Extraction of Remote Sensing Alteration Information Based on Integrated Spectral Mixture Analysis and Fractal Analysis

Extraction of Remote Sensing Alteration Information Based on Integrated Spectral Mixture Analysis and Fractal Analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Extraction of Remote Sensing Alteration Information Based on Integrated Spectral Mixture Analysis and Fractal Analysis Kai Qiao, Tao Luo, Shihao Ding, Licheng Quan, Jingui Kong, Yiwen Liu, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6598339/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract As a key target area in China's new round of strategic mineral exploration initiatives, Tibet possesses favorable metallogenic conditions shaped by its unique geological evolution and tectonic setting. In this paper, the Saga region of Tibet is the research object, and Level-2A Sentinel-2 imagery is utilized. By applying mixed pixel decomposition, interfering endmembers were identified, and spectral unmixing and reconstruction were performed, effectively avoiding the drawback of traditional methods that tend to remove mineral alteration signals and masking interference. Combined with band ratio analysis and principal component analysis (PCA), various types of remote sensing alteration anomalies in the region were extracted. Furthermore, the fractal box-counting method was employed to quantify the fractal dimensions of the different alteration anomalies, thereby delineating their spatial distribution and fractal structural characteristics. Based on these results, two prospective mineralization zones were identified. The results indicate that:(1) In areas of Tibet with low vegetation cover, applying spectral mixture analysis (SMA) effectively removes substantial background interference, thereby enabling the extraction of subtle remote sensing alteration anomalies. (2) The fractal dimensions of various remote sensing alteration anomalies were calculated using the fractal box-counting method over a spatial scale range of 0.765 to 6.123 km. These values quantitatively characterize the spatial fractal properties of the anomalies, and the differences in fractal dimensions among alteration types reflect the spatiotemporal heterogeneity of the mineralization system. (3) The high-potential mineralization zones identified in the composite contour map of fractal dimensions of alteration anomalies show strong spatial agreement with known mineralization sites. Additionally, two new prospective mineralization zones were delineated in their periphery, providing theoretical support and exploration targets for future prospecting in the study area. Earth and environmental sciences/Solid earth sciences Earth and environmental sciences/Space physics Physical sciences/Engineering Tibet Sentinel-2 spectral mixture analysis Fractal Theory Alteration Information Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 1. Introduction Hydrothermal alteration of wall rocks near ore bodies is primarily the result of interactions between various types of hydrothermal fluids and surrounding rocks, representing the imprints of the progressive enrichment and concentration of ore-forming materials. Quantitative information related to geological structures, ore-bearing geological bodies, and alteration associated with metal mineralization can be comprehensively captured and interpreted through remote sensing-derived alteration signals across various data sources. These signals serve as indicators of potentially altered rocks proximal to mineralization zones, and the alteration information extracted via remote sensing provides a geological basis for guiding mineral deposit prediction (Crowley 1984 ; Asadzadeh et al. 2024 ; Corrêa da Costa et al. 2024 ; Habashi Costa et al. 2024; Islam et al. 2024 ; Yang et al. 2024 ). As a key target area in China's new round of strategic mineral exploration initiatives, Tibet possesses favorable metallogenic conditions shaped by its unique geological evolution and tectonic setting (Huang et al. 2024 ; Lin et al. 2024 ; Zhao et al. 2024 ). Although the region's harsh natural environment imposes multiple technical challenges on conventional surface exploration methods, its sparse vegetation cover offers a distinct advantage for accurately identifying and extracting alteration anomalies using remote sensing techniques (Dong et al. 2022 ; Bai et al. 2023 ; Fu et al. 2023 ). Therefore, developing a high-precision remote sensing model for alteration information extraction has become urgent in advancing Tibet's new phase of mineral prospecting studies. At present, methods such as band ratio analysis (Abdelkader et al. 2024 ; Baid et al. 2024 ; Carvalho et al. 2025 ; Yang et al. 2025 ), principal component analysis (PCA) (Mbianya et al. 2021 ; Abedini et al. 2023 ; Hegab 2024 ; Hussain et al. 2025 ), spectral mixture analysis (SMA) (Zhao and Zhao 2019 ; Chen et al. 2022 ; Chen et al. 2025 ), spectral angle mapper (SAM) (Ren et al. 2022 ; Ghoneim et al. 2024 ; Yang and Tian 2024 ), and minimum noise fraction (MNF) transformation (Sikakwe et al. 2023; Yao et al. 2023 ; Yuan et al., 2025 )are widely applied in the extraction of mineralization-related alteration information. Among these methods, SMA has gradually attracted increasing attention from researchers due to its high accuracy, intuitive interpretability, and robustness. In remote sensing imagery, pixels represent the fundamental units of an image. However, due to spatial resolution limitations and small-scale surface features, a single pixel often contains spectral information from two or more distinct land cover types. Such pixels are referred to as mixed pixels (Naceur et al. 2004 ; Miao et al. 2007 ; Hosseinjani and Tangestani 2011 ). The spectral characteristics of remote sensing imagery represent a mixed response of the spectral signatures of various endmembers comprising each pixel. During the extraction of alteration information from remote sensing images, the widespread presence of mixed pixels significantly reduces the accuracy and reliability of land cover identification. Interfering surface features such as vegetation, water bodies, shadows, and fluvial sediments can severely affect the effectiveness of alteration anomaly detection. Therefore, to accurately identify the spectral signatures of useful endmembers, it is essential to perform spectral mixture analysis on the imagery. This process helps to minimize or eliminate the influence of interfering land cover types, improve spectral discrimination accuracy, and ultimately enhance the extraction of alteration information (Sharifi and Hosseingholizadeh 2019 ). Complex and irregular phenomena in nature often follow inherent scale-invariant laws. Based on this principle, Mandelbrot introduced the concept of fractal theory in 1983 (Mandelbrot 2019). As an effective method for quantitatively describing irregular shapes and spatial patterns, fractal theory has been widely applied to studying natural phenomena and complex systems (Kong and Ding 1991 ; Cheng 1995 ; Zuo and Wang 2020 ). This theory enables the extraction of deterministic and regular parameters from seemingly chaotic patterns, allowing the reconstruction of formation mechanisms of fractal structures and the inference of system evolution from stochastic processes. Many geological phenomena and spatial distributions in Earth systems exhibit pronounced fractal characteristics, particularly those associated with hydrothermal mineralization processes such as magma generation, migration, and intrusion, which often display varying degrees of self-similarity. On this basis, fractal theory serves as a powerful tool for identifying and extracting remote sensing-based alteration anomalies, thereby revealing the spatial organization and evolutionary patterns of mineralization. In recent years, fractal-based methods have been widely employed by researchers worldwide for remote sensing anomaly extraction and mineral exploration (Sun et al. 2018 ; Liu et al. 2021 ; Zhao et al. 2022 ; Chen et al. 2012). The Saga region of Tibet is situated in a key segment of the central-western India–Eurasia continental collision zone. It spans three major tectonic units from north to south: the Gangdese forearc basin, the Yarlung Zangbo suture zone, and the North Himalaya Tethyan fold-thrust belt. Although the area is characterized by a complex lithostratigraphic framework shaped by multiple phases of superimposed tectonic-magmatic activity, its high-altitude terrain, rugged topography, and limited accessibility pose significant challenges to field-based geological investigations. In this context, the present study focuses on the Saga region of Tibet. It employs Level-2A Sentinel-2 imagery to extract remote sensing alteration anomalies, including iron-staining, Al–OH, Mg–OH, and carbonate anomalies by the method of "resampling + mixed pixel decomposition + principal component analysis + mixing ratio + threshold segmentation" in this region. Furthermore, the fractal dimensions of the various alteration anomalies were calculated using fractal theory, enabling the characterization of their spatial distribution patterns and fractal structural features across the study area. Two potential areas have been identified using the composite contour map of fractal dimensions of alteration anomalies obtained from the superimposed analysis of various types of remotely sensed alteration anomalies (Fig. 1 ). 2. Overview of the Study Area 2.1. Geological Profile The study area is characterized by a wide distribution and diverse exposure of stratigraphic units, primarily comprising Paleozoic, Mesozoic, and Cenozoic sequences. Paleozoic strata are extensively exposed in the northern part of the region, while Mesozoic and Cenozoic formations are distributed throughout the entire area. The Paleozoic strata mainly include the Silurian–Ordovician Dajiling and Ziqu formations and the Permian Zhongba, Gangzhutan, and Gazhale formations. The Mesozoic sequences consist of the Triassic Kagongyan, Yachuyan, Niukuyan, and Tunjuri formations; the Jurassic Jipuyan, Jiabula, Angren, Chuangde, and Zongzhuo formations; and the Cretaceous Zezuweng, Dasangyan, and Danga formations. The Cenozoic deposits are dominated by Quaternary alluvial sediments (Fig. 2 ). Intrusive rocks are widespread across the region and are primarily composed of Neogene monzogranite and Cretaceous diabase. The main types of mineralization identified in the area include malachite, epidote, chlorite, Sericite, hematite, and limonite alterations. 2.2. Overview of Remote Sensing Data The Sentinel-2 satellite (Forouzan and Arfania 2020 ; Ding et al. 2023 ; Khashaba et al. 2023 ; Zhang et al. 2024 ; Amani et al. 2025 ; Boumahdi et al. 2025 ), part of the European Union's Copernicus Earth observation program, is equipped with a multispectral imager covering 13 spectral bands across the visible, near-infrared, and shortwave infrared regions. These bands offer spatial resolutions of 10 m, 20 m, and 60 m, respectively. This study utilized Level-2A Sentinel-2 imagery acquired on June 13, 2024, with satisfactory overall data quality (Fig. 3 a). Although Level-2A products provide bottom-of-atmosphere reflectance data that have already undergone radiometric calibration and atmospheric correction, reconstruction of the low-resolution bands is still required during the image processing stage. In this study, the S2 Resampling tool provided by SNAP software was employed to reconstruct the low-resolution Sentinel-2 bands to a spatial resolution of 10 m while preserving their spectral characteristics. For the subsequent spectral mixture analysis and remote sensing-based extraction of alteration anomalies, specific Sentinel-2 bands were excluded based on their primary functions: Bands 1, 9, and 10 are mainly designed for coastal aerosol, cirrus clouds, and water vapor detection, respectively; additionally, Bands 8 and 8a have similar central wavelengths, but Band 8a is more sensitive to rock alteration. Therefore, ENVI software was used in this study to synthesize the Sentinel-2 data in bands other than 1, 8, 9, and 10. The interference, such as the fourth system and rain and snow in the study area, was masked to obtain the processed Sentinel-2 image data (Fig. 3 b) to carry out the study's next step. 3. Research Methods 3.1. Sequential Maximum Angle Convex Cone Several models have been developed for spectral mixture analysis, among which the Linear Spectral Unmixing (LSU) model is the most widely applied due to its simplicity and ease of implementation (Zhang et al. 2010 ; Bhakthan et al. 2024). The LSU technique separates the spectra of each remote sensing data image by utilizing the research results, such as the combination of feature spectra for the characteristics of the mixed image elements. In the linear model of mixed pixels, the gray value of any pixel in given band reflects the co-reflections in that band from all places inside it, and its magnitude is linearly related to the area of the feature. The result of linear spectral separation is usually expressed as a series of gray-scale images of end-element spectra, also known as abundance images, in which the value of the image element indicates the proportion of the end-element spectra in the image element. The selection of end-elements is the prerequisite and key to the linear hybrid image decomposition method, and the type, number, and value of end-elements determine the success or failure of the hybrid image decomposition technique. The end elements are selected on the spectral image to represent the spectrum of the pure feature surface. The Sequential Maximum Angle Convex Cone (SMACC) method is a commonly used end-element extraction method for the Linear Spectral Unmixing model, which can simultaneously extract the end elements and their abundance proportion of each feature from the mixed image (Thompson et al. 2010 ; Filippi et al. 2022 ; Hamedianfar et al. 2024). The SMACC model's fundamental assumption is that a surface comprises a limited number of spectral endmembers. In a three-dimensional data cube representing remote sensing imagery—consisting of spatial dimensions (x-axis and y-axis) and spectral bands—the number of spectral bands is denoted by L, and the number of spectral endmembers by p. Each pixel within this cube can be interpreted as a vector in an L-dimensional spectral space, with each dimension corresponding to a specific spectral band. Thus, the spectral signature of each pixel is expressed as a linear combination of p spectral endmember vectors. Consequently, the original remote sensing image data can be mathematically described by: R = X + Noise = MS + Noise Here, M is the mixing matrix, where each vector m(n) represents the spectral signature of the nth endmember. The matrix S corresponds to the fractional abundance of each endmember, with a scaling factor g, and an abundance vector represented by a = { a₁, a₂, …, aₚ } ᵀ. Noise accounts for additive random disturbances in the remote sensing image data (Chen et al. 2018 ). There are two assumptions about the endmembers: nonnegativity and sum-to-one. Mixing matrix M and abundance matrix S are required for nonnegativity (a ≥ 0), and the sum-to-one constraint is used for the pixel fraction. The sum of all endmembers in a single pixel is one, i.e., ∑ q =1 p s ( q ) = 1. The simplex is represented as S x = { x ∈ R L : x = Mα , α ≥ 0, ∑ q =1 p s ( q ) = 1.