CN111103632B - Method for analyzing laterite-nickel ore in prediction area based on remote sensing detection - Google Patents

Abstract

The invention discloses a method for analyzing laterite-nickel ore in a prediction area based on remote sensing detection, comprising the following steps of: s1, obtaining remote sensing data of a prediction area; s2, establishing a remote sensing prospecting model; s3, interpreting remote sensing data according to the remote sensing prospecting model; and S4, remote sensing mining comprehensive prediction. The remote sensing prospecting comprehensive prediction is carried out by the method, the favorable section of the prospecting can be simply and efficiently determined, the target area of the prospecting is defined, and the laterite nickel ore is guided to prospecting.

Description

Method for analyzing laterite-nickel ore in prediction area based on remote sensing detection

Technical Field

The invention relates to a method for analyzing laterite-nickel ore in a prediction area based on remote sensing detection.

Background

Nickel is a hard and ductile metal element having ferromagnetism, is highly polished and resistant to corrosion, and is widely used for alloys (such as nickel steel and nickel silver) and catalyst materials.

Lateritic nickel ore is the main source of metallic nickel, and is formed under the geological background condition of continuous distributed ultrabasic rock. The specific mineralization process is that nickel element is separated out from the super-basic rock through mineralization effects such as oxidation, hydrolysis, carbonation and the like, and then leaching is carried out to secondary enrichment. The joint fissure zone and the structural fracture zone which are developed densely are important mineral-forming elements, and can introduce atmospheric rainfall into the ground, promote water-rock reaction and accelerate the leaching process. Topography is the most important element of ore control, which directly affects the effective accumulation of nickel ore bodies. The accumulation of ore bodies is facilitated at the positions of hills in front of mountains with gentle slopes, low mountain platforms, wide and gentle ridges and the like, and conversely, the accumulation of ore bodies is not facilitated by steep slopes with large slopes and river valleys with deep cutting.

In the aspect of ore deposit characteristics, the laterite type nickel ore has typical vertical zonation, and the characteristic of vertical zonation of the laterite type weathering crust is reflected. The residual red soil cap layer-detritus zone-bedrock are arranged from top to bottom in sequence. The boundary between the detritus zone and the residual laterite cover layer is not obvious, and the detritus zone has gradual change and mutation, and the thickness of the residual laterite cover layer is usually less than 1 m. The detritus band is in primarily graded contact with the bedrock. The rotten rock zone is the most main ore-containing layer, mainly comprises a nickel-containing limonite mineralized clay layer and a nickel-containing semiweathered residual layer, and the thickness of the rotten rock zone is controlled by the development depth of the structure and the water level of underground water. The downward developing joints-fractures bring surface water into the ground, promoting water-rock reactions.

Corresponding to the mineralization background of the laterite-type nickel ore, the prospecting mark mainly comprises basic-ultrabasic rock laterite weathered crust distributed in large area, is the most direct and main prospecting mark of the laterite-type nickel ore, and is the 'source' for forming the laterite-type nickel ore; secondly, a joint fissure zone and a structural fracture zone, which basically control the thickness of a detritus zone (ore-bearing layer) and determine the scale and the enrichment degree of an ore body; surface signs are most importantly topographical features: the parts such as mountain hills, short mountain platforms, wide and gentle ridges with gentle slopes are beneficial to effective accumulation of ore bodies.

The traditional method for searching the nickel laterite ore is mainly chemical exploration and physical exploration in field operation, a large amount of manpower, physical strength and financial resources are required to be consumed, particularly, high-strength field operation is difficult to realize in tropical rainforest areas where the nickel laterite ore is widely distributed, and a high-efficiency remote sensing technology is an important method for searching the nickel laterite ore. At present, the detection of the laterite-type nickel ore by using a remote sensing technology is mainly concentrated on the detection of a single ore finding mark, and the main ore finding mark is not comprehensively analyzed.

The invention summarizes the ore forming model and the ore finding mark of the laterite nickel ore on the basis of the previous work, extracts various ore finding mark information by utilizing the remote sensing technology, establishes the remote sensing ore finding model of the laterite nickel ore, applies the model to the actual work and obtains better results.

Disclosure of Invention

In order to solve the technical problems, the invention aims to provide a method for analyzing and predicting laterite-nickel ore in a region based on remote sensing detection.

