometry (ICP-MS). The testing procedure was as follows: the sample was decomposed with nitric acid, perchloric acid, and hydrofluoric acid, and the hydrofluoric acid was removed. After dissolving in aqua regia, the sample was transferred to a colorimetric tube, brought to volume, and shaken well. A portion of the solution was then diluted with dilute nitric acid and analyzed using a plasma mass spectrometer.

[0038] Geochemical data processing and the compilation of a series of maps were conducted based on 25 elemental indices analyzed and tested, and were compiled according to standard map sheets. A total of one set of geochemical maps was created—including original point data maps, geochemical maps, single-element anomaly maps, combined anomaly maps, comprehensive anomaly maps, mineral exploration prediction maps, and major comprehensive anomaly analysis maps. Map coordinate system: The Gyirong and Gunda map sheets adopt the Gaussian 6-degree zoning system (15 zones), the 2000 National Geodetic Coordinate System, and a 1:50,000 standard map sheet.

[0039] Mapping is based on MAPGIS, data processing and encryption are implemented using exponential weighting in GeoExpl, and parameter statistics, histograms, etc. are completed using self-developed programs.

[0040] Step 4: Geochemical Anomaly Interpretation and Evaluation. This step specifically includes the following: (a) Single-element anomaly delineation; Based on the geological structural characteristics of the work area, the geological background of the map sheets is relatively stable. Therefore, a unified lower limit for anomaly delineation is used for each map sheet. Three methods were employed to calculate the lower limit: First, the lower limit is calculated directly from the original data (encrypted data) using X±2S; second, the lower limit is calculated from the original data (encrypted data) after logarithmic transformation using X±2S, and the result is converted into true values ​​to obtain the lower limit; third, the lower limit is directly used from the data with a cumulative frequency of 90% of the original data (encrypted data). Since the area of ​​each 1:50,000 standard map sheet is less than 500 km², and the geological background is relatively uniform, filtering methods are generally not used for anomaly delineation, as filtering has a strong effect on reducing the intensity of anomalies.

[0041] Six sets of lower limits for anomalies were obtained using the three methods described above. These six sets of lower limits were then used for anomaly delineation, and the results showed that all delineated anomalies had defects. Either the lower limit was too low, causing anomalies to merge into a single, unseparable area, or the lower limit was too high, making it impossible to delineate any anomalies. A single lower limit was insufficient for anomaly delineation. Therefore, the technicians attempted to average these six sets of lower limits and then use the average value for anomaly delineation. This yielded better delineation results. After rounding and adjustments, the final lower limit value was determined, and three concentration zones (outer, middle, and inner) were established based on 1, 2, and 4 times the lower limit value.

[0042] (ii) Delineation of combined anomalies; Based on the single-element anomaly map, and considering cluster analysis, elemental relationships, and the geological and metallogenic background of the area, the elements are divided into six groups: Ag-Cd-Zn; Nb-Ba-Al2O3; Au-As-Hg; Li-Be-Rb-Sn-Ta-Sb; Cu-TFe2O3-Co-Cr-Mo-Ni; and Pb-Bi-UW. The main ore-forming elements are represented by surface colors in three zones: inner, middle, and outer. Other elements are represented by line colors, indicating only the outer zone of the anomaly.

[0043] Gyirong Region: Based on the single-element anomaly map, and considering cluster analysis, elemental metallogenic relationships, and the region's geological and metallogenic background, the elements are divided into five groups: Li-Nb-Rb-Ta-Sn-Al2O3; Cu-TFe2O3-Co-Cr-Ni-Ba; Au-Ag-Cd-Zn-Mo-Hg; Pb-W-Bi-Be-U; As-Sb. The main metallogenic elements are represented by surface colors in three zones: inner, middle, and outer. Other elements are represented by line colors, indicating only the outer anomaly zone.

[0044] (iii) Delineate prospective mineral exploration areas; 1. Principles for delineating prospective mineral exploration areas: Based on the relative relationship between the distribution of geochemical anomalies and strata, magma, structure, and mineral resources, and considering the differences in their metallogenic conditions and prospecting background, prospective areas are classified into three levels: I, II, and III. Class I prospective mineral areas: These areas possess superior mineralization geological conditions, with mineral deposits, mineral occurrences, and mineralized points already discovered. The mineralization elements exhibit high anomalies, with obvious anomaly concentration centers, high correlation between anomalies, and clear surface prospecting indicators. They possess excellent mineralization conditions and prospecting prospects, with the potential for further prospecting and the possibility of expanding the scale of discovered mineral deposits and finding new ore bodies or deposits.