} when Noise is zero, and the simplex shape is like a convex cone Cp = { r ∈R L : r = MS , α ≥ 0, ∑ q =1 p s ( q ) = 1, γ ≥ 0} (Nascimento and Dias 2005 ). As illustrated in Fig. 4 , the algorithm operates within the simplex Sx (where γ = 1 ), defined by a convex combination of spectral signatures. In the first iteration, the data points are projected in the direction f 1 , which emphasizes the contribution of the first endmember m a . The resulting projection identifies m a as the extreme point in that direction. In the second iteration, the data are re-projected along a new orthogonal direction, f 2 , allowing the algorithm to extract the second endmember m b . This iterative procedure continues, projecting the residual data in successively orthogonal directions until no additional distinct endmembers can be found. 3.2. Box-counting Fractal Method Since its introduction in the 1970s, fractal theory has experienced widespread application and continuous innovation within the quantitative analysis domain of Earth sciences (Qiu et al. 2015 ; Lyu et al. 2017 ; Maghsoudi et al. 2021). Alteration anomalies derived from remote sensing imagery typically exhibit complex and irregular spatial patterns, rendering traditional Euclidean geometry methods inadequate for accurately characterizing their spatial distribution. However, these alteration anomalies often display pronounced fractal self-similarity as an intrinsic feature. Therefore, fractal theory can effectively quantify the spatial self-similarity of alteration anomalies by calculating their fractal dimensions, providing a robust measure for characterizing the complexity of their spatial distributions. Several approaches have been developed for calculating fractal dimensions, including box-counting dimension (capacity dimension) (Ni et al. 2017 ; Dinç et al. 2023), information dimension (Sun et al. 2024 ; Shi and Xiao 2025), similarity dimension (Zhao et al. 2023 ; Cheng et al. 2024 ), correlation dimension (Liu et al. 2023 ; Liu et al. 2024 ), and Hausdorff dimension (Fernández-Martínez and Sánchez-Granero 2014 ; Hummer 2017). Among these, the box-counting dimension method has become the most widely adopted approach due to its straightforward theoretical foundation and computational simplicity. The implementation of the box-counting method involves overlaying the study area with a two-dimensional square grid of varying box sizes (side length = r ) and counting the number of boxes, N(r) , that contain alteration anomaly information at each scale. The relationship between the number of boxes N(r) and box size r follows a power-law distribution, as expressed by Eq.: N(r) = Cr − D Where C is a constant, and D is the box dimension subdimensional value. Take the logarithm of each side of the above equation: lnN(r)=-Dlnr + lnc According to the above formula, the log r-log N(r) curve is plotted and fitted by the least squares method. The slope of the resulting straight line represents the box dimension subdimension D . The correlation coefficient R 2 is also obtained, and the closer the value of R 2 is to 1, the better the fit is, and the better the consistency between the data and the power law model. In this study, the fractal dimension of remote sensing-derived alteration anomalies in the study area was calculated using the box-counting method described above, and corresponding fractal dimension contour maps were subsequently generated. The specific procedures are as follows: first, an initial observational scale ( r ) was defined, and two-dimensional orthogonal grids with side lengths of r = r₀, r₀/2, r₀/4 , and r₀/8 were used to overlay the study area (Fig. 5 ) sequentially. For each scale, the number of grid cells N(r) containing alteration anomaly information was recorded. Next, a log-log plot of r versus N(r) was constructed, and the box dimension ( D ), along with its coefficient of determination ( R² ), was obtained through least-squares linear regression. This approach enabled a quantitative characterization and evaluation of the spatial distribution of remote sensing alteration anomalies. 4. Results and Discussion 4.1. Spectral Mixture Analysis Using ENVI software, the Sentinel-2 imagery underwent a linear transformation, after which the SMACC method was applied to extract endmember spectra and abundance information. The processed imagery was decomposed into abundance images corresponding to nine distinct land-cover endmembers (Fig. 6 ) and one shadow image, with the total abundance summing to one. Additionally, spectral vectors were generated for each of the nine endmember components (Table 1 ). In the experiment, these nine endmember spectral vectors were compared and identified against standard spectra from a reference spectral library. By combining this spectral comparison with their spatial distribution patterns in the abundance images, endmember component 3 was identified as vegetation, while endmember components 4 and 6 were classified as other interference features. Table 1 Spectral vectors of nine endmember waveforms from remotely sensed images of the study area. Band2 Band 3 Band 4 Band 5 Band 6 Band 7 Band 8a Band 11 Band 12 1 4912 5948 6560 6675 6676 6734 6808 7893 6424 2 6512 6712 6640 5651 5453 5157 4789 2267 2162 3 3444 2256 2268 3435 3367 3691 4311 6460 6342 4 1675 2190 2122 2777 4259 4604 4920 4181 3007 5 6572 6916 6936 4633 4448 4396 4204 2853 2534 6 3856 4012 4372 3648 3618 4031 4278 5825 6136 7 2440 3432 4272 4548 4702 4913 4987 7146 6375 8 4524 5092 5044 5780 5638 5435 4974 2311 1975 9 2144 2636 2960 3083 3133 3325 3489 6116 5621 After determining the reflectance and abundance values of vegetation and interference endmembers, spectral unmixing was performed on each pixel band using a linear spectral unmixing model. Using Band 2 as an example, the abundance values of the identified interference endmember components (F₃, F₄, F₆) were multiplied by their corresponding reflectance values in Band 2. These products were then summed to create a composite image that represents vegetation and other interference endmembers. Next, the summed interference reflectance was subtracted from the original digital number (DN₂) of Band 2, resulting in an image that reflects only the non-vegetation components. This non-vegetation reflectance image was divided by its fractional abundance (1 − F₃ − F₄ − F₆), resulting in a compensated reflectance image that more accurately represents rock and soil endmembers by effectively removing interference contributions. Ultimately, spectral reconstruction for rock and soil pixels across all image bands in the study area was achieved through the reconstruction calculation (Table 2 ). Following spectral mixture analysis, rocky terrain in sparsely vegetated mountainous areas became distinctly apparent. Additionally, in densely vegetated regions, the vegetation signal was effectively suppressed without excessive masking, thereby preserving alteration information on rock surfaces and enhancing the reliability of subsequent alteration anomaly extraction in this study. Table 2 Calculation formula for reconstructed images of the study area. Band Calculation formula 2 (DN1-F3×3444-F4×1675-F6×3856)/༈1-F3-F4-F6༉ 3 (DN2-F3×2256-F4×2190-F6×4012)/༈1-F3-F4-F6༉ 4 (DN3-F3×2268-F4×2122-F6×4372)/༈1-F3-F4-F6༉ 5 (DN4-F3×3435-F4×2777-F6×3648)/༈1-F3-F4-F6༉ 6 (DN5-F3×3367-F4×4259-F6×3618)/༈1-F3-F4-F6༉ 7 (DN6-F3×3691-F4×4604-F6×4031)/༈1-F3-F4-F6༉ 8a (DN7-F3×4311-F4×4920-F6×4278)/༈1-F3-F4-F6༉ 11 (DN8-F3×6460-F4×4181-F6×5825)/༈1-F3-F4-F6༉ 12 (DN9-F3×6342-F4×3007-F6×6136)/༈1-F3-F4-F6༉ 4.2. Remote Sensing Alteration Anomaly Information Extraction In geological prospecting, the presence of altered rocks in a region does not necessarily mean that mineral deposits exist. Significant and widely distributed wall-rock alterations often occur in large to super-large ore deposits. Unaltered and altered rocks exhibit significant differences in mineral composition and lithology, resulting in distinct spectral reflectance characteristics. Therefore, differences in spectral properties serve as the theoretical basis for extracting alteration information through remote sensing. The Saga region exhibits various mineralogical alteration features associated with ore deposits, including malachitization, epidotization, chloritization, sericitization, hematitization, and limonitization. Minerals characteristic of such wall-rock alterations typically contain abundant Fe³⁺, Fe²⁺, OH⁻, or CO₃²⁻ groups. The electron vibration processes of these structural ions yield diagnostic spectral features in minerals enriched with these ions or groups, which are critical indicators in mineral exploration. In this study, diagnostic spectral features of typical alteration minerals were analyzed using reference spectral curves from the USGS spectral library to extract relevant alteration anomaly information. Representative spectral curves of these typical alteration minerals are presented in Fig. 7 . According to previous studies, PCA and band ratio methods are commonly utilized to extract remote sensing alteration information. In this study, PCA was applied to identify alteration anomalies associated with iron-staining and hydroxyl-bearing minerals (Al–OH and Mg–OH), based on their characteristic spectral absorption features (Table 3 ). Additionally, the band ratio method was employed to detect carbonate alteration anomalies by exploiting the distinct reflectance characteristics of carbonate minerals in Bands 7 and 12. Iron-staining minerals (such as hematite, limonite, and goethite) exhibit strong reflectance features in the wavelength ranges of 0.48–0.51 µm and 0.85–0.89 µm, corresponding to Band 2 and Band 8a of Sentinel-2 imagery. These minerals also display a characteristic absorption feature between 1.52–1.70 µm, corresponding to Sentinel-2 Band 11. Therefore, Bands 2, 4, 8a, and 11 were selected as an optimal combination for extracting iron-staining alteration anomalies. In principal component analysis, iron-staining anomalies are typically characterized by negative and identical contribution coefficient signs for Bands 2 and 8a, with an opposite coefficient sign for Band 11. Examination of the transformed eigenvectors indicated that the anomaly information related to iron-staining was primarily concentrated in the PC3 (Table 3 ). Consequently, PC3 was chosen as the principal component for the extraction of iron-staining anomalies. Table 3 Calculation formula for reconstructed images of the study area. Type Eigenvector Band 2 Band 4 Band 8A Band 11 Iron- staining PC1 0.313681 0.478547 0.517353 0.636351 PC2 0.617418 0.340468 0.160370 -0.690767 PC3 -0.568580 0.785637 -0.181369 -0.163086 PC4 -0.443977 -0.194549 0.820812 -0.302163 Type Eigenvector Band 6 Band 8A Band 11 Band 12 Al-OH PC1 -0.480088 -0.506195 -0.516806 -0.49618 PC2 0.468644 0.504798 -0.674377 -0.26602 PC3 -0.419601 0.146771 -0.480054 0.756269 PC4 0.611411 -0.68367 -0.218348 0.333311 Type Eigenvector Band 2 Band 8A Band 11 Band 12 Mg-OH PC1 -0.410902 -0.526081 -0.537109 -0.515667 PC2 0.571009 0.46045 -0.619561 -0.279426 PC3 0.543775 -0.362285 0.491561 -0.575698 PC4 0.457622 -0.61642 -0.293304 0.56972 OH groups alteration minerals primarily include Al–OH minerals (such as kaolinite, muscovite, and alunite) and Mg–OH minerals (such as chlorite, epidote). Al–OH minerals exhibit reflective spectral features in the wavelength range of 1.60–1.70 µm, corresponding to Sentinel-2 Band 11, and absorption features near 2.20 µm, corresponding to Sentinel-2 Band 12. Consequently, Bands 6, 8a, 11, and 12 were selected as the optimal band combination for identifying Al–OH mineral alteration anomalies. PCA was then performed using these selected bands (Bands 6, 8a, 11, and 12). The principal component related to Al–OH anomalies shows positive and identical contribution coefficients for Bands 8a and 11, while Bands 11 and 12 exhibit opposite signs. Examination of the transformed eigenvectors indicated that the inverse of principal component 4 (–PC4) best matched these criteria (Table 3 ). Therefore, –PC4 was utilized as the principal component for extracting Al–OH mineral alteration anomalies in this study. Mg–OH minerals exhibit an increasing reflectance trend in the wavelength range of 0.70–1.82 µm, corresponding to Bands 5–11 of Sentinel-2 imagery, and display characteristic absorption near 2.32 µm, corresponding to Sentinel-2 Band 12. Accordingly, Bands 2, 8a, 11, and 12 were selected as the optimal band combination for extracting Mg–OH mineral alteration anomalies. In the PCA, the principal component associated with Mg–OH anomalies typically shows positive and identical coefficient signs for Bands 2 and 11 and opposite signs for Bands 11 and 12. Based on these criteria, eigenvector PC3 was selected as the principal component for identifying Mg–OH mineral alteration anomalies. Carbonate minerals (such as calcite and dolomite) typically exhibit prominent reflectance peaks around 0.7–0.8 µm and 1.7–1.8 µm, corresponding to Bands 7 and 12 of Sentinel-2 imagery, respectively. Based on multiple experimental analyses, a hybrid band ratio method using Band 12/Band 8a and Band 7/Band 4 was adopted in this study. This mixed-ratio approach enhances the diagnostic spectral signatures of carbonate minerals, thereby effectively extracting associated alteration anomaly information. After the characteristic principal components or hybrid bandratio images representing the four alteration types had been extracted, we calculated the mean (µ) and standard deviation (σ) of the alteration-type images. Pixels with values greater than µ + Nσ were classified as remotesensing alteration anomalies. The multiplier N was set to 3 for ironstaining and Al–OH and Mg–OH anomalies and to 2 for carbonate anomalies. The final Sentinel-2 alteration anomaly map is shown in Fig. 8 . 