In order to solve the technical problem, the invention adopts the following technical scheme:

a method for analyzing laterite-nickel ore in a prediction area based on remote sensing detection comprises the following steps:

s1, obtaining remote sensing data of a prediction area

The remote sensing data comprises optical remote sensing images such as landsat and IKONOS and DEM elevation data;

s2, establishing a remote sensing prospecting model

Establishing a remote sensing prospecting model of the prediction region according to the mineralization mode of the laterite nickel ore of the prediction region;

the remote sensing prospecting model is a series of remote sensing prospecting mark combinations for describing formation and storage of a type of ore deposit; the remote sensing ore finding mark combination mainly comprises a lithology mark, a construction mark, a surface covering mark and a landform and landform mark;

wherein: the lithological mark is mainly a continuously exposed super-basic rock which is a 'source' for forming the laterite-type nickel ore; the construction mark is mainly a linear image, a buffer area within a certain distance is a mineralization favorable area, and the surface covering mark is mainly a bare red soil area; the landform mark is mainly an area with a small gradient; the mountain front hills, the low mountain platforms and the wide and gentle ridge parts are main effective accumulation areas of ore bodies, the linear buffer areas with the length of 150 meters, the gentle areas with the gradient of less than 15 degrees and the bare areas of laterite are areas with favorable conditions for ore formation, and the areas with various favorable conditions for ore formation are favorable sections for ore formation and are key areas for finding ore;

s3, interpreting the remote sensing data according to the remote sensing prospecting model

The method comprises lithologic structure remote sensing interpretation and landform remote sensing interpretation, wherein the lithologic structure remote sensing interpretation and the landform remote sensing interpretation are bare red soil remote sensing interpretation;

s4, remote sensing comprehensive prediction of prospecting;

according to the principle of similar analogy, on the basis of analyzing various favorable conditions of the nickel mining area, various geology and remote sensing interpretation information of the prediction area are comprehensively predicted to comprehensively predict favorable sections of the nickel mining area.

Further, the step S1 specifically includes:

checking whether the original image space coordinate system is correct;

checking whether the cloud and snow amount of the original image exceeds 15%;

control data analysis, which is to check and analyze the reasonability of the position of the control point and the correctness of the coordinate; note that the point positions of the control points should be uniformly distributed, and there should be a common point between adjacent scenes, and the higher the overlap degree is, the higher the priority for laying the common point is; the control area is larger than the prediction area range, control points are added to the mountain land appropriately, and points must be set in the adjacent scene overlapping area;

the DEM quality is checked, in particular the detection of edges and outliers, and corrected.

Further, the step S1 further includes a standard image making process: image registration, image fusion, color mixing after fusion, orthorectification and image mosaic;

1) Image registration

For original data which is not fused with panchromatic data and multispectral data, selecting the same-name feature points with obvious characteristics as registration control points to perform original image registration by utilizing the panchromatic data and the multispectral data corresponding to the panchromatic data; during registration, the registration control points are uniformly distributed in the registration unit, including a mountain region; the registration area is larger than the patch area; the number of the registration control points of each scene is 9-15, and the registration control points are properly added in the mountain land;

and (4) after registration, carrying out registration precision check on the image by adopting a 'sliding window curtain' method, and carrying out the next step of data fusion after the precision is qualified.

2) Image fusion

In the remote sensing image processing process, an IHS (induction heating system) transformation and/or PANSHARP (Pancluster fusion) fusion method is adopted to fuse the images, and the selection of the image fusion method follows the following principle:

the texture information can be clearly expressed, and main ground objects such as water bodies, buildings, cultivated land and roads can be highlighted;

the image spectral characteristics are real, accurate and free of spectral abnormality;

the characteristics of various land types are obvious, the boundary is clear, and different land type information can be distinguished through visual interpretation;

the fused image has uniform tone, moderate contrast and color close to natural and true color.

3) Color mixing after fusion

Adjusting the hue of the fused image by adopting a linear or non-linear stretching method, a brightness contrast ratio method, a color balance method, a chroma method, a saturation method and a brightness method; the images after color mixing treatment have clear texture, uniform color tone, moderate contrast and colors close to natural and true colors, and important ground types can be clearly distinguished.

Further, a typical nickel ore area mainly comprises the following characteristics: the mining area is mainly red brown, the road and supporting facilities are mainly purple red, the image characteristics are obvious, and a large-area red soil exposed area is formed; the substrate of the mining area is super-basic rock, and the mining area is located in three linear buffer areas of 150 meters; the topography of the mining area is relatively flat, the slope is less than 25 degrees, the slope of most mining areas is 5-10 degrees, the elevation of the mining area is between 50-250m, and 50-150m is mainly used.

Further, the step S3 mainly includes: according to the remote sensing ore finding model, firstly, the exposed range of the super-basic rock is interpreted, the spatial distribution of the super-basic rock is defined, secondly, a linear structure comprising joints, cracks and faults is interpreted, then, an exposed laterite weathering crust area is extracted, the landform and the landform are comprehensively interpreted by combining remote sensing images according to the elevation and gradient information extracted by the DEM, particularly, an effective accumulation area of ore bodies with gentle landform is interpreted, and finally, comprehensive analysis is carried out according to the known characteristics of the ore mining area to determine the beneficial mining section of the prediction area.

Further, in the step S4, a prediction area, that is, a linear structure 150m buffer area, a slope of which is less than 15 °, a laterite bare area, and a semi-bare area in the prediction area are subjected to spatial superposition analysis, so as to obtain a multiple-attribute structure; if a certain area has three favorable conditions for mineral formation, namely the area has the most similar mineral formation characteristics with a typical mining area, the area is a first-level favorable mining section for comprehensive prediction of remote sensing mineral finding; if a certain region has two kinds of favorable conditions for ore formation, the method is a secondary favorable ore formation section for remote sensing ore finding comprehensive prediction.