[0045] Level II prospective mineral areas: These areas possess favorable mineralization conditions. Mineral occurrences or mineralized points have been discovered within the area, but the mineralizing elements exhibit high anomalies, with obvious anomaly concentration centers and high overlap of anomaly elements. They possess relatively good mineralization geological conditions, and further work may lead to the discovery of mineralization clues.

[0046] Level III prospective mineral exploration area: It has certain mineralization geological conditions, but no mineralization clues have been found in the area. Geochemical anomalies are present, but the anomaly values ​​of mineralization elements are low and the anomaly concentration center is not obvious. Compared with the regional mineralization conditions, it has certain mineral exploration potential.

[0047] Naming principle: Geographical name + main metallogenic period + mineral type + prospective mineral exploration area.

[0048] 2. Delineation of prospective mineral exploration areas: (1) The degree of favorability of the geological background to mineralization: Compared with the ore-controlling strata, rock strata (body) and ore-controlling structures in the region, its development degree and characteristics in this area have certain mineralization geological conditions.

[0049] (2) Geochemical background and anomalous features: geochemical background field characteristics, geochemical anomalous scale, intensity, zonation, and combination of anomalous elements. The geochemical background and anomalous features show that the region has a high degree of enrichment of ore-forming elements and may be enriched into minerals.

[0050] (3) Mineralization clues have been discovered in the area. Compared with the mineral deposits (points) of the same type discovered in the region, the mineral type, degree of mineralization, mineralization type and genetic type are comparable.

[0051] Based on the above principles, the Gyirong area is divided into the Malebiluo-Labila Class III prospective mineral exploration area, and the Gunda area is divided into the Shenglong Class I prospective mineral exploration area and the Quedangxi Class III prospective mineral exploration area.

[0052] (iv) Delineating mineral exploration target areas; Based on the indicators for prospecting target areas in the "Technical Requirements for Mineral Prospect Survey," the determination of the target area category comprehensively considers factors such as the degree of mineralization favorability, engineering control status, mineralization intensity, the richness and intensity of comprehensive geophysical, chemical, and remote sensing prospecting information, the sufficiency of predictive basis, and the size of resource potential. Therefore, prospecting target areas are divided into three levels: A, B, and C. Category A: Excellent mineralization conditions, good mineralization, and geological, geophysical, chemical, and remote sensing data all indicate a favorable area for mineralization. Sufficient data support is available, the prospecting prospects are good, and there is potential to find medium-sized or larger deposits with significant resource potential.

[0053] Category B: Areas with favorable mineralization conditions and good mineralization indications, supported by geological, geophysical, chemical, and remote sensing data. There is sufficient data to support these findings, indicating potential for mineral exploration and resource availability, with the possibility of discovering medium to small-sized deposits.

[0054] Category C: Possesses certain mineralization geological conditions, with geochemical anomalies or mineralization indications, but lacks sufficient data and its resource potential is unclear.

[0055] Naming principles for mineral exploration target areas: geographical name + mineral type + target area (level).

[0056] Based on a comprehensive analysis of the geological and geochemical characteristics of the work area, and according to metallogenic regularities and prospecting indicators, three prospecting target areas were delineated in the Gyirong and Gunda areas. (See details...) Figure 2 .

[0057] Step 5: Geological survey of the Shenglong beryllium-lithium polymetallic prospecting target area; In this step, high-level prospecting target areas are prioritized for 1:10,000 scale geological mapping to locate spodumene orebodies. Large-scale 1:2000 geological and petrogeochemical profiles are then deployed at the locations of discovered orebodies. Conventional mineral geological surveys cannot quickly and efficiently determine the outcrops of adjacent ore-bearing pegmatite veins. Based on the surface outcrops discovered through previous geochemical and geological methods, hyperspectral remote sensing is used to obtain the spectral characteristics of known orebodies, effectively determining the distribution of spodumene-bearing pegmatites within the region.

[0058] Step Six: Hyperspectral Remote Sensing Data Acquisition and Processing; In this step, the ZY-1 02D satellite was used to identify lithium deposits in pegmatites using hyperspectral imaging. ZY-1 02D is China's first operational civilian hyperspectral satellite, equipped with both a hyperspectral camera and a multispectral camera. This experiment primarily used imaging data from the hyperspectral camera. The hyperspectral data covers the 0.40-2.50 μm spectral range, containing 166 spectral bands, with a spatial resolution of 30 meters and a swath width of 60 kilometers, providing precise identification capabilities for surface materials. The ZY-1 02D ground processing system produces L1-level product data using standard procedures. The main file for the AHSI payload L1A-level product is in GeoTiff format. The product data package includes GeoTiff data files, XML description files, RPC parameter files, browsing map files, coverage vector files, observation geometry files, and calibration coefficient files.