4.3. Characterization of Alteration Anomalies in Remote Sensing Using Fractal Dimensions The fractal dimension provides a useful indicator for locating ore deposits. In this study, box-counting fractal analysis was applied to four alteration-anomaly layers derived from Sentinel-2 imagery: iron staining, Al–OH, Mg–OH, and carbonate. This analysis aimed to pinpoint the most promising mineralization sites in the study area and provide theoretical guidance for future exploration efforts. Four two-dimensional orthogonal grids with cell sizes of 6.123 km, 3.062 km, 1.531 km, and 0.765 km were successively superimposed on the study area (Fig. 5 ). For each grid scale, the number of square cells N(r) containing the respective alteration anomaly was recorded. The natural logarithm of cell size, ln r, was plotted on the xaxis and ln N(r) on the yaxis; leastsquares regression lines were then fitted separately for each alteration type. The slope of each regression line yields the boxcounting dimension D . The statistical results are summarized in Table 4 , and the ln N(r) – ln r regression plots are shown in Fig. 9 . Table 4 Calculation formula for reconstructed images of the study area. Type fractal dimension Type fractal dimension r /km N ( r ) ln r Ln N ( r ) r /km N ( r ) ln r Ln N ( r ) Iron- staining 6.123 12 5.77455 3.46574 Al-OH 6.123 11 5.42935 3.46574 3.062 35 4.74493 4.15888 3.062 33 4.60517 4.15888 1.531 115 3.55535 4.85203 1.531 100 3.49651 4.85203 0.765 322 2.48491 5.54518 0.765 228 2.3979 5.54518 Mg-OH 6.123 12 5.54908 3.46574 Carbonate 6.123 10 4.61512 3.46574 3.062 34 4.7362 4.15888 3.062 24 4.00733 4.15888 1.531 114 3.52636 4.85203 1.531 55 3.17805 4.85203 0.765 257 2.48491 5.54518 0.765 101 2.30259 5.54518 Using the fractal method described above, fractal dimensions for each alteration type were calculated over the scale range of 0.765–6.123 km. The results are as follows: the ironstaining anomalies have an overall fractal dimension of 1.59541 with R² = 0.9968; the Al–OH anomalies, 1.47198 with R² = 0.9921; the Mg–OH anomalies, 1.50074 with R² = 0.9916; and the carbonate anomalies, 1.12053 with R² = 0.9925. The highest fractal dimension value of the iron-stained alteration in the study area indicates significant spatial structural complexity. This complexity may be closely related to mineralization centers formed by the oxidation of metal sulfides or the intersection of fracture sites. It is presumed to be a result of the main metallogenic period. Therefore, areas with high-D zones should be prioritized as core targets for exploration. The similar fractal dimensions of the Al–OH and Mg–OH anomalies imply comparable spatial complexities, likely reflecting different stages of the same mineralizing system; highD areas delineate the probable boundaries of this system. By contrast, the lower fractal dimension of the carbonate anomalies points to a more uniform spatial distribution characteristic of peripheral or late-stage alteration zones. All alteration types exhibit R² > 0.99, confirming pronounced statistical self-similarity within the selected scale domain. The fractal dimension statistics of the previously described division grid were calculated to determine the fractal dimension value of the etching anomalies within each grid range. Each grid was assigned a fractal dimension value based on its center. Using Surfer software, the Kriging interpolation method was then employed to create contour maps depicting the various alteration anomalies. The resulting contour map illustrates the fractal dimension information for different alteration anomalies, as shown in Fig. 10 . The fractal dimension contour maps for the various alteration types show that ironstaining anomalies are markedly stronger, in both magnitude and areal extent, than the other alteration types, whereas the highvalue zones for carbonate alteration are the least extensive and least intense. High fractaldimension values for ironstaining, Al–OH, and Mg–OH anomalies are concentrated principally within the Silurian–Ordovician Dajiling Formation and the Jurassic Zongzhuo Formation. The Dajiling Formation is dominated by quartz schist, quartz sandstone, and muscovite-bearing quartzite, whereas the Zongzhuo Formation consists mainly of argillaceous siltstone, shale, and limestone. These strata are relatively unconsolidated, highly cracked, and rich in minerals that are readily altered, thereby providing favorable pathways and reactive sites for hydrothermal fluids. The lithologic and structural characteristics likely control the enrichment of the alteration anomalies in these units. Additionally, the anomalies of iron-staining, Al–OH, and Mg–OH show a strong spatial relationship with most known sites of Cu and Fe mineralization; the deposits are generally found within their respective high-value zones. This coherence reflects intense hydrothermal activity and robust mineralization in the area. Accordingly, both the Dajiling and Zongzhuo formations not only provide favorable conditions for alteration development but also represent the most prospective stratigraphic units for exploration. By contrast, carbonate-related alteration anomalies display high values only in limited portions of the Zongzhuo Formation and exhibit a comparatively weak spatial match with known mineralization points. This suggests that their predictive significance is limited and that they may indicate mineral potential only under specific structural settings or lithologic combinations. Overall, the fractal characteristics of the different alteration anomalies reveal their spatial complexity and genetic mechanisms. When combined with the spatial coupling between each alteration type and existing mineralization sites, these features constitute a valuable basis for regional prospecting, offering scientific support for identifying potential mineralization targets. To integrate the fractaldimension contour maps of the various alteration types in the Saga area, we normalized each dataset to prevent the spatial characteristics of any one alteration type from being masked by large absolute values during a simple arithmetic overlay. This normalization effectively eliminated the problem of subdued anomaly zones in the composite layer that would otherwise arise from disproportionately high values in a single alteration map. The resulting composite fractaldimension contour map shows that high-value zones are concentrated mainly within the Silurian–Ordovician Dajiling Formation and the Jurassic Zongzhuo Formation, exhibiting a conspicuous concentricdiffusion pattern. Based on the distribution of previously identified mineralization sites, areas with composite fractal values exceeding 0.7 were designated as high-potential mineralization zones, and two additional prospective targets were delineated outside the known mineralization sites (Fig. 11 ). 4.4. Results of Field Survey and Laboratory Analysis Field Survey is a critical approach for validating indoor remote sensing interpretation results. Based on the preceding findings, a geological traverse survey was conducted in Mineralization Potential Zone I. The stratigraphy of this zone is primarily composed of the Dajiling Formation. During the field investigation, several well-defined quartz vein outcrops were observed near high-value anomaly zones. Some of these quartz veins exhibited evident signs of malachite staining and chalcopyrite mineralization. Quartz vein samples were collected, prepared into thin sections, and examined under a polarizing microscope. The petrographic analysis revealed multiple types of mineralization within the quartz veins, including Cu mineralization (Fig. 12 a), malachitization (Fig. 12 b), sericitization (Fig. 12 c), pyritization (Fig. 12 d), limonitization (Fig. 12 e), and galena mineralization (Fig. 12 f). In terms of structural features, the rocks predominantly exhibit hydrothermal infill and replacement textures, including brecciated, vein-like, stockwork, fine vein-disseminated, and massive textures. Among these, fine vein-disseminated, densely disseminated, and brecciated textures are the most common. The fine and densely disseminated textures occur mainly within stratiform ore bodies, whereas brecciated textures are typically associated with vein-type deposits. Furthermore, in schistositized marble, disseminated limonitization, chloritization, and spotty limonitization are observed, with localized limonitization forming banded patterns along the schistositized planes. The above field observations corroborate our earlier analysis of the lithological characteristics of the Dajiling Formation. The rocks of this formation are generally loosely consolidated, extensively cracked, and rich in minerals susceptible to alteration, thereby providing favorable pathways and reactive sites for hydrothermal fluid infiltration, migration, and mineralization. These conditions have promoted the enrichment of the alteration anomalies observed. This finding not only validates the reliability of the methods applied in this study but also furnishes critical evidence for the verification of additional mineralization potential zones. Future work will involve groundtruthing another prospective zone within the study area. 5. Conclusions (1)Introducing spectral mixture analysis into the study area enables the separation of reflectance and abundance values for the various interference components contained in mixed pixels. Spectral unmixing and endmember reconstruction enhance the detectability of subtle mineralization-related alteration anomalies in the remotesensing imagery. By accurately quantifying the spectral contribution of each surface feature, this approach prevents the loss of alteration signatures that often occurs with conventional masking techniques due to over-correction, and it markedly improves both the accuracy and completeness of alteration information extraction under diverse surfacecover conditions. (2)The fractal dimensions of the various remotesensing alteration anomalies were determined using fractal analysis, thereby quantitatively characterizing their spatial self-similarity. The ironstaining anomalies have a fractal dimension of 1.59541 with R² = 0.9968; the Al–OH anomalies, 1.47198 with R² = 0.9921; the Mg–OH anomalies, 1.50074 with R² = 0.9916; and the carbonate anomalies, 1.12053 with R² = 0.9925. Notably, all alteration types exhibit R² values greater than 0.99, confirming that the alteration anomalies in the study area display pronounced statistical self-similarity within the selected scale range. (3)The highpotential mineralization zones delineated in the fractaldimension contour map of alteration anomalies show a strong spatial correspondence with known mineralization sites. In addition, two new prospective zones were identified on the periphery of the existing mineralization sites. These findings furnish a sound theoretical basis and clear exploration priorities for the next stage of prospecting in the study area. Declarations Author contributions Conceptualization: K.Q.; methodology: K.Q.; software: K.Q. and S.D.; validation: K.Q., and J.K.; formal analysis: K.Q. and T.L.; investigation: K.Q.; L.Q.; and Z.R.; resources: Z.R. and S.G.; data curation: K.Q.; writing—original draft preparation: J.K.; writing—review and editing: Y.L. and S.D.; visualization: K.Q.; L.Q.; and S.D.; supervision: T.L. and Y.H.; project administration: T.L.; funding acquisition: Y.H. Funding This work was supported by the China Geological Survey Project (DD20240014, DD20243088), the Ministry of Natural Resources' new round of scientific and technological support project for finding mineral breakthroughs (ZKKJ202427-03), Tibet Science and Technology Program Projects (XZ202401YD0006-07). Data Availability The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request. Competing interests The authors declare no competing interests. References Abdelkader, M. A. et al. Advanced exploration of rare metal mineralization through integrated remote sensing and geophysical analysis of structurally-controlled hydrothermal alterations. J. Geochem. Explor. 267 , 107598 (2024). Abedini, M., Ziaii, M., Timkin, T. & Pour, A. B. Machine Learning (ML)-Based Copper Mineralization Prospectivity Mapping (MPM) Using Mining Geochemistry Method and Remote Sensing Satellite Data. Remote Sens. 15 (15), 3708 (2023). Amani, S., Shafizadeh-Moghadam, H. & Morid, S. Integrating sentinel-2 and sentinel-3 for actual evapotranspiration estimation across diverse climate zones using the sen-ET plugin and machine learning models. Earth Sci. Inf. 18 (4), 338 (2025). Asadzadeh, S., Chabrillat, S., Cudahy, T., Rashidi, B. & de Souza Filho, C. R. Alteration Mineral Mapping of the Shadan Porphyry Cu-Au Deposit (Iran) Using Airborne Imaging Spectroscopic Data: Implications for Exploration Drilling. Econ. Geol. 119 (1), 139–160 (2024). Bai, L. Y. et al. Predictive Prospecting Using Remote Sensing in a Mountainous Terrestrial Volcanic Area, in Western Bangongco–Nujiang Mineralization Belt, Tibet. Remote Sens. 15 (19), 4851 (2023). Baid, S. et al. Integrating geochemical insights and remote sensing for enhanced identification of hydrothermal alterations in the Igoudrane region, Anti-Atlas, Morocco. J. Afr. Earth Sc. 218 , 105368 (2024). Bhakthan, S. M. & Loganathan, A. A hyperspectral unmixing model using convolutional vision transformer. Earth Sci. Inf. 17 (3), 2255–2273 (2024). Boumahdi, M., García-Pedrero, A., Lillo-Saavedra, M. & Gonzalo-Martin, C. A new benchmark for spatiotemporal fusion of Sentinel-2 and Sentinel-3 OLCI images. Earth Sci. Inf. 18 (2), 349 (2025). Carvalho, M., Cardoso-Fernandes, J., González, F. J. & Teodoro, A. C. Comparative Performance of Sentinel-2 and Landsat-9 DataforRawMaterials’ Exploration OnshoreandinCoastal Areas. Remote Sens. 17 (2), 305 (2025). Chen, J. et al. Using geochemical data for prospecting target areas by the sequential maximum angle convex cone method in the Manzhouli area, China. Geochem. J. 52 (1), 13–27 (2018). Chen, Q. et al. ASTER and GF-5 Satellite Data for Mapping Hydrothermal Alteration Minerals in the Longtoushan Pb-Zn Deposit, SW China. Remote Sens. 14 (5), 1253 (2022). Chen, J. et al. Metallogenic prediction based on fractal theory and machine learning in Duobaoshan Area, Heilongjiang Province. (2024). Ore Geol Rev 106030. Chen, Q. et al. Remote sensing identification of hydrothermal alteration minerals in the Duobuza porphyry copper mining area in Tibet using WorldView-3 and GF-5 data: The impact of spatial and spectral resolution. Ore Geol. Rev. 180 , 106573 (2025). Cheng, Q. M. The perimeter-area fractal model and its application to geology. Math. Geol. 27 , 69–82 (1995). Cheng, L. et al. Self-similarity study based on the particle sizes of coal-series diatomite. Sci. Rep. 14 (1), 7502 (2024). Corrêa da Costa, M. A., Carneiro Naleto, J. L., Perrotta, M. M., Soares Monteiro, L. V. & de Souza Filho, C. R. Visible-Near Infrared–Short-Wave Infrared Spectroscopy and Mineral Mapping of Hydrothermal Alteration Zones at the Brejuí W-Mo Skarn Deposit, Seridó Mobile Belt, Borborema Province, Brazil. Econ. Geol. 119 (5), 1171–1198 (2024). Crowley, J. K. Near-infrared reflectance of zunyite; implications for field mapping and remote-sensing detection of hydrothermally altered high alumina rocks. Econ. Geol. 79 (3), 553–557 (1984). Ding, H. F. et al. Research on Scale Improvement of Geochemical Exploration Based on Remote Sensing Image Fusion. Remote Sens 15(8): 1993 (2023). Dinç Göğüş, Ö., Avşar, E., Develi, K. & Çalık, A. Quantifying the Rock Damage Intensity Controlled by Mineral Compositions: Insights from Fractal Analyses. Fractal Fract. 7 (5), 383 (2023). Dong, X. F., Gan, F. P., Li, N., Zhang, S. F. & Li, T. T. Mineral mapping in the Duolong porphyry and epithermal ore district, Tibet, using the Gaofen-5 satellite hyperspectral remote sensing data. Ore Geol. Rev. 151 , 105222 (2022). Fernández-Martínez, M. & Sánchez-Granero, M. A. Fractal dimension for fractal structures: A Hausdorff approach revisited. J. Math. Anal. Appl. 409 (1), 321–330 (2014). Filippi, A. M. et al. Comparison of Image Endmember- and Object-Based Classification of Very-High-Spatial-Resolution Unmanned Aircraft System (UAS) Narrow-Band Images for Mapping Riparian Forests and Other Land Covers. Land 11 (2), 246 (2022). Forouzan, M. & Arfania, R. Integration of the bands of ASTER, OLI, MSI remote sensing sensors for detection of hydrothermal alterations in southwestern area of the Ardestan, Isfahan Province, Central Iran. Egypt. J. Remote Sens. Space Sci. 23 (2), 145–157 (2020). Fu, Y. F., Cheng, Q. M., Jing, L. H., Ye, B. & Fu, H. Z. Mineral Prospectivity Mapping of Porphyry Copper Deposits Based on Remote Sensing Imagery and Geochemical Data in the Duolong Ore District, Tibet. Remote Sens. 15 (2), 439 (2023). Ghoneim, S. M., Ali, H. F., Abdelrahman, K., Fnais, M. S. & Lala, A. M. S. Integrating remote sensing and geophysical data for mapping potential gold mineralization localities at Abu Marawat area, central Eastern Desert, Egypt. Sci. Rep. 14 (1), 18273 (2024). Habashi, J., Oskouei, M. M., Moghadam, H. J. & Pour, A. B. Optimizing alteration mineral detection: A fusion of multispectral and hyperspectral remote sensing techniques in the Sar-e-Chah-e Shur, Iran. Remote Sensing Applications: Society and Environment 35: 101249. (2024). Hamedianfar, A. et al. Leveraging High-Resolution Long-Wave Infrared Hyperspectral Laboratory Imaging Data for Mineral Identification Using Machine Learning Methods. Remote Sens. 15 (19), 4806 (2023). Hegab, M. A. E. R. A multi-disciplinary approach for uranium exploration using remote sensing and airborne gamma-ray spectrometry data in the Gebel Duwi area, Central Eastern Desert, Egypt. Sci. Rep. 14 (1), 19739 (2024). Hosseinjani, M. & Tangestani, M. H. Mapping alteration minerals using sub-pixel unmixing of ASTER data in the Sarduiyeh area, SE Kerman, Iran. Int. J. Digit. Earth . 