Compared with the prior art, the invention has the beneficial technical effects that:

the remote sensing prospecting comprehensive prediction is carried out through the principle and the method, the favorable section of the prospecting can be determined simply and efficiently, and meanwhile, the remote sensing nickel-based information graph is decoded, so that convenience is provided for prospecting.

Drawings

The invention is further illustrated in the following description with reference to the drawings.

FIG. 1 is a Landsat-8 image;

FIG. 2 is an IKONOS image;

FIG. 3 is a DEM failure diagram;

FIG. 4 is a DEM qualification diagram;

FIG. 5 is a schematic view of an image before color matching;

FIG. 6 is a schematic diagram of a color-mixed image;

FIG. 7 is a mosaic image;

FIG. 8: a is a remote sensing image map; b is a lithologic structure remote sensing interpretation graph; c is a slope remote sensing interpretation graph;

FIG. 9 is a remote sensing prospecting model of a laterite-type nickel ore in a prediction area;

FIG. 10 is a schematic diagram of a lithology remote sensing interpretation mark;

FIG. 11 is a schematic diagram of a remote sensing interpretation mark of a linear image;

FIG. 12 is a lithology construct remote sensing interpretation map;

FIG. 13 is a gradient map based on DEM;

FIG. 14 shows remote sensing image characteristics for different surface coverage types;

FIG. 15 is an explanatory diagram of bare red soil information in a work area;

FIG. 16 is a diagram of remote sensing comprehensive prediction of prospecting;

FIG. 17 is a remote sensing image of the target area of the remote sensing prospecting mine.

Detailed Description

The method of the present invention is described in detail below with reference to a specific embodiment.

Example 1

A method for analyzing laterite-nickel ore in a prediction area based on remote sensing detection specifically comprises the following steps:

remote sensing data acquisition

Remote sensing data of Landsat-8 satellites and IKONOS satellites (see figures 1 and 2) shot in 2013 are used as remote sensing information data. Landsat series satellites and IKONOS satellite parameters are shown in tables 1 and 2. And the corresponding remote sensing interpretation work is completed by combining the geological data map completed at an early stage, the 30-meter DEM and related auxiliary data such as a gradient map, a topographic map, a contour map and the like generated by the DEM.

TABLE 1 Landsat series satellite essential parameters

TABLE 2 IKONOS satellite essential

Data quality analysis

The primary data quality analysis object is a landsat image, an IKONOS image and DEM data, and comprises the following contents:

1. checking whether the original image space coordinate system is correct.

2. And checking whether the cloud and snow amount of the original image exceeds 15 percent or not.

3. The analysis of the control data is carried out, and checking and analyzing the reasonability of the position of the control point and the correctness of the coordinate. Note that the points of the control points should be evenly distributed and there should be a common point between adjacent scenes, the higher the overlap, the higher the priority of laying the common point. The control area is larger than the prediction area range, control points are added to the mountainous region properly, and points must be set in the adjacent scene overlapping area.

4. And checking DEM quality, particularly checking edge connection and abnormal values, and correcting. The comparison of the before and after correction DEM images is shown in fig. 3 and 4.

Standard image production

After 5 processes of image registration, image fusion, color mixing after fusion, orthorectification and image mosaic, the original image is finally made into a standard image for interpretation work.

Image registration

For original data which is not fused with panchromatic data and multispectral data, panchromatic data and multispectral data corresponding to the panchromatic data are used, and homonymy feature points with obvious features are selected as registration control points to conduct original image registration. During registration, the registration control points are uniformly distributed in the registration unit (including mountainous regions), and the registration area is larger than the area range. The number of the registration control points of each scene is between 9 and 15, and the registration control points are added in the mountain land appropriately.

The same source panchromatic and multispectral images synchronously acquired by the method are selected to form a geometric polynomial model, and the order is not more than 2. The method comprises the steps of firstly registering the areas with flat and hilly lands and small side view angles, and reserving the wave band number, sequence and sampling interval of original images for registered images. The resampling method adopts a bilinear interpolation method or a cubic convolution interpolation method.

And (4) after registration, carrying out registration precision check on the image by adopting a 'sliding window curtain' method, and carrying out the next step of data fusion after the precision is qualified.

Image fusion

In the process of processing remote sensing images, the commonly adopted fusion methods include IHS transformation, principal component transformation, weighted product, ratio transformation, wavelet transformation, high-pass filtering, BROVERY, PANSHARP fusion and other methods, wherein the IHS transformation and PANSHARP fusion methods have better effect on image fusion, and the image fusion method is selected according to the following principle:

can clearly express the texture information and can highlight main ground objects (such as water bodies, buildings, cultivated land, roads and the like).

The image spectral characteristics should be true and accurate without spectral anomalies.

The characteristics of various land types are obvious, the boundary is clear, and different land type information can be distinguished through visual interpretation.

The fused image has uniform tone, moderate contrast and color close to natural and true color.