[0059] The main processing steps include: data preparation and import, radiometric calibration, atmospheric correction, bad band removal, and geometric correction; Step 7: Rock surface sampling and feature analysis; In this step, an ASD spectrometer is used to collect visible to short-wave infrared ground-based spectral information of spodumene pegmatite and ore-free rock masses. The collected results are analyzed and compared with standard single-mineral spectra in commonly used spectral databases to analyze the spectral characteristics of spodumene pegmatite and related ore-free rock masses, identify differences in spectral features, and prepare for the extraction of spodumene-bearing pegmatite. (I) The ASD spectrometer, specifically the FieldSpec series spectrometer designed and manufactured by ASD Corporation in the United States, is relatively widely used in remote sensing applications in China. Its applications have expanded to include precision agriculture, forestry, marine and inland waters, snow and ice, environmental pollution monitoring, meteorology, geology and mineral resources, ground calibration, and education, among other fields. The specific operational procedure for acquiring ground spectra of rocks and minerals using an ASD spectrometer is as follows: 1. The RS3 software configuration is as follows: (1) Warm up the instrument for 1-30 minutes. During the first hour of operation, click the DC icon or repeat the optimization every 5-10 minutes.

[0060] (2) Connect the computer cable after preheating.

[0061] (3) Start the computer and enter the operating system.

[0062] (4) Double-click the HH icon to start the ASD operating program. Start High Contrast RS³, suitable for use in the field.

[0063] (5) Open the “Spectrum Save” option (Alt S or Control / Spectrum Save).

[0064] (6) Enter the following information: Path name - Enter the folder where you want to store the data. A good approach is to name the folder after the current date, such as C:\160907; Base name - The base name of the data file. File extensions are automatically ordered numerically. For example, if the first file is named C:\160907\SPECTRUM.000, the next file stored will be C:\160907\SPECTRUM.001; Starting spectrum Num (the starting spectrum file number) - Up to 1000 spectrum data files can be stored under the same file name; The number of files to save is the number of files to save in a single observation. It is generally set to 15, but no less than 10 (to ensure that at least one spectral curve is less affected by direct sunlight, white hats, and waves). Interval Between Saves is the time interval between each save, which is the sampling interval when saving the spectral curve. It is generally set to 1 second (one wave cycle). Comment prompts - Add necessary prompts.

[0065] (7) After filling in all the information for the stored files, click OK to close this window.

[0066] (8) Next, use Alt+C / C++ or Control+Adjust Configuration to enter the following information: 2. The operation of RS3 software is as follows: (1) Point the spectrometer with or without a lens at the white board, and make sure the white board is optimally illuminated and only the white board is in the field of view of the lens.

[0067] (2) Click the OPT icon to optimize the spectrometer's integration time. Note that re-optimizing every 15-20 minutes or when lighting and environmental conditions (such as cloud cover, humidity changes, sun movement, etc.) change is beneficial for data quality. In fact, it is recommended to re-optimize before each time you change the target to collect spectral data.

[0068] (3) Place the spectrometer above the object to be measured (Note: the orientation of the spectrometer should be the same as when collecting the reference spectrum from the whiteboard). At this time, the interface will display the relative reflectance spectrum.

[0069] (4) Press the space bar to save the current spectral curve. You will hear a prompt sound after pressing the space bar.

[0070] 3. The field measurement workflow is as follows: (1) In a dark room where the lighting conditions can be controlled, first turn on the power of the spectrometer, and then turn on the power of the computer.

[0071] (2) Create a folder to record spectral data.

[0072] (3) Fix the bare fiber optic probe connected to the spectrometer host onto the tripod.

[0073] (4) Start the RS3 software.

[0074] (5) Preheat the instrument for 15 to 30 minutes.

[0075] (6) Align the probe vertically with the standard whiteboard for optimization. First, click the OPT icon. When the curve is stable and the rawDN value of the maximum height is about half of the interface height, click the WR icon. When the obtained spectral reflectance curve is relatively stable and is basically a straight line with a reflectance value of 1.0, the optimization is complete. Otherwise, perform the optimization again. Or when the red saturation appears on the left, perform the optimization again until the optimization is successful.

[0076] (7) Click ALT+S and select the path and file name where you want to store the data.

[0077] (8) The probe is vertically aligned with the center of the object being measured, 0.10m from the sample surface.