4 (6), 487–504 (2011). Huang, M. L. et al. Linking Porphyry Cu Formation to Tectonic Change in Postsubduction Settings: A Case Study from the Giant Yulong Belt, Eastern Tibet. Econ. Geol. 119 (2), 279–304 (2024). Hummer, D. R. Fractal distribution of mineral species among the crystallographic point groups. Am. Mineral. 106 (10), 1574–1579 (2021). Hussain, M., Li, H., Beshr, A. M. & Memon, F. A. Remote sensing-based structural and lithological mapping for prospecting polymetallic mineralization at Xiaoshan region, China. Earth Sci. Inf. 18 (1), 82 (2025). Islam, N. U. et al. Mineralogical mapping and lithological discrimination by using ASTER remote sensing data in the Chitral region, Khyber Pakhtunkhwa, Northern Pakistan. Earth Sci. Inf. 17 (6), 6075–6094 (2024). Khashaba, S. M. A., El-Shibiny, N. H., Hassan, S. M., Takazawa, E. & Khedr, M. Z. Application of remote sensing data integration in detecting mineralized granitic zones: A case study of the Gabal Al-Ijlah Al-Hamra, Central Eastern Desert, Egypt. J. Afr. Earth Sc. 200 , 104855 (2023). Kong, F. C. & Ding, G. Y. The implications the fractal dimension values of lineaments. Earthquake 5 , 33–37 (1991). Lin, B. et al. Multipulsed Magmatism and Duration of the Hydrothermal System of the Giant Jiama Porphyry Cu System, Tibet, China. Econ. Geol. 119 (1), 201–217 (2024). Liu, Z. H. et al. Fractal and Multifractal Characteristics of Lineaments in the Qianhe Graben and Its Tectonic Significance Using Remote Sensing Images. Remote Sens. 13 (4), 587 (2021). Liu, H. et al. Fractal dimension used as a proxy to understand the spatial distribution for carlin-type gold deposits. Ore Geol. Rev. 158 , 105534 (2023). Liu, Y. et al. Fractal-Based Pattern Quantification of Mineral Grains: A Case Study of Yichun Rare-Metal Granite. Fractal Fract. 8 (1), 49 (2024). Lyu, C., Cheng, Q. M., Zuo, R. G. & Wang, X. P. Mapping Spatial Distribution Characteristics of Lineaments Extracted from Remote Sensing Image Using Fractal and Multifractal Models. J. Earth Sci. 28 , 507–515 (2017). Maghsoudi Moud, F., van Ruitenbeek, F., Hewson, R. & van der Meijde, M. An Approach to Accuracy Assessment of ASTER Derived Mineral Maps. Remote Sens. 13 (13), 2499 (2021). Mandelbrot, B. B. The Fractal Geometry of Nature pp. 17–179 (WH Freeman, 1983). Mbianya, G. N. et al. Remote sensing satellite-based structural/alteration mapping for gold exploration in the Ketté goldfield, Eastern Cameroon. J. Afr. Earth Sc. 184 , 104386 (2021). Miao, L. D., Qi, H. R. & Szu, H. A maximum entropy approach to unsupervised mixed-pixel decomposition. IEEE Trans. Image Process. 16 (4), 1008–1021 (2007). Nabatian, G., Songjian, A., Pour, A. B., Abdollahi, F. & Habashi, J. Integration of ASTER imagery and field data for chromite exploration in the Eastern Khoy Ophiolite Complex, NW Iran. Earth Sci. Inf. 18 (2), 245 (2025). Naceur, M. S., Loghmari, M. A. & Boussema, M. R. The contribution of the sources separation method in the decomposition of mixed pixels. IEEE Trans. Geosci. Remote Sens. 42 (11), 2642–2653 (2004). Nascimento, J. M. P. & Dias, J. M. B. Vertex component analysis: a fast algorithm to unmix hyper-spectral data. IEEE Trans. Geosci. Rem. Sens. 43 (4), 898–910 (2005). Ni, C. Z., Zhang, S. T., Chen, Z., Yan, Y. F. & Li, Y. J. Mapping the spatial distribution and characteristics of lineaments using fractal and multifractal models: A case study from Northeastern Yunnan Province, China. Sci. Rep. 7 (1), 10511 (2017). Qiu, J. T., Zhang, C. & Hu, X. Integration of Concentration-Area Fractal Modeling and Spectral Angle Mapper for Ferric Iron Alteration Mapping and Uranium Exploration in the Xiemisitan Area, NW China. Remote Sens. 7 (10), 13878–13894 (2015). Ren, Z. L., Zhai, Q. P. & Sun, L. A Novel Method for Hyperspectral Mineral Mapping Based on Clustering-Matching and Nonnegative Matrix Factorization. Remote Sens. 14 (4), 1042 (2022). Sharifi, A. & Hosseingholizadeh, M. H. Validation of extracted endmembers from hyperspectral images. Int. Archives Photogrammetry Remote Sens. Spat. Inform. Sci. 42 , 989–992 (2019). Shi, R. & Xiao, H. G. Particle Flow Simulation of the Mechanical Properties and Fracture Behavior of Multi-Mineral Rock Models with Different Fractal Dimensions. Fractal Fract. 9 (1), 13 (2024). Sikakwe, G. U. Mineral exploration employing drones, contemporary geological satellite remote sensing and geographical information system (GIS) procedures: A review. Remote Sens. Applications: Soc. Environ. 31 , 100988 (2023). Sun, T. et al. Fractal and multifractal characteristics of regional fractures in Tongling metallogenic Area. Nonferrous Met. Eng. 8 , 111–115 (2018). Sun, T. et al. Fractal-Based Multi-Criteria Feature Selection to Enhance Predictive Capability of AI-Driven Mineral Prospectivity Mapping. Fractal Fract. 8 (4), 224 (2024). Thompson, D. R., Mandrake, L., Gilmore, M. S. & Castano, R. Superpixel endmember detection. IEEE Trans. Geosci. Remote Sens. 48 (11), 4023–4033 (2010). Yang, C., Jia, H. K., Dong, L. F., Zhao, H. S. & Zhao, M. H. Selection of Landsat-8 Operational Land Imager (OLI) Optimal Band Combinations for Mapping Alteration Zones. Remote Sens. 16 (2), 392 (2024). Yang, S. H. & Tian, S. F. Extraction of Alteration Information from Hyperspectral Data Base on Kernel Extreme Learning Machine. Remote Sens. 16 (19), 3646 (2024). Yang, D., Shao, H., Chang, H., Zhang, R. & Ding, B. Landsat 8 and ASTER remote sensing data for ore prospecting in the postmasburg manganese ore field, South Africa. Earth Sci. Inf. 18 (1), 1–16 (2025). Yao, F. J. et al. Review on the development of multi- and hyperspectral remote sensing technology for exploration of copper-gold deposits. Ore Geol. Rev. 162 , 105732 (2023). Yuan, C. J., Zhao, J., Liu, Z. B., Liu, X. F. & Tang, J. Detecting mineralization via white mica spectral footprints: From field-based sampling to satellite hyperspectral remote sensing. Ore Geol. Rev. 181 , 106588 (2025). Zhang, H. C., Zhang, Y. N. & Duan, F. A robust and fast geometry based unmixing algorithm for hyperspectral imagery. Int. J. Remote Sens. 31 (20), 5481–5493 (2010). Zhang, G. et al. Comprehensive Multi-Source remote sensing data integration for enhanced mineralization alteration extraction and geological structure interpretation in the Lala region of Sichuan Province. Ore Geol. Rev. 168 , 106032 (2024). Zhao, H. Q. & Zhao, X. S. Nonlinear unmixing of minerals based on the log and continuum removal model. Eur. J. Remote Sens. 52 (1), 277–293 (2019). Zhao, F. W., Hu, J. H., Yang, Y. N., Xiao, H. X. & Ma, F. C. Cross-Scale Study on Lime Modified Phosphogypsum Cemented Backfill by Fractal Theory. Minerals 12 (4), 403 (2022). Zhao, Z., Qiao, K., Liu, Y. W., Chen, J. & Li, C. L. Geochemical Data Mining by Integrated Multivariate Component Data Analysis: The Heilongjiang Duobaoshan Area (China) Case Study. Minerals 12 (8), 1035 (2022). Zhao, Z. H. et al. Spatial Analysis of Structure and Metal Mineralization Based on Fractal Theory and Fry Analysis: A Case Study in Nenjiang – Heihe. Metallogenic Belt Minerals . 13 (3), 313 (2023). Zhao, Q., Yan, Y., Tonai, S., Dilek, Y. & Zhu, Z. F. Timing of India-Asia collision and significant coupling between them around 51 Ma: Insights from the activation history of the Zhongba-Gyangze thrust in southern Tibet. Geology 52 (1), 61–66 (2024). Zuo, R. G. & Wang, J. L. ArcFractal: An ArcGIS add-in for processing geoscience data using fractal/ multifractal models. Nat. Resour. Res. 29 , 3–12 (2020). Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6598339","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":454854117,"identity":"92cbee71-fcc7-4d8c-9b84-f7ae656f0bde","order_by":0,"name":"Kai Qiao","email":"","orcid":"","institution":"Civil-Military Integration Center of China Geological Survey","correspondingAuthor":false,"prefix":"","firstName":"Kai","middleName":"","lastName":"Qiao","suffix":""},{"id":454854118,"identity":"8e10ece9-e00d-4992-a9a1-9d379a97bc27","order_by":1,"name":"Tao Luo","email":"","orcid":"","institution":"Civil-Military Integration Center of China Geological Survey","correspondingAuthor":false,"prefix":"","firstName":"Tao","middleName":"","lastName":"Luo","suffix":""},{"id":454854126,"identity":"3b9a5804-08e2-42bf-9378-18443dfc8afa","order_by":2,"name":"Shihao Ding","email":"","orcid":"","institution":"Civil-Military Integration Center of China Geological Survey","correspondingAuthor":false,"prefix":"","firstName":"Shihao","middleName":"","lastName":"Ding","suffix":""},{"id":454854127,"identity":"c80372f8-3ae0-4c8c-a851-cc99fbaa0b2e","order_by":3,"name":"Licheng Quan","email":"","orcid":"","institution":"Civil-Military Integration Center of China Geological Survey","correspondingAuthor":false,"prefix":"","firstName":"Licheng","middleName":"","lastName":"Quan","suffix":""},{"id":454854129,"identity":"44dea651-f7e5-4662-ba63-ba14f8dc88b2","order_by":4,"name":"Jingui Kong","email":"","orcid":"","institution":"Shandong Provincial Nuclear Industry Geological Group 273","correspondingAuthor":false,"prefix":"","firstName":"Jingui","middleName":"","lastName":"Kong","suffix":""},{"id":454854130,"identity":"12422554-5a32-48ad-94cb-0fda8bdb613e","order_by":5,"name":"Yiwen Liu","email":"","orcid":"","institution":"Liaoning Datang International Shendong Thermal Power Co.,Ltd","correspondingAuthor":false,"prefix":"","firstName":"Yiwen","middleName":"","lastName":"Liu","suffix":""},{"id":454854131,"identity":"e6079638-566f-4afa-afd5-70f9f382fd15","order_by":6,"name":"Zhiwen Ren","email":"","orcid":"","institution":"Civil-Military Integration Center of China Geological Survey","correspondingAuthor":false,"prefix":"","firstName":"Zhiwen","middleName":"","lastName":"Ren","suffix":""},{"id":454854132,"identity":"8e7b921d-a844-457c-8ee8-091b2fbd3307","order_by":7,"name":"Shisong Gong","email":"","orcid":"","institution":"Civil-Military Integration Center of China Geological Survey","correspondingAuthor":false,"prefix":"","firstName":"Shisong","middleName":"","lastName":"Gong","suffix":""},{"id":454854133,"identity":"df6f4049-171e-416a-9f7b-114801b18d0c","order_by":8,"name":"Yong Huang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8klEQVRIiWNgGAWjYBACAwhlw8DAzNxAkpY0oBZG0rQcBmJitZhLJD97+LXtfDR/O1DLj191DPLuPQYMP3fg1mI5I83cWLbtdu6Mw4wNjL19hxkMz5wxYOw9g8dhNxLMpCWBWhqAWpgZew4wGM5IS2BmbMOnJf0bUMu53PkQLXXEaMkxk/zYdiB3A0gLww9mBnmJ5AP4tZx5UybNcC45dyNQy8HehsM8BjyHDxzsxaflePo2yR9ldrnzzh8++ODHnzo5+fbGxgc/8WgBAWZeNgjjANA9PAYHgAz8GoBx+OMPjAlkyDcQUj8KRsEoGAUjDQAA6p5WlE1B0iEAAAAASUVORK5CYII=","orcid":"","institution":"Civil-Military Integration Center of China Geological Survey","correspondingAuthor":true,"prefix":"","firstName":"Yong","middleName":"","lastName":"Huang","suffix":""}],"badges":[],"createdAt":"2025-05-06 02:53:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6598339/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6598339/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":82656161,"identity":"a4702022-11aa-4ec3-b02b-cc8cc0fa3b7d","added_by":"auto","created_at":"2025-05-13 18:51:16","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":430806,"visible":true,"origin":"","legend":"\u003cp\u003eTechnical workflow of the study area.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/655d81150c2071081a25d48e.png"},{"id":82655768,"identity":"44b5758d-0cf6-4d51-8719-86da490fa638","added_by":"auto","created_at":"2025-05-13 18:43:16","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1177887,"visible":true,"origin":"","legend":"\u003cp\u003eGeological map of the Saga region.\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/d194bcf67ad2b5318e46aaf2.jpeg"},{"id":82656640,"identity":"1a3e91a1-7812-43c6-94e7-a022c57bf298","added_by":"auto","created_at":"2025-05-13 18:59:16","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":960358,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of Sentinel-2 image mask before (a) and after (b) in the Saga region.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/cf7fba74d17e54ca4ed51e70.jpeg"},{"id":82656164,"identity":"946bf6dc-bdb2-460c-9aa6-27356c7faf11","added_by":"auto","created_at":"2025-05-13 18:51:16","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":192107,"visible":true,"origin":"","legend":"\u003cp\u003eModel diagram of the SMACC algorithm (modified from the literature [48.49]).\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/d907ea1911f195178d79226a.png"},{"id":82655765,"identity":"50e54668-975b-453a-8aa1-6e9665af4781","added_by":"auto","created_at":"2025-05-13 18:43:16","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":173143,"visible":true,"origin":"","legend":"\u003cp\u003eSchematic diagrams of fractal dimension scale division steps (\u003cem\u003er\u003c/em\u003e = 6.123km).\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/860ea93d8b19887d05beaa25.jpeg"},{"id":82655774,"identity":"04d36dc3-6e6c-472c-b595-d39ec1120d12","added_by":"auto","created_at":"2025-05-13 18:43:16","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":1853169,"visible":true,"origin":"","legend":"\u003cp\u003eShows abundant images of each end element after SMACC decomposition (a-j: end elements 1-9).\u003c/p\u003e","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/6f78cdd6095b0660cf60d93b.jpeg"},{"id":82656170,"identity":"10837408-377c-4ee0-8937-e36811f4b52d","added_by":"auto","created_at":"2025-05-13 18:51:16","extension":"jpeg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":487569,"visible":true,"origin":"","legend":"\u003cp\u003eSpectral profiles of iron-stained (a), Al-OH (b), Mg-OH (c), and carbonate (d) were obtained from the Saga area.\u003c/p\u003e","description":"","filename":"floatimage7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/25a299f5d0ffadbadd23a77d.jpeg"},{"id":82656167,"identity":"edc5fbff-1fd5-4789-875f-c016c79fbb3c","added_by":"auto","created_at":"2025-05-13 18:51:16","extension":"jpeg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":2447345,"visible":true,"origin":"","legend":"\u003cp\u003eInformation map of Iron-stained (a), Al-OH (b), Mg-OH (c), and carbonate (d) alteration anomalies in the Saga area.\u003c/p\u003e","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/64f3b97d2bcd8cae119864d5.jpeg"},{"id":82656641,"identity":"bdd4829c-cb5c-4e7f-81c3-9ef1d38b24fc","added_by":"auto","created_at":"2025-05-13 18:59:16","extension":"jpeg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":348296,"visible":true,"origin":"","legend":"\u003cp\u003eLinear plots of the fractal dimension \u003cem\u003elnN(r)-lnr\u003c/em\u003efit the alteration anomaly information for Iron-staining (a), Al-OH (b), Mg-OH (c), and carbonate (d) in the Saga area.\u003c/p\u003e","description":"","filename":"floatimage9.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/28ff2a3a99f717916c2d8e0c.jpeg"},{"id":82656165,"identity":"012a06f8-d186-4743-9fdf-2f2a56adaa34","added_by":"auto","created_at":"2025-05-13 18:51:16","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":1783910,"visible":true,"origin":"","legend":"\u003cp\u003eFractional dimensional contour maps of alteration anomalies for Iron-stained (a), Al-OH (b), Mg-OH (c), and carbonate (d) in the Saga area.\u003c/p\u003e","description":"","filename":"floatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/1c2d0a484adee8d51bd643a8.png"},{"id":82656172,"identity":"0f6ab192-7442-4d2b-ba3c-e1c59b544fa3","added_by":"auto","created_at":"2025-05-13 18:51:16","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":1677271,"visible":true,"origin":"","legend":"\u003cp\u003eFractional dimensional contour map of the alteration anomaly in the Saga area.