The fusion mainly adopts PANSHARP fusion algorithm. The fusion algorithm can ensure the whole color of the ground object to be real, can also ensure the outline definition of the boundary of the ground object, and is favorable for visual interpretation in the later period.

Color mixing after fusion

And adjusting the hue of the fused image by adopting methods such as linear or nonlinear stretching, brightness contrast, color balance, chroma, saturation, lightness and the like. The images after color mixing treatment should have clear texture, uniform color tone, moderate contrast, and color close to natural color, so that important types can be clearly distinguished. See fig. 5 and 6 for comparison images before and after color matching

Orthographic correction

Orthorectification is usually applied to single scene data for different orbits and different time phases of images. This time using a rational function model for orthorectification.

Image mosaicing

The mosaic lines select obvious boundary lines such as linear ground features or land parcel boundaries and the like so as to eliminate the splicing seams in the mosaic images as much as possible, ensure the integrity of the same land parcel when different images are mosaiced and be beneficial to interpretation. And the embedded image should avoid cloud, fog, snow and other areas with relatively poor quality, so that the embedded part has no crack, blur and double image. When the time phases or the qualities of two adjacent images are not greatly different, the texture and the color of the images are kept in natural transition, and when the time phase difference is large and the difference of the ground feature is obvious, the texture and the color of the images are kept, but the spectral features in the same land are kept consistent. Meanwhile, data with better quality is used as much as possible, if the data is covered by snow and clouds, the attractiveness of the joint line is not required to be considered when the data with better quality is used, and therefore the overall cloud amount of the data is less and the visual interpretation influence of ground objects is minimized.

And inlaying the corrected image with the same sampling interval with the superposition accuracy meeting the requirement. As shown in fig. 7.

Remote sensing interpretation platform

The remote sensing image interpretation work is carried out in ENVI, PCI, ERDAS and eCoginization, comprehensive prediction analysis is carried out in ArcMap, and the final result is stored in a MapGIS recognizable format.

Remote sensing prospecting model

Ore forming mode of laterite-nickel ore in prediction area

1. Geological background

The method mainly comprises the steps of developing continuously distributed basic and super-basic rocks in a prediction area, wherein the lithology is mainly serpentine olivine (Du), periclase (Hz), pyroxene (Pyr) and serpentine (S), and in the super-basic rocks, later rock walls or dikes are developed, and the rocks usually have serpentine with different degrees. Most ultrabasic rocks have weathering changes at different degrees on the top and are covered by brownish red and yellowish red weathering crust, the thickness of the weathering layer is generally 5-10m, and the thickness of the weathering layer in individual sections can reach more than 20 m. The wide distribution of basic and super basic rocks creates a good mineral source for the laterite-type nickel ore.

The loose packing of the predicted zone is widely distributed, which presents no small difficulty for lithology-construction interpretation.

2. Process of forming ore

The nickel element is separated out from the ultrabasic rock through the mineralization effects of oxidation, hydrolysis, carbonation and the like, and then leached to be enriched for secondary. The densely developed linear structures (joint fracture zone and structural fracture zone) are important mineral-forming elements, and can guide atmospheric rainfall into the ground, promote water-rock reaction and accelerate the leaching process. Topography is the most important mineral control element, which directly affects the effective accumulation of nickel ore bodies. The accumulation of ore bodies is facilitated at parts such as mountain hills with gentle slopes, low mountain platforms, wide and gentle ridges and the like, and conversely, the accumulation of ore bodies is not facilitated at steep slopes with larger slopes and valleys with deeper cuts.

3. Characteristics of the deposit

The laterite type nickel ore has typical vertical zonation, and reflects the vertical zonation characteristic of the laterite type weathering crust. The residual red soil cap-detritus zone-bedrock are formed from top to bottom in turn. The boundary between the detritus zone and the residual laterite cover layer is not obvious, and the detritus zone has gradual change and mutation, and the thickness of the residual laterite cover layer is usually less than 1 m. The detritus band is in primarily progressive contact with the bedrock. The detritus zone is the most important mineral-containing layer, mainly including nickel-containing limonite mineralized clay layer and nickel-containing semiweathered residual layer, and the thickness of the detritus zone is controlled by the depth of the structure development and the groundwater level. The downward-developing joints-fractures bring surface water into the ground, promoting water-rock reactions.

4. Ore-finding mark

The mineral body is mainly produced in a complete weathering zone of the serpentine olivine and the Fanghui olivine, basic-ultrabasic rock weathered shells are distributed in a large area, and the laterite weathered shells are the most direct and main mineral finding marks of the laterite nickel ore and are the 'sources' for forming the laterite nickel ore. The prediction area is almost covered by a mauve and dark red iron clay layer, and the mineralization area and the non-mineralization area have larger contrast in the aspects of soil, vegetation and the like. Although the mineralization area is thick in weathering layer, because the iron content is high, the vegetation does not develop, only the shrubs with low, short, thin and sparse properties exist, most shrubs are half-covered areas, and although the weathering crust is thin, the tropical rain forest develops quite well, and the vegetation covers well. The second is the joint fissure zone and the tectonic fracture zone, which basically controls the thickness of the detritus zone (ore-bearing layer) and determines the size and enrichment degree of ore body. The most important of the surface signs are the landform, such as the hills with gentle slopes, the low mountain platforms, the wide and gentle ridges and the like, which are beneficial to the effective accumulation of ore bodies.