[0078] (9) During the measurement, after the spectrum stabilizes, press the space bar on the computer to start collecting the reflectance spectrum information of the ground objects. After the collection is completed, move the probe to the next target for measurement.

[0079] (10) After the measurement is completed, first turn off the computer, then turn off the power of the spectrometer, and perform necessary cleaning and tidying.

[0080] 4. ViewSpecPro performs the following post-processing of data: (1) Double-click the icon to open the software, click File to open the file. The default path is the ViewSpecPro folder. To ensure that the open and save paths are the folder where the data is stored, you need to modify the open path.

[0081] (2) Click Setup, modify the Input Directory path, and at the same time change the Output Directory path to the folder where the spectral data is stored by default.

[0082] (3) Open a file, click View-Graph Data, and a window will appear. The most important items in the window menu are Format and Export. Click Format to directly convert the spectrum between DN, reflection (No Derivatite, 1st Derivatite, 2nd Derivatite), and Transmittance. The most frequently used options are reflection (No Derivatite, 1st Derivatite). Click Export to output in image or text format. To output an image, first select the format to save, usually JPG, then click FileBrowse to select the storage path, and click Export. To output a text file, select Text-Browse, select the storage path, and click Export. You can choose to output in list or table format.

[0083] (4) ASCII code output: Select the data to be exported, click Process ASCII Export in the main menu, select the data format (DN, Reflectance, Radiance / Irradiance) in the dialog box, and click OK. The file can be saved as a .txt file.

[0084] (5) Clicking on View header info in the main menu will display some specific information about this data, such as the instrument serial number, the time of data acquisition, the integration time at that time, and the versions of the operating software and data analysis software, etc.

[0085] (6) Clicking "Process" in the main menu allows for data statistics, transformations, and other analyses. The most commonly used tool is "Statistics," which can analyze up to 14 data points simultaneously, calculating the mean, average, and standard deviation. The generated files will be named with different file extensions. For example, the extension for the mean will be .mn.

[0086] (II) The spectral characteristics of the rock surface are analyzed as follows: Based on the geological and metallogenic background conditions of the study area, four types of rock spectra were mainly collected, such as... Figure 3 , 6As shown in Figures 9, 12, and 15: spodumene pegmatite, mineral-free granite pegmatite, tourmaline granite, and albite granite. The average spectrum of each rock type was used as the ground standard spectrum for this type of rock. The spectral absorption characteristic parameters were calculated using the IDL DISPEC spectral analysis plugin developed by the University of Twente, including five types of indicators: absorption location, absorption depth, absorption width, absorption area, and absorption object type. Finally, the spectral reflectance, outer envelope diagram, and convex hull ratio diagram after envelope removal were output. Comparison with relevant characteristic parameters provides a more intuitive display of spectral characteristics, serving as the basic data for spectral characteristic analysis. Below are explanations of relevant parameters and terminology: Absorption position (AP): The wavelength at which the reflectance is lowest in a spectral absorption valley; Absorption depth (AD): The distance from the point of lowest reflectivity to the normalized envelope within a certain absorption range of a certain wavelength band. Absorption width (AW): The spectral bandwidth at half the maximum absorption depth; Absorption area (AA): The area enclosed by the spectral absorption curve and the normalized envelope; Spectral absorption symmetry (AS): The common logarithm of the ratio of the area to the area to the left of the region with the vertical line of the absorption position as the boundary.

[0087] Envelope: The convex hull curve of each spectral curve.

[0088] Envelope removal (also known as continuum removal): This method compares the reflectance value of the spectral curve at the corresponding wavelength with the reflectance value of the envelope to obtain a new spectral curve. After envelope removal, the relative values ​​at "peak" points are all 1, and the relative values ​​at non-"peak" points are all less than 1, forming several absorption valleys. By removing the envelope from the spectral curve, spectral features are highlighted, thereby enhancing the absorption and reflection characteristics of the spectral curve. This method normalizes the reflectance to 0-1.0 and normalizes the absorption characteristics of the spectrum to a consistent spectral background, which is beneficial for comparison and classification with other spectral curves, such as... Figure 4 , 5 As shown in 7, 8, 10, 11, 13, 14, 16, and 17.