\u003c/p\u003e","description":"","filename":"floatimage11.png","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/876cface473c107e690b3079.png"},{"id":82655776,"identity":"1af8856b-b24a-40df-92e8-4216e1fe922a","added_by":"auto","created_at":"2025-05-13 18:43:16","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":1802682,"visible":true,"origin":"","legend":"\u003cp\u003ePhotomicrographs of mineralization: (a) Chalcopyrite and pyrite in a quartz vein; (b) Malachite alteration in a quartz vein; (c) Sericite–quartz symbiosis in Sericite–quartz schist; (d, e) Pyrite, limonite, and rutile coexisting in a quartz vein; (f) Galena in a quartz vein. Ccp–chalcopyrite; Py–pyrite; Mlc–malachite; Qz–quartz; Ser–sericite; Rt –rutile; Lm–limonite; Gn–galena.\u003c/p\u003e","description":"","filename":"floatimage12.png","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/20096ff8e69e42ef218f9988.png"},{"id":84067729,"identity":"c5e57ce8-b831-446e-9856-d1ad43450c60","added_by":"auto","created_at":"2025-06-06 11:31:29","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":14440228,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6598339/v1/982206f2-7d94-414b-bd89-fef6088809e3.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Extraction of Remote Sensing Alteration Information Based on Integrated Spectral Mixture Analysis and Fractal Analysis","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eHydrothermal alteration of wall rocks near ore bodies is primarily the result of interactions between various types of hydrothermal fluids and surrounding rocks, representing the imprints of the progressive enrichment and concentration of ore-forming materials. Quantitative information related to geological structures, ore-bearing geological bodies, and alteration associated with metal mineralization can be comprehensively captured and interpreted through remote sensing-derived alteration signals across various data sources. These signals serve as indicators of potentially altered rocks proximal to mineralization zones, and the alteration information extracted via remote sensing provides a geological basis for guiding mineral deposit prediction (Crowley \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e1984\u003c/span\u003e; Asadzadeh et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Corr\u0026ecirc;a da Costa et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Habashi Costa et al. 2024; Islam et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Yang et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). As a key target area in China's new round of strategic mineral exploration initiatives, Tibet possesses favorable metallogenic conditions shaped by its unique geological evolution and tectonic setting (Huang et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Lin et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Zhao et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Although the region's harsh natural environment imposes multiple technical challenges on conventional surface exploration methods, its sparse vegetation cover offers a distinct advantage for accurately identifying and extracting alteration anomalies using remote sensing techniques (Dong et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Bai et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Fu et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Therefore, developing a high-precision remote sensing model for alteration information extraction has become urgent in advancing Tibet's new phase of mineral prospecting studies.\u003c/p\u003e \u003cp\u003eAt present, methods such as band ratio analysis (Abdelkader et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Baid et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Carvalho et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Yang et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), principal component analysis (PCA) (Mbianya et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Abedini et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Hegab \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Hussain et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), spectral mixture analysis (SMA) (Zhao and Zhao \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Chen et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Chen et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), spectral angle mapper (SAM) (Ren et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Ghoneim et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Yang and Tian \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and minimum noise fraction (MNF) transformation (Sikakwe et al. 2023; Yao et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Yuan et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2025\u003c/span\u003e)are widely applied in the extraction of mineralization-related alteration information. Among these methods, SMA has gradually attracted increasing attention from researchers due to its high accuracy, intuitive interpretability, and robustness. In remote sensing imagery, pixels represent the fundamental units of an image. However, due to spatial resolution limitations and small-scale surface features, a single pixel often contains spectral information from two or more distinct land cover types. Such pixels are referred to as mixed pixels (Naceur et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Miao et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Hosseinjani and Tangestani \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). The spectral characteristics of remote sensing imagery represent a mixed response of the spectral signatures of various endmembers comprising each pixel. During the extraction of alteration information from remote sensing images, the widespread presence of mixed pixels significantly reduces the accuracy and reliability of land cover identification. Interfering surface features such as vegetation, water bodies, shadows, and fluvial sediments can severely affect the effectiveness of alteration anomaly detection. Therefore, to accurately identify the spectral signatures of useful endmembers, it is essential to perform spectral mixture analysis on the imagery. This process helps to minimize or eliminate the influence of interfering land cover types, improve spectral discrimination accuracy, and ultimately enhance the extraction of alteration information (Sharifi and Hosseingholizadeh \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eComplex and irregular phenomena in nature often follow inherent scale-invariant laws. Based on this principle, Mandelbrot introduced the concept of fractal theory in 1983 (Mandelbrot 2019). As an effective method for quantitatively describing irregular shapes and spatial patterns, fractal theory has been widely applied to studying natural phenomena and complex systems (Kong and Ding \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; Cheng \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e1995\u003c/span\u003e; Zuo and Wang \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). This theory enables the extraction of deterministic and regular parameters from seemingly chaotic patterns, allowing the reconstruction of formation mechanisms of fractal structures and the inference of system evolution from stochastic processes. Many geological phenomena and spatial distributions in Earth systems exhibit pronounced fractal characteristics, particularly those associated with hydrothermal mineralization processes such as magma generation, migration, and intrusion, which often display varying degrees of self-similarity. On this basis, fractal theory serves as a powerful tool for identifying and extracting remote sensing-based alteration anomalies, thereby revealing the spatial organization and evolutionary patterns of mineralization. In recent years, fractal-based methods have been widely employed by researchers worldwide for remote sensing anomaly extraction and mineral exploration (Sun et al. \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Liu et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Zhao et al. \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Chen et al. 2012).\u003c/p\u003e \u003cp\u003eThe Saga region of Tibet is situated in a key segment of the central-western India\u0026ndash;Eurasia continental collision zone. It spans three major tectonic units from north to south: the Gangdese forearc basin, the Yarlung Zangbo suture zone, and the North Himalaya Tethyan fold-thrust belt. Although the area is characterized by a complex lithostratigraphic framework shaped by multiple phases of superimposed tectonic-magmatic activity, its high-altitude terrain, rugged topography, and limited accessibility pose significant challenges to field-based geological investigations. In this context, the present study focuses on the Saga region of Tibet. It employs Level-2A Sentinel-2 imagery to extract remote sensing alteration anomalies, including iron-staining, Al\u0026ndash;OH, Mg\u0026ndash;OH, and carbonate anomalies by the method of \"resampling\u0026thinsp;+\u0026thinsp;mixed pixel decomposition\u0026thinsp;+\u0026thinsp;principal component analysis\u0026thinsp;+\u0026thinsp;mixing ratio\u0026thinsp;+\u0026thinsp;threshold segmentation\" in this region. Furthermore, the fractal dimensions of the various alteration anomalies were calculated using fractal theory, enabling the characterization of their spatial distribution patterns and fractal structural features across the study area. Two potential areas have been identified using the composite contour map of fractal dimensions of alteration anomalies obtained from the superimposed analysis of various types of remotely sensed alteration anomalies (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"2. Overview of the Study Area","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Geological Profile\u003c/h2\u003e \u003cp\u003eThe study area is characterized by a wide distribution and diverse exposure of stratigraphic units, primarily comprising Paleozoic, Mesozoic, and Cenozoic sequences. Paleozoic strata are extensively exposed in the northern part of the region, while Mesozoic and Cenozoic formations are distributed throughout the entire area. The Paleozoic strata mainly include the Silurian\u0026ndash;Ordovician Dajiling and Ziqu formations and the Permian Zhongba, Gangzhutan, and Gazhale formations. The Mesozoic sequences consist of the Triassic Kagongyan, Yachuyan, Niukuyan, and Tunjuri formations; the Jurassic Jipuyan, Jiabula, Angren, Chuangde, and Zongzhuo formations; and the Cretaceous Zezuweng, Dasangyan, and Danga formations. The Cenozoic deposits are dominated by Quaternary alluvial sediments (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Intrusive rocks are widespread across the region and are primarily composed of Neogene monzogranite and Cretaceous diabase. The main types of mineralization identified in the area include malachite, epidote, chlorite, Sericite, hematite, and limonite alterations.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Overview of Remote Sensing Data\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe Sentinel-2 satellite (Forouzan and Arfania \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Ding et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Khashaba et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Zhang et al. \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Amani et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Boumahdi et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), part of the European Union's Copernicus Earth observation program, is equipped with a multispectral imager covering 13 spectral bands across the visible, near-infrared, and shortwave infrared regions. These bands offer spatial resolutions of 10 m, 20 m, and 60 m, respectively.\u003c/p\u003e \u003cp\u003eThis study utilized Level-2A Sentinel-2 imagery acquired on June 13, 2024, with satisfactory overall data quality (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). Although Level-2A products provide bottom-of-atmosphere reflectance data that have already undergone radiometric calibration and atmospheric correction, reconstruction of the low-resolution bands is still required during the image processing stage. In this study, the S2 Resampling tool provided by SNAP software was employed to reconstruct the low-resolution Sentinel-2 bands to a spatial resolution of 10 m while preserving their spectral characteristics. For the subsequent spectral mixture analysis and remote sensing-based extraction of alteration anomalies, specific Sentinel-2 bands were excluded based on their primary functions: Bands 1, 9, and 10 are mainly designed for coastal aerosol, cirrus clouds, and water vapor detection, respectively; additionally, Bands 8 and 8a have similar central wavelengths, but Band 8a is more sensitive to rock alteration. Therefore, ENVI software was used in this study to synthesize the Sentinel-2 data in bands other than 1, 8, 9, and 10. The interference, such as the fourth system and rain and snow in the study area, was masked to obtain the processed Sentinel-2 image data (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb) to carry out the study's next step.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"3. Research Methods","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Sequential Maximum Angle Convex Cone\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eSeveral models have been developed for spectral mixture analysis, among which the Linear Spectral Unmixing (LSU) model is the most widely applied due to its simplicity and ease of implementation (Zhang et al. \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Bhakthan et al. 2024). The LSU technique separates the spectra of each remote sensing data image by utilizing the research results, such as the combination of feature spectra for the characteristics of the mixed image elements. In the linear model of mixed pixels, the gray value of any pixel in given band reflects the co-reflections in that band from all places inside it, and its magnitude is linearly related to the area of the feature. The result of linear spectral separation is usually expressed as a series of gray-scale images of end-element spectra, also known as abundance images, in which the value of the image element indicates the proportion of the end-element spectra in the image element. The selection of end-elements is the prerequisite and key to the linear hybrid image decomposition method, and the type, number, and value of end-elements determine the success or failure of the hybrid image decomposition technique. The end elements are selected on the spectral image to represent the spectrum of the pure feature surface. The Sequential Maximum Angle Convex Cone (SMACC) method is a commonly used end-element extraction method for the Linear Spectral Unmixing model, which can simultaneously extract the end elements and their abundance proportion of each feature from the mixed image (Thompson et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Filippi et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Hamedianfar et al. 2024).