5. Typical field analysis

Near the prediction zone, there is a laterite nickel ore site being mined. The mining area is mainly reddish brown, the supporting facilities such as roads are mainly purplish red, the image characteristics are obvious, and the laterite bare area with large area is provided (as shown in figure 8 a). The basement of the mine is ultrabasic rock with two near-south-north and one north-east linear formations passing through the mine, all located within the 150 meter buffers of the three linear formations (see fig. 8 b). The terrain of the mining area is relatively flat, the slopes are all less than 25 degrees, and the slopes of most mining areas are 5 degrees to 10 degrees (as shown in fig. 8c, 1 represents that the slopes are less than 5 degrees, 2 represents that the slopes are 5 degrees to 10 degrees, 3 represents that the slopes are 10 degrees to 15 degrees, and 4 represents that the slopes are 15 degrees to 25 degrees). The elevation of the mining area is also between 50 and 250m, and is mainly between 50 and 150 m.

The lithology-structure condition, the earth surface covering characteristic, the landform, particularly the slope, the elevation and other characteristics of a typical mining area well reflect the ore forming mode of the laterite-type nickel ore in the area, have typical representativeness and provide a foundation for establishing a remote sensing ore finding model of a prediction area.

Remote sensing prospecting model for prediction area

The remote sensing prospecting model is a combination of a series of remote sensing prospecting marks for describing formation and storage of a type of ore deposit under the current technical condition. The remote sensing prospecting mark of the laterite-type nickel ore in the prediction area mainly comprises a lithological mark, a construction mark, a surface covering mark and a topographic feature mark. The lithologic sign is mainly a continuously exposed ultrabasic rock which is a 'material source' for forming the laterite-type nickel ore; the construction mark is mainly a linear image, and a buffer area within a certain distance is an ore-forming favorable area; the ground surface covering mark is mainly a naked red upper area, which is beneficial to infiltration of atmospheric rainfall, promotes water rock reaction and accelerates the leaching process; the landform signs are mainly regions with small gradient, such as forward hills, short mountain platforms, wide and gentle ridges and the like, which are main effective accumulation areas of ore bodies. The 150m buffer zone with linear structure, the gentle area with gradient less than 15 degrees and the laterite bare area are areas with favorable conditions for mineral formation, and the areas with various favorable conditions for mineral formation are favorable sections for mineral formation and are key areas for mineral formation. The remote sensing prospecting model of the laterite-type nickel ore in the prediction area is shown in figure 9.

According to the remote sensing prospecting model, the work firstly needs to interpret the exposure range of the ultra-basic rock and define the spatial distribution of the ultra-basic rock. It interprets linear structures including joints, fractures, faults, etc. And then extracting the exposed laterite weathering crust area, comprehensively interpreting the landform and the landform by combining a remote sensing image according to the elevation and gradient information extracted by the DEM, particularly interpreting the effective ore body accumulation area with gentle landform, and finally comprehensively analyzing according to the known characteristics of the mining area to determine the mining favorable area of the prediction area.

Remote sensing geological interpretation

Remote sensing geological interpretation method

With the progress of remote sensing technology and computer technology, the geological remote sensing interpretation method has been developed from the past pure visual interpretation into human-computer interaction interpretation which can fully utilize the characteristics of spatial resolution, spectral characteristics, time and the like of remote sensing images. The remote sensing interpretation is carried out on the digital image which is accurately geometrically corrected, when the types and the ground objects are identified and the characteristics of the ground objects are judged, the image is processed at any time, signals are enhanced or improved, the images are amplified or reduced, and the spectral characteristics and the geometric data of all parts can be measured at any time. General remote sensing interpretation methods mainly include 4 methods, namely an interpretation method, a search method, a similarity method and a comprehensive analysis method, and the 4 methods are often comprehensively used in the actual interpretation process.

A translation method: the geological information such as rock stratums, rock masses, structures and the like is directly extracted from the image map by using the interpretation marks. The method is mainly used for delineating the boundary of the geologic body, and has obvious effect.

A pursuit method: according to the unclear traces displayed on the image by the spreading or extending rules of various information such as strata, rock mass, geological structures and the like, tracking and tracing are carried out, and geological boundary lines, linear structures and the like are defined or sketched. The method is mainly used for delineating the boundary, the fold turning end and the large fracture of the geologic body, and has obvious effect.

A similarity method: the attribute of the geologic body or the geological phenomenon with certain remote sensing hidden information characteristics in the adjacent areas is deduced by taking the image characteristics of the known geologic body or the geological phenomenon as reference.