[0089] according to Figure 4 , 5The spectral characteristics of rocks shown in Figures 7, 8, 10, 11, 13, 14, 16, and 17 reveal that the four rock types exhibit significant similarities in their waveforms within the 400-2500 nm range, aside from differences in overall reflectance values. All four rock types show absorption peaks at 1400 nm and 1900 nm, as well as a 2200 nm absorption peak formed by AL-OH minerals, primarily white mica and clay minerals, with only minor differences. The main differences ultimately lie between 2370 nm and 2410 nm, primarily manifested in significant variations in the positions of absorption peaks within this range. Furthermore, there are marked differences in various spectral absorption parameters. However, compared to absorption position, absorption depth, absorption width, absorption area, and absorption target, direct waveform comparison reveals that the difference in spectral slope is even greater within this range. To mitigate the impact of resolution differences between hyperspectral imagery and ground-based spectral measurements, the spectral slope parameter was added. Figure 18 As shown in the figure, the slope of different lithologies in the 2370nm-2410nm range was calculated in Python.

[0090] Because the reflectance values ​​of hyperspectral remote sensing images are multiplied by 10000 after atmospheric correction, to ensure consistency with the units of spectral slope in subsequent images, the slope of the ground spectral interval is multiplied by 10000. From the spectral interval slopes, it can be observed that the interval slope * 10000 for spodumene-bearing pegmatites is negative, less than -1.5, while the interval slope * 10000 for granite pegmatites and tourmaline granites (without spodumene) is positive. The boundary of albite granite is straight, with a spectral interval slope * 10000 of -0.78. Figure 19 As shown.

[0091] Step 8: Extraction of spodumene pegmatite from hyperspectral remote sensing satellite imagery; Based on relevant rock surface spectral analysis, a process for extracting spodumene pegmatites from hyperspectral remote sensing satellite imagery was determined. First, the extent of the pegmatite veins was determined using high spatial resolution remote sensing imagery. Then, spectral angle matching was used to match the overall waveform characteristics of the hyperspectral remote sensing imagery with the spectra of images showing known spodumene pegmatite locations. Finally, the extraction accuracy of spodumene pegmatites was further improved by calculating the slope of the imagery in the 2370nm-2410nm spectral range. (I) The high spatial resolution remote sensing interpretation of the granite pegmatite dikes is as follows: In high-altitude areas of Tibet, where vegetation cover is low and bedrock exposure is relatively good, high spatial resolution remote sensing imagery provides a high degree of interpretation for granite pegmatite veins. Granite pegmatite veins, rich in light-colored minerals such as quartz and feldspar (high SiO2 content), have high reflectivity in the visible-near-infrared band, typically appearing as a light yellowish-brown to white hue in remote sensing images. Compared to the surrounding rock, their color is significantly lighter, making them easy to identify. Morphologically, they often occur as veins, lenses, or irregular masses, distributed along fault zones or contact zones of rock masses. For example: Figure 20 The northeast-trending granite pegmatites, with a light yellow hue and banded or wavy patterns, are distributed at the boundary between the calcareous slate on the northwest side and the glacial sediments on the southeast side. Due to their distinct tonal and morphological characteristics in high-spatial-resolution remote sensing imagery, visual interpretation of the granite pegmatite area can be effectively utilized. Figure 21-24 The results of a partial interpretation of the Jilong map.

[0092] (II) The spectral matching method for spodumene pegmatite is as follows: Because the pixel size of hyperspectral remote sensing images differs significantly in scale from the spot size of instruments used in ground-based spectral measurements, and because the spectra of remote sensing images are more complex mixed spectra, ground-measured rock spectra were not directly used as samples. Instead, known field coordinates of spodumene pegmatites were projected onto the hyperspectral remote sensing image using ENVI software as endmembers for spectral matching, preparing for the initial extraction of the spodumene pegmatite extent. Figure 25 It can be seen that extracting spectral matching endmembers from the image spectrum maintains consistency with the ground-measured spectrum in terms of spectral absorption peak characteristics, thus ensuring the consistency of waveforms in spectral matching calculations.

[0093] The Spectral Angle Mapper (SAM) algorithm in ENVI software was used to initially extract the extent of spodumene pegmatites. SAM is a remote sensing object classification algorithm based on the geometric similarity of spectral curves. Its core principle is to measure the similarity by calculating the angle θ between two vectors in n-dimensional space; the smaller the angle, the higher the matching degree. SAM is insensitive to changes in brightness because it only focuses on the shape of the spectral curve (vector direction) and is not affected by changes in overall brightness caused by light intensity or terrain shadows. Rock and mineral identification, as a relatively strict category in SAM, requires a small spectral angle value: <0.1 radians.

[0094] Since the spectral image lacks prominent features at the 2200nm absorption peak position, Python programming was used to calculate the depth of the 2200nm absorption peak, further assisting in verifying the results of the spectral angle extraction. Figure 26 As shown.