\u003c/p\u003e \u003cp\u003eThe SMACC model's fundamental assumption is that a surface comprises a limited number of spectral endmembers. In a three-dimensional data cube representing remote sensing imagery\u0026mdash;consisting of spatial dimensions (x-axis and y-axis) and spectral bands\u0026mdash;the number of spectral bands is denoted by L, and the number of spectral endmembers by p. Each pixel within this cube can be interpreted as a vector in an L-dimensional spectral space, with each dimension corresponding to a specific spectral band. Thus, the spectral signature of each pixel is expressed as a linear combination of p spectral endmember vectors. Consequently, the original remote sensing image data can be mathematically described by:\u003c/p\u003e \u003cp\u003e \u003cem\u003eR\u0026thinsp;=\u0026thinsp;X\u0026thinsp;+\u0026thinsp;Noise\u0026thinsp;=\u0026thinsp;MS\u0026thinsp;+\u0026thinsp;Noise\u003c/em\u003e \u003c/p\u003e \u003cp\u003eHere, M is the mixing matrix, where each vector m(n) represents the spectral signature of the nth endmember. The matrix S corresponds to the fractional abundance of each endmember, with a scaling factor g, and an abundance vector represented by a = {\u003cem\u003ea₁, a₂, \u0026hellip;, aₚ\u003c/em\u003e} ᵀ. Noise accounts for additive random disturbances in the remote sensing image data (Chen et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003eThere are two assumptions about the endmembers: nonnegativity and sum-to-one. Mixing matrix M and abundance matrix S are required for nonnegativity (a\u0026thinsp;\u0026ge;\u0026thinsp;0), and the sum-to-one constraint is used for the pixel fraction. The sum of all endmembers in a single pixel is one, i.e., \u0026sum;\u003csub\u003e\u003cem\u003eq\u003c/em\u003e=1\u003c/sub\u003e\u003csup\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sup\u003e\u003cem\u003es\u003c/em\u003e(\u003cem\u003eq\u003c/em\u003e)\u0026thinsp;=\u0026thinsp;1. The simplex is represented as S\u003cem\u003ex\u003c/em\u003e = {\u003cem\u003ex\u003c/em\u003e\u0026isin;\u003cem\u003eR\u003c/em\u003e\u003csup\u003e\u003cem\u003eL\u003c/em\u003e\u003c/sup\u003e: \u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;\u003cem\u003eMα\u003c/em\u003e, \u003cem\u003eα\u003c/em\u003e\u0026thinsp;\u0026ge;\u0026thinsp;0, \u0026sum;\u003csub\u003e\u003cem\u003eq\u003c/em\u003e=1\u003c/sub\u003e\u003csup\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sup\u003e\u003cem\u003es\u003c/em\u003e(\u003cem\u003eq\u003c/em\u003e)\u0026thinsp;=\u0026thinsp;1.} when Noise is zero, and the simplex shape is like a convex cone Cp = {\u003cem\u003er\u003c/em\u003e\u0026isin;R\u003csup\u003eL\u003c/sup\u003e: \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;\u003cem\u003eMS\u003c/em\u003e, \u003cem\u003eα\u003c/em\u003e\u0026thinsp;\u0026ge;\u0026thinsp;0, \u0026sum;\u003csub\u003e\u003cem\u003eq\u003c/em\u003e=1\u003c/sub\u003e\u003csup\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sup\u003e\u003cem\u003es\u003c/em\u003e(\u003cem\u003eq\u003c/em\u003e)\u0026thinsp;=\u0026thinsp;1, \u003cem\u003eγ\u003c/em\u003e\u0026thinsp;\u0026ge;\u0026thinsp;0} (Nascimento and Dias \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2005\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAs illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, the algorithm operates within the simplex \u003cem\u003eSx\u003c/em\u003e (where \u003cem\u003eγ\u0026thinsp;=\u0026thinsp;1\u003c/em\u003e), defined by a convex combination of spectral signatures. In the first iteration, the data points are projected in the direction \u003cem\u003ef\u003c/em\u003e\u003csub\u003e1\u003c/sub\u003e, which emphasizes the contribution of the first endmember \u003cem\u003em\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e. The resulting projection identifies \u003cem\u003em\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e as the extreme point in that direction. In the second iteration, the data are re-projected along a new orthogonal direction, \u003cem\u003ef\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e, allowing the algorithm to extract the second endmember \u003cem\u003em\u003c/em\u003e\u003csub\u003e\u003cem\u003eb\u003c/em\u003e\u003c/sub\u003e. This iterative procedure continues, projecting the residual data in successively orthogonal directions until no additional distinct endmembers can be found.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Box-counting Fractal Method\u003c/h2\u003e \u003cp\u003eSince its introduction in the 1970s, fractal theory has experienced widespread application and continuous innovation within the quantitative analysis domain of Earth sciences (Qiu et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Lyu et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Maghsoudi et al. 2021). Alteration anomalies derived from remote sensing imagery typically exhibit complex and irregular spatial patterns, rendering traditional Euclidean geometry methods inadequate for accurately characterizing their spatial distribution. However, these alteration anomalies often display pronounced fractal self-similarity as an intrinsic feature. Therefore, fractal theory can effectively quantify the spatial self-similarity of alteration anomalies by calculating their fractal dimensions, providing a robust measure for characterizing the complexity of their spatial distributions. Several approaches have been developed for calculating fractal dimensions, including box-counting dimension (capacity dimension) (Ni et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Din\u0026ccedil; et al. 2023), information dimension (Sun et al. \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Shi and Xiao 2025), similarity dimension (Zhao et al. \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Cheng et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), correlation dimension (Liu et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Liu et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and Hausdorff dimension (Fern\u0026aacute;ndez-Mart\u0026iacute;nez and S\u0026aacute;nchez-Granero \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Hummer 2017). Among these, the box-counting dimension method has become the most widely adopted approach due to its straightforward theoretical foundation and computational simplicity.\u003c/p\u003e \u003cp\u003eThe implementation of the box-counting method involves overlaying the study area with a two-dimensional square grid of varying box sizes (side length\u0026thinsp;=\u0026thinsp;\u003cem\u003er\u003c/em\u003e) and counting the number of boxes, \u003cem\u003eN(r)\u003c/em\u003e, that contain alteration anomaly information at each scale. The relationship between the number of boxes \u003cem\u003eN(r)\u003c/em\u003e and box size r follows a power-law distribution, as expressed by Eq.:\u003c/p\u003e \u003cp\u003e \u003cem\u003eN(r)\u0026thinsp;=\u0026thinsp;Cr\u003c/em\u003e \u003csup\u003e \u003cem\u003e\u0026minus;\u0026thinsp;D\u003c/em\u003e \u003c/sup\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eWhere C is a constant, and D is the box dimension subdimensional value. Take the logarithm of each side of the above equation:\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003elnN(r)=-Dlnr\u0026thinsp;+\u0026thinsp;lnc\u003c/em\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eAccording to the above formula, the \u003cem\u003elog r-log N(r)\u003c/em\u003e curve is plotted and fitted by the least squares method. The slope of the resulting straight line represents the box dimension subdimension \u003cem\u003eD\u003c/em\u003e. The correlation coefficient \u003cem\u003eR\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e is also obtained, and the closer the value of \u003cem\u003eR\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e is to 1, the better the fit is, and the better the consistency between the data and the power law model. In this study, the fractal dimension of remote sensing-derived alteration anomalies in the study area was calculated using the box-counting method described above, and corresponding fractal dimension contour maps were subsequently generated. The specific procedures are as follows: first, an initial observational scale (\u003cem\u003er\u003c/em\u003e) was defined, and two-dimensional orthogonal grids with side lengths of \u003cem\u003er\u0026thinsp;=\u0026thinsp;r₀, r₀/2, r₀/4\u003c/em\u003e, and \u003cem\u003er₀/8\u003c/em\u003e were used to overlay the study area (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) sequentially. For each scale, the number of grid cells \u003cem\u003eN(r)\u003c/em\u003e containing alteration anomaly information was recorded. Next, a log-log plot of r versus \u003cem\u003eN(r)\u003c/em\u003e was constructed, and the box dimension (\u003cem\u003eD\u003c/em\u003e), along with its coefficient of determination (\u003cem\u003eR\u0026sup2;\u003c/em\u003e), was obtained through least-squares linear regression. This approach enabled a quantitative characterization and evaluation of the spatial distribution of remote sensing alteration anomalies.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4. Results and Discussion","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Spectral Mixture Analysis\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eUsing ENVI software, the Sentinel-2 imagery underwent a linear transformation, after which the SMACC method was applied to extract endmember spectra and abundance information. The processed imagery was decomposed into abundance images corresponding to nine distinct land-cover endmembers (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e) and one shadow image, with the total abundance summing to one. Additionally, spectral vectors were generated for each of the nine endmember components (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). In the experiment, these nine endmember spectral vectors were compared and identified against standard spectra from a reference spectral library. By combining this spectral comparison with their spatial distribution patterns in the abundance images, endmember component 3 was identified as vegetation, while endmember components 4 and 6 were classified as other interference features.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSpectral vectors of nine endmember waveforms from remotely sensed images of the study area.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBand 3\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBand 4\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eBand 5\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eBand 6\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBand 7\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eBand 8a\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eBand 11\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003eBand 12\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4912\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5948\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e6560\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e6675\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e6676\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e6734\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e6808\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7893\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e6424\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6512\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e6640\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5651\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e5453\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e5157\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e4789\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2267\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e2162\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3444\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2256\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2268\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e3435\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3367\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e3691\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e4311\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e6460\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e6342\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1675\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2190\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2122\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2777\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4259\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e4604\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e4920\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4181\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e3007\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6572\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6916\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e6936\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4633\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4448\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e4396\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e4204\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2853\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e2534\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3856\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4012\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4372\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e3648\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3618\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e4031\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e4278\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e5825\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e6136\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2440\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3432\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4272\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4548\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4702\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e4913\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e4987\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7146\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e6375\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4524\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5092\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e5044\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e5638\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e5435\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e4974\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2311\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e1975\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2144\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2636\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2960\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e3083\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3133\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e3325\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e3489\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e6116\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e5621\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eAfter determining the reflectance and abundance values of vegetation and interference endmembers, spectral unmixing was performed on each pixel band using a linear spectral unmixing model. Using Band 2 as an example, the abundance values of the identified interference endmember components (F₃, F₄, F₆) were multiplied by their corresponding reflectance values in Band 2. These products were then summed to create a composite image that represents vegetation and other interference endmembers. Next, the summed interference reflectance was subtracted from the original digital number (DN₂) of Band 2, resulting in an image that reflects only the non-vegetation components. This non-vegetation reflectance image was divided by its fractional abundance (1\u0026thinsp;\u0026minus;\u0026thinsp;F₃ \u0026minus; F₄ \u0026minus; F₆), resulting in a compensated reflectance image that more accurately represents rock and soil endmembers by effectively removing interference contributions. Ultimately, spectral reconstruction for rock and soil pixels across all image bands in the study area was achieved through the reconstruction calculation (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Following spectral mixture analysis, rocky terrain in sparsely vegetated mountainous areas became distinctly apparent. Additionally, in densely vegetated regions, the vegetation signal was effectively suppressed without excessive masking, thereby preserving alteration information on rock surfaces and enhancing the reliability of subsequent alteration anomaly extraction in this study.