Comprehensive analysis method: when the interpretation mark is not obvious and the interpretation is difficult, the generation condition and the control condition of the causal relationship of the control geological unit can be comprehensively analyzed, and finally the purpose of interpretation is achieved. For example, in the process of linear image interpretation, the length, fracture property, joint zone or cleavage zone, rock control distribution rule, and the mutual relationship among linear structures in different directions are mainly aimed at.

In consideration of multi-source and multi-scale remote sensing images used in the project, the remote sensing image interpretation of each thematic element adopts various methods of combining human-computer interaction and visual interpretation, combining image interaction interpretation with different resolutions, combining image interaction interpretation with different enhancement modes and combining comprehensive research and known data.

In principle, remote sensing interpretation generally adopts an interpretation method from "face → line → point" to "point → line → face", namely from macro to micro, and then from micro to macro, and the interpretation is gradual and repeated, and finally the purpose of accurate interpretation is achieved.

Lithological structure remote sensing interpretation

The main exposed lithological unit in the prediction area is basic-ultrabasic rock (MTosu), the lithology is mainly serpentine olivine, and the unit is a 'material source' for forming the nickel ore with the shape of red top, and is the most important ore-containing layer. The quaternary loose deposits (Q) are distributed widely in the prediction area, mainly flood-fill deposits and residual slope deposits. The image difference between the basement-ultrabasement rock and loose accumulations is significant. The basic-super-basic rock is mainly in green-dark green color, and purple red or gray brown patches are locally clamped, which is the display of bare soil; the speckled texture is fine and smooth; the river development is basically avoided, and the ridge is relatively rounded; there was essentially no trace of human activity (fig. 10 a). Loose accumulations are mainly gray-brown, which is the display of cultivated land, and irregular green patches or strips are locally clamped, which is the display of vegetation; mainly comprising regular strip-shaped textures (cultivated land) and linear textures (road); the terrain is flat, and the image is basically displayed without shadow; human activities were strong, water system developed, and many David lakes (FIG. 10 b).

The lithology information of the prediction area is interpreted according to the remote sensing interpretation mark and the general method of remote sensing geological interpretation, and the result is shown in figure 12. It should be noted that although the quaternary loose accumulation is decoded, the basement in the prediction area is the basic-super basic rock, which determines that the whole prediction area is the area to be predicted for mineralization prediction.

The main linear structures of the prediction region are fault, joint and crack. According to the remote sensing geological principle, the interpretation signs of the fault are as follows:

(1) The negative terrain is linear in shape, such as rivers, valleys, lakes and basins, has obvious directionality, extends far and is different from general erosion negative terrain. As shown in fig. 11.

(2) And the terrains, beaks, cliffs or fault triangular surfaces are linearly distributed. The steep bank, the steep cliff and the bealock on the terrain are linearly distributed and extend for a certain distance, which is the characteristic of most fractures, and fractures younger or with new activities often have fault triangular surfaces.

(3) Mountain or stratum dislocation. The two disks are twisted relatively to form a linear abnormal image.

(4) The image has a distinct linear image of a right angle bend of the water system, a dislocation of the ridge, and a crossing of the ridge.

(5) The adjacent water systems are synchronously twisted. This is one of the topographical displays of a translation fault, and is typical of synchronous distortion of a tributary parallel water system.

(6) The landscape features on both sides are different. Large-scale fracture, landscape features such as landform, water system, color tone, shadow and the like on two sides are often greatly different.

Joints and cracks are mainly linear or nearly linear furrows and a plurality of linearly-distributed smaller-scale furrows. Compared with the remote sensing interpretation mark of the fault, the linear image has no obvious image characteristics such as synchronous bending of a water system, dislocation of a ridge and the like. As shown in fig. 11.

According to the image characteristics and the remote sensing interpretation mark, the interpretation of the linear structure of the prediction area is completed by adopting a general method of remote sensing geological interpretation, and the lithology-structure remote sensing interpretation result is shown in figure 12.

Statistics shows that the area of bare basic-super basic property of the prediction region exceeds 55km2, accounts for 45% of the total area (123 km 2) of the prediction region, the rest regions are shallow covering basic-super basic rock, and the condition of a mining source is good.

Remote sensing interpretation of landform and landform

According to the mineralization mode of the laterite-nickel ore, the gentle area with a smaller gradient is a main effective accumulation area of an ore body, such as a mountain front, a low mountain platform, a wide and gentle ridge and the like. The gradient is the most important landform factor of the ore formation, and the flat and gentle area with the gradient less than 15 degrees is the most favorable ore formation. Therefore, the main interpretation element of the terrain and features of the predicted area is the slope. Based on DEM with 30 m spatial resolution, the gradient of the prediction area is divided into six levels of 0-5 degrees, 5-10 degrees, 10-15 degrees, 15-25 degrees, 25-35 degrees and 35-90 degrees. The slope of the prediction area is divided through a terrain analysis module of ArcMap software and is mapped, and the result is shown in FIG. 13.

The area of each grade in the prediction region is shown in table 1.