[0095] (III) The slope of the image spectral interval is calculated as follows: like Figure 27 , 28 As shown, ground-based spectra revealed a significant difference in the spectral slope of spodumene pegmatites within the 2370-2410 nm range compared to other non-mineralized pegmatites or surrounding rocks, a characteristic also observed in the image spectra of spodumene pegmatites. Based on the spectral slope within the characteristic range of spodumene pegmatite image spectra, a range was determined using threshold segmentation to obtain the image range meeting the slope condition: the slope between the average slope of lithium deposits in the image spectra of spodumene pegmatites and the slope of the ground spectrum of the fresh surface of the spodumene pegmatites.

[0096] Below is the core algorithm for calculating the spectral slope of an image using Python programming: It utilizes polynomial fitting to calculate the spectral slope within a specified wavelength range and saves the fitting result to the output image. The polynomial fitting part of the `spectral slope()` function is specifically implemented using the `numpy.polyfit()` function. The algorithm steps are as follows: 1. For each pixel in the input image, extract spectral data within a specified wavelength range.

[0097] 2. Use the numpy.polyfit() function to perform polynomial fitting on the spectral data. The fitting order is specified by the user (default is order 1, order 2 is optional).

[0098] 3. Save the fitted coefficients to the output image. For a first-order fit, the output includes the slope and intercept; for a second-order fit, it also includes the quadratic coefficients.

[0099] The core of the algorithm is: Python CopyCode c=numpy.polyfit(Sw,Ss,fitorder) This line of code is the core of the algorithm, where Sw is the wavelength array, Ss is the corresponding spectral reflectance or radiance array, and fitorder is the fitting order. The numpy.polyfit() function returns the coefficient array c of the fitted polynomial.

[0100] Step Nine: Prediction of spodumene pegmatite veins within a specific area; Satellite remote sensing spectral characteristics of spodumene-bearing pegmatites are used as positive samples, while spectral characteristics of spodumene-free pegmatites and granites are used as negative samples. The possible locations of ore bodies are interpreted from satellite remote sensing imagery using the spectral characteristics of known ore bodies. Furthermore, by using the spectral characteristics of spodumene-free pegmatites and granites as negative samples, pegmatites with low spectral correlation within the delineated ore body area are eliminated, thus enabling rapid and efficient delineation and prediction of ore bodies.

[0101] The accuracy of spodumene pegmatite location is further improved by overlaying visual interpretation results, spectral matching, and slope calculation results. The specific process is as follows: the classification raster files obtained after spectral matching and oblique intersection calculations are converted into vector files; the intersection is then performed in ArcGIS software to obtain the spodumene pegmatite region extracted from hyperspectral remote sensing data; the visual interpretation results of the pegmatite veins are then overlaid, and the veins falling within the spodumene pegmatite region are identified as spodumene pegmatite veins. Figure 29 As shown.

[0102] Step 10: Verify the prediction results; The location of the lithium-bearing pegmatite vein was predicted by combining steps five, six, seven, eight, and nine, and then verified by field surveys.

[0103] This method integrates geological, geochemical, and hyperspectral remote sensing techniques for mineral exploration prediction, achieving an organic combination of these methods. This makes the predicted mineral types more targeted and the predicted area locations more accurate, thus significantly improving the prediction accuracy and exploration efficiency of pegmatite-type lithium deposits to meet the growing lithium demand. In particular, the method of extracting hyperspectral remote sensing data allows for difference analysis between the extracted spectral characteristics of relevant spodumene pegmatites and lithium-free pegmatites, enabling the determination of an effective prediction range.

[0104] Meanwhile, compared with traditional lithium exploration methods, this method can first accurately and quickly identify pegmatite-type lithium deposits and make prospecting predictions in areas where it is difficult for humans to reach, thus saving a lot of manpower and material resources.

[0105] Against the favorable metallogenic background of the Himalayan metallogenic belt, this method conducts detailed and comprehensive mineral exploration prediction from large to small scales, avoiding the shortcomings of other existing technologies that only carry out remote sensing interpretation of lithium-bearing pegmatites at the mining area scale or establish short-wave infrared spectral libraries, without prioritizing the acquisition of mineralized areas and making effective mineral exploration predictions for the surrounding areas.

[0106] The present invention has been described above in conjunction with the preferred embodiments, but the present invention is not limited to the embodiments disclosed above, but should cover various modifications and equivalent combinations made in accordance with the essence of the present invention.