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCalculation formula for reconstructed images of the study area.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBand\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCalculation formula\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(DN1-F3\u0026times;3444-F4\u0026times;1675-F6\u0026times;3856)/༈1-F3-F4-F6༉\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(DN2-F3\u0026times;2256-F4\u0026times;2190-F6\u0026times;4012)/༈1-F3-F4-F6༉\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(DN3-F3\u0026times;2268-F4\u0026times;2122-F6\u0026times;4372)/༈1-F3-F4-F6༉\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(DN4-F3\u0026times;3435-F4\u0026times;2777-F6\u0026times;3648)/༈1-F3-F4-F6༉\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(DN5-F3\u0026times;3367-F4\u0026times;4259-F6\u0026times;3618)/༈1-F3-F4-F6༉\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(DN6-F3\u0026times;3691-F4\u0026times;4604-F6\u0026times;4031)/༈1-F3-F4-F6༉\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8a\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(DN7-F3\u0026times;4311-F4\u0026times;4920-F6\u0026times;4278)/༈1-F3-F4-F6༉\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(DN8-F3\u0026times;6460-F4\u0026times;4181-F6\u0026times;5825)/༈1-F3-F4-F6༉\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(DN9-F3\u0026times;6342-F4\u0026times;3007-F6\u0026times;6136)/༈1-F3-F4-F6༉\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Remote Sensing Alteration Anomaly Information Extraction\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eIn geological prospecting, the presence of altered rocks in a region does not necessarily mean that mineral deposits exist. Significant and widely distributed wall-rock alterations often occur in large to super-large ore deposits. Unaltered and altered rocks exhibit significant differences in mineral composition and lithology, resulting in distinct spectral reflectance characteristics. Therefore, differences in spectral properties serve as the theoretical basis for extracting alteration information through remote sensing. The Saga region exhibits various mineralogical alteration features associated with ore deposits, including malachitization, epidotization, chloritization, sericitization, hematitization, and limonitization. Minerals characteristic of such wall-rock alterations typically contain abundant Fe\u0026sup3;⁺, Fe\u0026sup2;⁺, OH⁻, or CO₃\u0026sup2;⁻ groups. The electron vibration processes of these structural ions yield diagnostic spectral features in minerals enriched with these ions or groups, which are critical indicators in mineral exploration. In this study, diagnostic spectral features of typical alteration minerals were analyzed using reference spectral curves from the USGS spectral library to extract relevant alteration anomaly information. Representative spectral curves of these typical alteration minerals are presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eAccording to previous studies, PCA and band ratio methods are commonly utilized to extract remote sensing alteration information. In this study, PCA was applied to identify alteration anomalies associated with iron-staining and hydroxyl-bearing minerals (Al\u0026ndash;OH and Mg\u0026ndash;OH), based on their characteristic spectral absorption features (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Additionally, the band ratio method was employed to detect carbonate alteration anomalies by exploiting the distinct reflectance characteristics of carbonate minerals in Bands 7 and 12.\u003c/p\u003e \u003cp\u003eIron-staining minerals (such as hematite, limonite, and goethite) exhibit strong reflectance features in the wavelength ranges of 0.48\u0026ndash;0.51 \u0026micro;m and 0.85\u0026ndash;0.89 \u0026micro;m, corresponding to Band 2 and Band 8a of Sentinel-2 imagery. These minerals also display a characteristic absorption feature between 1.52\u0026ndash;1.70 \u0026micro;m, corresponding to Sentinel-2 Band 11. Therefore, Bands 2, 4, 8a, and 11 were selected as an optimal combination for extracting iron-staining alteration anomalies. In principal component analysis, iron-staining anomalies are typically characterized by negative and identical contribution coefficient signs for Bands 2 and 8a, with an opposite coefficient sign for Band 11. Examination of the transformed eigenvectors indicated that the anomaly information related to iron-staining was primarily concentrated in the PC3 (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Consequently, PC3 was chosen as the principal component for the extraction of iron-staining anomalies.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCalculation formula for reconstructed images of the study area.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eType\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEigenvector\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBand 2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBand 4\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eBand 8A\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eBand 11\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eIron-\u003c/p\u003e \u003cp\u003estaining\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePC1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.313681\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.478547\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.517353\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.636351\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePC2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.617418\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.340468\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.160370\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.690767\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePC3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.568580\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.785637\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.181369\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.163086\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePC4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.443977\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.194549\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.820812\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.302163\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eType\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eEigenvector\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eBand 6\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eBand 8A\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003eBand 11\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003eBand 12\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eAl-OH\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePC1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.480088\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.506195\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.516806\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.49618\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePC2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.468644\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.504798\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.674377\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.26602\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePC3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.419601\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.146771\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.480054\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.756269\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePC4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.611411\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.68367\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.218348\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.333311\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eType\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eEigenvector\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eBand 2\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eBand 8A\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003eBand 11\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003eBand 12\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eMg-OH\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePC1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.410902\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.526081\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.537109\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.515667\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePC2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.571009\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.46045\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.619561\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.279426\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePC3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.543775\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.362285\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.491561\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.575698\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePC4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.457622\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.61642\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.293304\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.56972\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eOH groups alteration minerals primarily include Al\u0026ndash;OH minerals (such as kaolinite, muscovite, and alunite) and Mg\u0026ndash;OH minerals (such as chlorite, epidote). Al\u0026ndash;OH minerals exhibit reflective spectral features in the wavelength range of 1.60\u0026ndash;1.70 \u0026micro;m, corresponding to Sentinel-2 Band 11, and absorption features near 2.20 \u0026micro;m, corresponding to Sentinel-2 Band 12. Consequently, Bands 6, 8a, 11, and 12 were selected as the optimal band combination for identifying Al\u0026ndash;OH mineral alteration anomalies. PCA was then performed using these selected bands (Bands 6, 8a, 11, and 12). The principal component related to Al\u0026ndash;OH anomalies shows positive and identical contribution coefficients for Bands 8a and 11, while Bands 11 and 12 exhibit opposite signs. Examination of the transformed eigenvectors indicated that the inverse of principal component 4 (\u0026ndash;PC4) best matched these criteria (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Therefore, \u0026ndash;PC4 was utilized as the principal component for extracting Al\u0026ndash;OH mineral alteration anomalies in this study.\u003c/p\u003e \u003cp\u003eMg\u0026ndash;OH minerals exhibit an increasing reflectance trend in the wavelength range of 0.70\u0026ndash;1.82 \u0026micro;m, corresponding to Bands 5\u0026ndash;11 of Sentinel-2 imagery, and display characteristic absorption near 2.32 \u0026micro;m, corresponding to Sentinel-2 Band 12. Accordingly, Bands 2, 8a, 11, and 12 were selected as the optimal band combination for extracting Mg\u0026ndash;OH mineral alteration anomalies. In the PCA, the principal component associated with Mg\u0026ndash;OH anomalies typically shows positive and identical coefficient signs for Bands 2 and 11 and opposite signs for Bands 11 and 12. Based on these criteria, eigenvector PC3 was selected as the principal component for identifying Mg\u0026ndash;OH mineral alteration anomalies.\u003c/p\u003e \u003cp\u003eCarbonate minerals (such as calcite and dolomite) typically exhibit prominent reflectance peaks around 0.7\u0026ndash;0.8 \u0026micro;m and 1.7\u0026ndash;1.8 \u0026micro;m, corresponding to Bands 7 and 12 of Sentinel-2 imagery, respectively. Based on multiple experimental analyses, a hybrid band ratio method using Band 12/Band 8a and Band 7/Band 4 was adopted in this study. This mixed-ratio approach enhances the diagnostic spectral signatures of carbonate minerals, thereby effectively extracting associated alteration anomaly information.\u003c/p\u003e \u003cp\u003eAfter the characteristic principal components or hybrid bandratio images representing the four alteration types had been extracted, we calculated the mean (\u0026micro;) and standard deviation (σ) of the alteration-type images. Pixels with values greater than \u0026micro;\u0026thinsp;+\u0026thinsp;Nσ were classified as remotesensing alteration anomalies. The multiplier N was set to 3 for ironstaining and Al\u0026ndash;OH and Mg\u0026ndash;OH anomalies and to 2 for carbonate anomalies. The final Sentinel-2 alteration anomaly map is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Characterization of Alteration Anomalies in Remote Sensing Using Fractal Dimensions\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe fractal dimension provides a useful indicator for locating ore deposits. In this study, box-counting fractal analysis was applied to four alteration-anomaly layers derived from Sentinel-2 imagery: iron staining, Al\u0026ndash;OH, Mg\u0026ndash;OH, and carbonate. This analysis aimed to pinpoint the most promising mineralization sites in the study area and provide theoretical guidance for future exploration efforts. Four two-dimensional orthogonal grids with cell sizes of 6.123 km, 3.062 km, 1.531 km, and 0.765 km were successively superimposed on the study area (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). For each grid scale, the number of square cells \u003cem\u003eN(r)\u003c/em\u003e containing the respective alteration anomaly was recorded. The natural logarithm of cell size, ln r, was plotted on the xaxis and ln \u003cem\u003eN(r)\u003c/em\u003e on the yaxis; leastsquares regression lines were then fitted separately for each alteration type. The slope of each regression line yields the boxcounting dimension \u003cem\u003eD\u003c/em\u003e. The statistical results are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, and the \u003cem\u003eln N(r) \u0026ndash; ln r\u003c/em\u003e regression plots are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCalculation formula for reconstructed images of the study area.