TABLE 1 area statistics table for each grade

Grade of grade	Area (Km 2)	Ratio (%)
			0°-5°	72.57	59.00
5°-10°	22.62	18.39
			10°-15°	8.80	7.16
15°-25°	14.70	11.95
			25°-35°	3.91	3.17
35°-90°	0.43	0.35

As can be seen from Table 1, the terrain of the prediction area is flat, and the area with the gradient less than 5 degrees exceeds 72km2 and is close to 60 percent of the total area of the prediction area; the area of the favorable mineralised terrain with a gradient of less than 15 deg. exceeds about 104km2, exceeding 84% of the total area of the prediction zone. The laterite-type nickel ore in the prediction area has superior ore-forming terrain conditions.

Remote interpretation of dew red soil

According to an ore-forming model, the bare laterite weathering crust is beneficial to infiltration of atmospheric rainfall, promotes water-rock reaction, accelerates the leaching process and is a beneficial section of ore formation. The main earth surface coverage types of the prediction area are three types, namely a laterite bare area, a half laterite bare area and a non-laterite bare area. The remote sensing characteristics of the three coverage types are obvious and the discrimination is high, as shown in fig. 14a, 14b and 14c.

In contrast to vegetation, laterites are highly reflective at the red (R) band and low reflective at the near infrared slope (NIR), and the laterite information is enhanced with a normalized vegetation index NDVI (NDVI = (NIR-R)/(NIR + R)). The low value of NDVI is a lateritic bare area, the high value of NDVI is a non-lateritic covered area (vegetation cover), and the medium value of NDVI is a half lateritic bare area. And extracting the earth surface coverage information by adopting a density segmentation method. As in fig. 15.

The areas of the various types of land cover within the predicted area are shown in table 2.

TABLE 2 area statistics table for various land cover types

Grade of grade	Area (km 2)	Ratio (%)
			Semi-red soil bare	13.42	10.91
Red soil bare	17.27	14.04
			Non-laterite bare	92.35	75.08

As can be seen from Table 1, the areas of the laterite bare and half laterite bare areas in the prediction area exceed 30km2, and are close to 25 percent of the total area of the prediction area, so that the areas are favorable mineralization sections of the laterite type nickel ores in the prediction area.

Principle and method

The nickel mineralization information extraction and mineralization prediction based on remote sensing geological interpretation are based on the principle of similar analogy, namely that the various geological and remote sensing interpretation information of a prediction region is comprehensively predicted to comprehensively predict the favorable section of nickel mineralization of the prediction region on the basis of analyzing various mineralization favorable conditions of a typical nickel mineralization region from the fact that the known conjecture is unknown.

And (3) taking a source as a basis, defining a bedrock-ultrabedrock area as a prediction area, and performing spatial superposition analysis on a linear structure buffer area with the length of 150m, a slope smaller than 15 degrees, a laterite bare area and a semi-bare area in the prediction area to obtain a multiple-attribute structure. If a certain area has three kinds of favorable conditions for mineral formation, namely the mineral formation characteristics which are most similar to those of a typical mining area, the area is a first-level favorable mining section for comprehensive prediction of remote sensing mineral finding. If a certain area has two kinds of favorable conditions for mineral formation, the method is a secondary favorable mineral formation section for comprehensive prediction of remote sensing mineral exploration.

Comprehensive remote sensing prospecting result

The remote sensing prospecting comprehensive prediction is carried out according to the principle and the method, favorable sections of the formed ore are determined, target areas of the prospecting are defined, and a remote sensing prospecting comprehensive prediction image is shown in figure 16.

I. The II, III, IV, V and VI target areas are mainly positioned in the south of the prediction area and comprise a primary mineralization favorable area and a secondary mineralization favorable area with a certain area, and the target areas comprise more than two mineralization favorable conditions.

The remote-sensed images of the respective target regions are shown in fig. 17.

The target area of the mineral exploration defined by the remote sensing technology is the key area of the next work, and key exploration is carried out by applying other technical methods and means such as geophysical prospecting, chemical prospecting and drilling.

The research on the mineralization mode of the laterite-type nickel ore is strengthened, in particular to the research on the influence mechanism of a river on the enrichment of an ore body and the later-stage transformation mechanism of the ore body.

The above-described embodiments are only intended to illustrate the preferred embodiments of the present invention, and not to limit the scope of the present invention, and various modifications and improvements made to the technical solution of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims (2)

1. A method for analyzing laterite-nickel ore in a prediction area based on remote sensing detection is characterized by comprising the following steps:

s1, obtaining remote sensing data of a prediction area

The remote sensing data comprises a remote sensing image and DEM elevation data;

the step S1 specifically includes:

checking whether the original image space coordinate system is correct;

checking whether the cloud and snow amount of the original image exceeds 15 percent;

control data analysis, which is to check and analyze the reasonability of the position of the control point and the correctness of the coordinate; note that the point locations of the control points should be uniformly distributed, and there should be a common point between adjacent scenes, the higher the overlap degree is, the higher the priority of laying the common point is; the control area is larger than the prediction area range, control points are added to the mountain land appropriately, and points must be set in the adjacent scene overlapping area;

checking DEM quality, particularly checking edge connection and abnormal values, and correcting;