Claims

1. A comprehensive prospecting and prediction method for pegmatite-type lithium deposits, characterized in that, Includes the following steps: Step 1: Delineate the large-scale geochemical exploration work area based on small-scale geochemical anomalies; The scope of geochemical work at the standard map sheet of 1:50,000 was delineated based on the metallogenic belt and the 1:200,000 geochemical anomaly. Among them, the 1:200,000 scale geochemical survey is a geochemical survey that has been basically fully covered by the state. The lithium element anomaly at the 1:200,000 scale can provide a direct basis for selecting areas for carrying out larger scale 1:50,000 stream sediment surveys. Step 2: 1:50,000 scale sediment sampling of the river system; 1:50,000 stream sediment measurements were conducted to further delineate geochemical anomaly zones of different levels. The quality control methods and indicators for the 1:50,000 stream sediment measurements were carried out in accordance with the requirements of the China Geological Survey's "Geochemical Survey Specification (1:50000)" (DZ / T 0011-2015) and "Geological and Mineral Laboratory Testing Quality Management Specification" (DZ / T 0130-2006), as well as other quality technical standards, specifications, and systems. According to the "Geochemical Survey Specification" (1:50000) and the landscape geochemical conditions within the survey area, the survey area was divided into three main zones: an intensive sampling zone, a basic sampling density zone, and a non-sampling zone. The basic sampling density for stream sediment measurements was 4-8 points / km², with actual sampling based on the principle of effectively controlling the anomaly range and avoiding continuous blank areas in the survey area. The sampling techniques included: GPS verification, determination of sampling grain size, and determination of sampling points. Step 3: Testing of river sediment samples and data processing and mapping; The testing employed a comprehensive analytical approach combining multiple methods, including inductively coupled plasma mass spectrometry (ICP-MS), X-ray fluorescence spectrometry (XRF), foam adsorption-inductively coupled plasma mass spectrometry (ICP-MS), atomic fluorescence spectrometry (AFS), and emission spectrometry (ES). The tested elements included Au, Be, Co, Cr, Cu, Li, Mo, Nb, Ni, Sb, Sn, U, W, Zn, Pb, Ta, Rb, As, Hg, Ag, Al, Ba, Fe, Cd, and Bi, totaling 25 elements. Geochemical data processing and the compilation of a series of maps were carried out based on 25 elemental indicators analyzed and tested, and were compiled according to standard map sheets. A total of one set of geochemical maps was compiled, including original point data maps, geochemical maps, single-element anomaly maps, combined anomaly maps, comprehensive anomaly maps, mineral exploration prediction maps, and major comprehensive anomaly analysis maps. The coordinate system used for mapping is Gaussian 6-degree zoning with 15 zones, 2000 National Geodetic Coordinate System, and 1:50,000 standard map sheets. Mapping uses MAPGIS as the basic platform, data processing and encryption are implemented in GeoExpl using index weighting, and parameter statistics, histograms, etc. are completed using self-developed programs. Step 4: Geochemical anomaly interpretation and evaluation, which specifically includes the following: (i) Delineation of single-element anomalies; (ii) Delineation of combined anomalies; (iii) Delineation of prospective mineral exploration areas; (iv) Delineation of prospective mineral exploration target areas. Step 5: Geological survey of the mineral exploration target area; Prioritize high-level prospecting target areas to conduct 1:10,000 special geological mapping to find spodumene ore bodies, and deploy large-scale 1:2000 geological profiles and petrogeochemical profiles at the locations of discovered ore bodies; Step Six: Hyperspectral Remote Sensing Data Acquisition and Processing; Using satellites to identify lithium deposits in pegmatites; The main processing steps include: data preparation and import, radiometric calibration, atmospheric correction, bad band removal, and geometric correction; Step 7: Rock surface sampling and feature analysis; Ground spectral information of spodumene pegmatite and ore-free rock bodies in the visible to short-wave infrared range was collected using an ASD spectrometer. The collected results were analyzed and compared with the standard spectra of single minerals in commonly used spectral databases. The spectral characteristics of spodumene pegmatite and related ore-free rock bodies were analyzed to find differences in spectral characteristics, in order to prepare for the extraction of spodumene-bearing pegmatite. Step 8: Extraction of spodumene pegmatite from hyperspectral remote sensing satellite imagery; By analyzing the relevant rock surface spectra, the extraction process of spodumene pegmatite from hyperspectral remote sensing satellite images was determined. First, the range of pegmatite veins was determined using high spatial resolution remote sensing images. Then, the overall waveform characteristics of the hyperspectral remote sensing images were matched with the image spectra of known spodumene pegmatite locations through spectral angle matching. Finally, the extraction accuracy of spodumene pegmatite was further improved by calculating the slope of the image in the 2370nm-2410nm spectral range. Step Nine: Prediction of spodumene pegmatite veins within a specific area; Satellite remote sensing spectral characteristics of spodumene-bearing pegmatites are used as positive samples, while spectral characteristics of spodumene-free pegmatites and spodumene-free granites are used as negative samples. The possible locations of ore bodies are interpreted from satellite remote sensing images based on the spectral characteristics of known ore bodies. Step 10: Verify the prediction results; The location of the lithium-bearing pegmatite vein was predicted by combining steps five, six, seven, eight, and nine, and then verified by field surveys.