s2, establishing a remote sensing prospecting model

Establishing a remote sensing prospecting model of the prediction area according to the mineralization mode of the laterite-nickel ore of the prediction area;

the remote sensing prospecting model is a series of remote sensing prospecting mark combinations for describing formation and storage of a type of ore deposit; the remote sensing prospecting mark combination mainly comprises a lithology mark, a construction mark, a surface covering mark and a landform mark;

wherein: the lithological mark is mainly continuous exposed ultra-basic rock and is a 'source' for forming the laterite-type nickel ore; the structure mark is mainly a linear image, a buffer area within a certain distance is a mining favorable area, and the ground surface covering mark is mainly a bare red soil area; the landform signs are mainly areas with small gradient; the mountain front hills, the low mountain platforms and the wide and gentle ridge parts are main effective accumulation areas of ore bodies, the linear buffer areas with the length of 150 meters, the gentle areas with the gradient of less than 15 degrees and the bare red soil areas are areas with favorable conditions for ore formation, and the areas with various favorable conditions for ore formation are favorable sections for ore formation and are key areas for finding ores;

s3, interpreting the remote sensing data according to the remote sensing prospecting model

The method comprises lithologic structure remote sensing interpretation and landform remote sensing interpretation, wherein the exposed laterite remote sensing interpretation is interpreted;

s4, remote sensing comprehensive prediction of prospecting;

according to the principle of similar analogy, on the basis of analyzing various favorable conditions of the formed ores of a typical nickel ore area, comprehensively predicting favorable sections of the nickel formed ores of the prediction area by comprehensively predicting various geological and remote sensing interpretation information of the prediction area;

a typical nickel mine primarily includes the following features: the mining area is mainly red brown, the road and supporting facilities are mainly purple red, the image characteristics are obvious, and a large-area red soil exposed area is formed; the basement of the mining area is super-basic rock, and the mining area is located in three buffer areas of 150 meters in linear structure; the landform of the mining area is relatively flat, the slope is less than 25 degrees, the slope of most mining areas is 5 degrees to 10 degrees, the elevation of the mining area is between 50m and 250m, and the elevation of the mining area is mainly 50m to 150 m;

the step S3 mainly includes: according to a remote sensing ore finding model, firstly, interpreting the exposure range of the super-basic rock, delineating the spatial distribution of the super-basic rock, then interpreting a linear structure comprising joints, cracks and faults, then extracting an exposed laterite weathering crust area, comprehensively interpreting the landform and the landform by combining a remote sensing image according to the elevation and gradient information extracted by DEM, particularly interpreting an effective ore body accumulation area with gentle terrain, and finally, comprehensively analyzing according to the known characteristics of an ore mining area to determine an ore forming favorable area section of a prediction area;

in the step S4, performing spatial superposition analysis on the prediction area, namely a linear structure buffer area with the length of 150m, the slope of less than 15 degrees, a red soil bare area and a semi-bare area in the prediction area to obtain a multiple attribute structure; if a certain area has three favorable conditions for mineral formation, namely the area has the most similar characteristics of mineral formation to typical mining areas, the area is a first-level favorable mining section for comprehensive prediction of remote sensing mineral prospecting; if a certain area has two kinds of favorable conditions for ore formation, the method is a secondary favorable ore formation section for comprehensive prediction of remote sensing ore finding.

2. The method for analyzing and predicting lateritic nickel ores in a district based on remote sensing detection according to claim 1, characterized in that the step S1 further includes a standard image making process: image registration, image fusion, color mixing after fusion, orthorectification and image mosaic;

1) Image registration

For original data which is not fused with panchromatic data and multispectral data, selecting the same-name feature points with obvious characteristics as registration control points to perform original image registration by utilizing the panchromatic data and the multispectral data corresponding to the panchromatic data; during registration, the registration control points are uniformly distributed in the registration unit, including a mountain region; the registration area is larger than the patch area; the number of the registration control points of each scene is 9-15, and the registration control points are added in the mountain land appropriately;

after registration, the images are subjected to registration precision check by adopting a 'sliding window curtain' method, and after the precision is qualified, the next step of data fusion is carried out;

2) Image fusion

In the remote sensing image processing process, an IHS (induction heating system) transformation and/or PANSHARP (Pancluster fusion) fusion method is adopted to fuse the images, and the selection of the image fusion method follows the following principle:

the texture information can be clearly expressed, and the main ground objects such as water bodies, buildings, cultivated land and roads can be highlighted;

the image spectral characteristics are real, accurate and free of spectral abnormality;

the characteristics of various land types are obvious, the boundary is clear, and different land type information can be distinguished through visual interpretation;

the fused image has uniform tone, moderate contrast and color close to natural true color;

3) Color mixing after fusion

Adjusting the hue of the fused image by adopting a linear or non-linear stretching method, a brightness contrast ratio method, a color balance method, a chroma method, a saturation method and a brightness method; the images after color mixing treatment should have clear texture, uniform color tone, moderate contrast, and color close to natural color, so that important types can be clearly distinguished.

Images