2. The comprehensive prospecting and prediction method for pegmatite-type lithium deposits according to claim 1, characterized in that, Step four, section (a), specifically the single-element anomaly identification, is as follows: Based on the geological structure characteristics of the work area, the geological background of the map sheets is relatively stable. Therefore, a unified lower limit for anomaly delineation is adopted for each map sheet. Among them, three methods are used to calculate the lower limit for anomaly: First, the lower limit for anomaly is calculated directly from the original data (encrypted data) according to X±2S; Second, the lower limit for anomaly is calculated by performing a logarithmic transformation on the original data (encrypted data) and then calculating according to X±2S, and converting the result into a true value to obtain the lower limit for anomaly; Third, the data with a cumulative frequency of 90% of the original data (encrypted data) is directly used as the lower limit for anomaly. Six sets of abnormal lower limits were obtained using the above three methods. These six sets of abnormal lower limits were used to delineate anomalies. The average of these six sets of abnormal lower limits was then used to delineate anomalies, which yielded good delineation results. After rounding and adjustment, the abnormal lower limit value was finally determined. Based on the range of 1, 2, and 4 times the abnormal lower limit value, the three-level concentration zones of outer, middle, and inner anomalies were determined.

3. The comprehensive prospecting and prediction method for pegmatite-type lithium deposits according to claim 1, characterized in that, Step four, section (ii), specifically the identification of combined anomalies, is as follows: Based on the single-element anomaly map, and taking into account factors such as cluster analysis, the metallogenic relationship of the elements, and the geological and metallogenic background of the area, the elements are divided into several groups. The main metallogenic elements are represented by surface colors in three zones: inner, middle, and outer. Other elements are represented by line colors, which only indicate the range of the outer anomaly zone.

4. The comprehensive prospecting and prediction method for pegmatite-type lithium deposits according to claim 1, characterized in that, Step four, section (iv), specifically delineating the mineral exploration target area, is as follows: Based on the indicators of prospecting target areas in the "Technical Requirements for Mineral Prospect Survey", the determination of the category of prospecting target areas takes into account factors such as the degree of mineralization favorability, engineering control, mineralization intensity, richness and intensity of comprehensive prospecting information (physical, chemical and remote sensing), sufficiency of prediction basis, and size of resource potential. Prospecting target areas are divided into three levels: A, B, and C. Through comprehensive analysis of the geological and geochemical characteristics of the work area, and based on mineralization regularity and prospecting indicators, three prospecting target areas are delineated.

5. The comprehensive prospecting and prediction method for pegmatite-type lithium deposits according to claim 1, characterized in that, In step eight, for spectral matching of spodumene pegmatite, the spectral angle matching algorithm of ENVI software is used to initially extract the range of spodumene pegmatite.

6. The comprehensive prospecting and prediction method for pegmatite-type lithium deposits according to claim 1, characterized in that, In step eight, for the calculation of the slope of the image spectral interval, the core algorithm for calculating the slope of the image spectral interval is obtained by Python programming: using polynomial fitting to calculate the spectral slope within a specified wavelength range and saving the fitting result to the output image. The polynomial fitting part in the spectralslope() function is specifically implemented using the numpy.polyfit() function.

7. The comprehensive prospecting and prediction method for pegmatite-type lithium deposits according to claim 1, characterized in that, In step nine, the visual interpretation results, spectral matching, and slope calculation results of the pegmatite veins are superimposed to further improve the accuracy of the spodumene pegmatite location; the specific process is as follows: The classification raster file obtained after spectral matching and oblique intersection calculation is converted into a vector file. The intersection is then taken in ArcGIS software to obtain the spodumene pegmatite region extracted from the hyperspectral remote sensing data. The visual interpretation results of the pegmatite veins are then superimposed to identify the veins falling within the spodumene pegmatite region as spodumene pegmatite veins.

Images