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eType\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e \u003cp\u003efractal dimension\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eType\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c10\" namest=\"c7\"\u003e \u003cp\u003efractal dimension\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003er\u003c/em\u003e/km\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003eN\u003c/em\u003e(\u003cem\u003er\u003c/em\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eln\u003cem\u003er\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLn\u003cem\u003eN\u003c/em\u003e(\u003cem\u003er\u003c/em\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003cem\u003er\u003c/em\u003e/km\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cem\u003eN\u003c/em\u003e(\u003cem\u003er\u003c/em\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eln\u003cem\u003er\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003eLn\u003cem\u003eN\u003c/em\u003e(\u003cem\u003er\u003c/em\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eIron-\u003c/p\u003e \u003cp\u003estaining\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.123\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5.77455\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.46574\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eAl-OH\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e6.123\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e5.42935\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e3.46574\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.062\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.74493\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.15888\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3.062\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e4.60517\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e4.15888\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e115\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.55535\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.85203\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e3.49651\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e4.85203\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.765\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e322\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.48491\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.54518\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.765\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e228\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e2.3979\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e5.54518\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eMg-OH\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.123\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5.54908\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.46574\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eCarbonate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e6.123\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e4.61512\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e3.46574\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.062\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.7362\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.15888\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3.062\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e4.00733\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e4.15888\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e114\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.52636\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.85203\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e3.17805\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e4.85203\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.765\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e257\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.48491\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.54518\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.765\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e101\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e2.30259\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e5.54518\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eUsing the fractal method described above, fractal dimensions for each alteration type were calculated over the scale range of 0.765\u0026ndash;6.123 km. The results are as follows: the ironstaining anomalies have an overall fractal dimension of 1.59541 with \u003cem\u003eR\u0026sup2;\u003c/em\u003e = 0.9968; the Al\u0026ndash;OH anomalies, 1.47198 with \u003cem\u003eR\u0026sup2;\u003c/em\u003e = 0.9921; the Mg\u0026ndash;OH anomalies, 1.50074 with \u003cem\u003eR\u0026sup2;\u003c/em\u003e = 0.9916; and the carbonate anomalies, 1.12053 with \u003cem\u003eR\u0026sup2;\u003c/em\u003e = 0.9925. The highest fractal dimension value of the iron-stained alteration in the study area indicates significant spatial structural complexity. This complexity may be closely related to mineralization centers formed by the oxidation of metal sulfides or the intersection of fracture sites. It is presumed to be a result of the main metallogenic period. Therefore, areas with high-D zones should be prioritized as core targets for exploration. The similar fractal dimensions of the Al\u0026ndash;OH and Mg\u0026ndash;OH anomalies imply comparable spatial complexities, likely reflecting different stages of the same mineralizing system; highD areas delineate the probable boundaries of this system. By contrast, the lower fractal dimension of the carbonate anomalies points to a more uniform spatial distribution characteristic of peripheral or late-stage alteration zones. All alteration types exhibit R\u0026sup2; \u0026gt; 0.99, confirming pronounced statistical self-similarity within the selected scale domain.\u003c/p\u003e \u003cp\u003eThe fractal dimension statistics of the previously described division grid were calculated to determine the fractal dimension value of the etching anomalies within each grid range. Each grid was assigned a fractal dimension value based on its center. Using Surfer software, the Kriging interpolation method was then employed to create contour maps depicting the various alteration anomalies. The resulting contour map illustrates the fractal dimension information for different alteration anomalies, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe fractal dimension contour maps for the various alteration types show that ironstaining anomalies are markedly stronger, in both magnitude and areal extent, than the other alteration types, whereas the highvalue zones for carbonate alteration are the least extensive and least intense. High fractaldimension values for ironstaining, Al\u0026ndash;OH, and Mg\u0026ndash;OH anomalies are concentrated principally within the Silurian\u0026ndash;Ordovician Dajiling Formation and the Jurassic Zongzhuo Formation. The Dajiling Formation is dominated by quartz schist, quartz sandstone, and muscovite-bearing quartzite, whereas the Zongzhuo Formation consists mainly of argillaceous siltstone, shale, and limestone. These strata are relatively unconsolidated, highly cracked, and rich in minerals that are readily altered, thereby providing favorable pathways and reactive sites for hydrothermal fluids. The lithologic and structural characteristics likely control the enrichment of the alteration anomalies in these units. Additionally, the anomalies of iron-staining, Al\u0026ndash;OH, and Mg\u0026ndash;OH show a strong spatial relationship with most known sites of Cu and Fe mineralization; the deposits are generally found within their respective high-value zones. This coherence reflects intense hydrothermal activity and robust mineralization in the area. Accordingly, both the Dajiling and Zongzhuo formations not only provide favorable conditions for alteration development but also represent the most prospective stratigraphic units for exploration. By contrast, carbonate-related alteration anomalies display high values only in limited portions of the Zongzhuo Formation and exhibit a comparatively weak spatial match with known mineralization points. This suggests that their predictive significance is limited and that they may indicate mineral potential only under specific structural settings or lithologic combinations.\u003c/p\u003e \u003cp\u003eOverall, the fractal characteristics of the different alteration anomalies reveal their spatial complexity and genetic mechanisms. When combined with the spatial coupling between each alteration type and existing mineralization sites, these features constitute a valuable basis for regional prospecting, offering scientific support for identifying potential mineralization targets. To integrate the fractaldimension contour maps of the various alteration types in the Saga area, we normalized each dataset to prevent the spatial characteristics of any one alteration type from being masked by large absolute values during a simple arithmetic overlay. This normalization effectively eliminated the problem of subdued anomaly zones in the composite layer that would otherwise arise from disproportionately high values in a single alteration map. The resulting composite fractaldimension contour map shows that high-value zones are concentrated mainly within the Silurian\u0026ndash;Ordovician Dajiling Formation and the Jurassic Zongzhuo Formation, exhibiting a conspicuous concentricdiffusion pattern. Based on the distribution of previously identified mineralization sites, areas with composite fractal values exceeding 0.7 were designated as high-potential mineralization zones, and two additional prospective targets were delineated outside the known mineralization sites (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.4. Results of Field Survey and Laboratory Analysis\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eField Survey is a critical approach for validating indoor remote sensing interpretation results. Based on the preceding findings, a geological traverse survey was conducted in Mineralization Potential Zone I. The stratigraphy of this zone is primarily composed of the Dajiling Formation. During the field investigation, several well-defined quartz vein outcrops were observed near high-value anomaly zones. Some of these quartz veins exhibited evident signs of malachite staining and chalcopyrite mineralization. Quartz vein samples were collected, prepared into thin sections, and examined under a polarizing microscope. The petrographic analysis revealed multiple types of mineralization within the quartz veins, including Cu mineralization (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003ea), malachitization (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003eb), sericitization (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003ec), pyritization (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003ed), limonitization (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003ee), and galena mineralization (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003ef). In terms of structural features, the rocks predominantly exhibit hydrothermal infill and replacement textures, including brecciated, vein-like, stockwork, fine vein-disseminated, and massive textures. Among these, fine vein-disseminated, densely disseminated, and brecciated textures are the most common. The fine and densely disseminated textures occur mainly within stratiform ore bodies, whereas brecciated textures are typically associated with vein-type deposits. Furthermore, in schistositized marble, disseminated limonitization, chloritization, and spotty limonitization are observed, with localized limonitization forming banded patterns along the schistositized planes. The above field observations corroborate our earlier analysis of the lithological characteristics of the Dajiling Formation. The rocks of this formation are generally loosely consolidated, extensively cracked, and rich in minerals susceptible to alteration, thereby providing favorable pathways and reactive sites for hydrothermal fluid infiltration, migration, and mineralization. These conditions have promoted the enrichment of the alteration anomalies observed. This finding not only validates the reliability of the methods applied in this study but also furnishes critical evidence for the verification of additional mineralization potential zones. Future work will involve groundtruthing another prospective zone within the study area.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"5. Conclusions","content":"\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003e(1)Introducing spectral mixture analysis into the study area enables the separation of reflectance and abundance values for the various interference components contained in mixed pixels. Spectral unmixing and endmember reconstruction enhance the detectability of subtle mineralization-related alteration anomalies in the remotesensing imagery. By accurately quantifying the spectral contribution of each surface feature, this approach prevents the loss of alteration signatures that often occurs with conventional masking techniques due to over-correction, and it markedly improves both the accuracy and completeness of alteration information extraction under diverse surfacecover conditions.\u003c/p\u003e \u003cp\u003e(2)The fractal dimensions of the various remotesensing alteration anomalies were determined using fractal analysis, thereby quantitatively characterizing their spatial self-similarity. The ironstaining anomalies have a fractal dimension of 1.59541 with R\u0026sup2; = 0.9968; the Al\u0026ndash;OH anomalies, 1.47198 with R\u0026sup2; = 0.9921; the Mg\u0026ndash;OH anomalies, 1.50074 with R\u0026sup2; = 0.9916; and the carbonate anomalies, 1.12053 with R\u0026sup2; = 0.9925. Notably, all alteration types exhibit R\u0026sup2; values greater than 0.99, confirming that the alteration anomalies in the study area display pronounced statistical self-similarity within the selected scale range.\u003c/p\u003e \u003cp\u003e(3)The highpotential mineralization zones delineated in the fractaldimension contour map of alteration anomalies show a strong spatial correspondence with known mineralization sites. In addition, two new prospective zones were identified on the periphery of the existing mineralization sites. These findings furnish a sound theoretical basis and clear exploration priorities for the next stage of prospecting in the study area.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor contributions\u0026nbsp;\u003c/strong\u003eConceptualization: K.Q.; methodology: \u0026nbsp;K.Q.; software: \u0026nbsp;K.Q. and S.D.; validation: \u0026nbsp;K.Q., and J.K.; formal analysis: \u0026nbsp;K.Q. and T.L.; investigation: \u0026nbsp;K.Q.; L.Q.; and Z.R.; resources: \u0026nbsp;Z.R. and S.G.; data curation: K.Q.; writing\u0026mdash;original draft preparation: \u0026nbsp;J.K.; writing\u0026mdash;review and editing: \u0026nbsp;Y.L. and S.D.; visualization: \u0026nbsp;K.Q.; L.Q.; and S.D.; supervision: \u0026nbsp;T.L. and Y.H.; project administration: \u0026nbsp;T.L.; funding acquisition: \u0026nbsp;Y.H.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u0026nbsp;\u003c/strong\u003eThis work was supported by the China Geological Survey Project (DD20240014, DD20243088), the Ministry of Natural Resources\u0026apos; new round of scientific and technological support project for finding mineral breakthroughs (ZKKJ202427-03), Tibet Science and Technology Program Projects (XZ202401YD0006-07).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u0026nbsp;\u003c/strong\u003eThe datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u0026nbsp;\u003c/strong\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAbdelkader, M. A. et al. Advanced exploration of rare metal mineralization through integrated remote sensing and geophysical analysis of structurally-controlled hydrothermal alterations. \u003cem\u003eJ. Geochem. Explor.\u003c/em\u003e \u003cb\u003e267\u003c/b\u003e, 107598 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbedini, M., Ziaii, M., Timkin, T. \u0026amp; Pour, A. B. Machine Learning (ML)-Based Copper Mineralization Prospectivity Mapping (MPM) Using Mining Geochemistry Method and Remote Sensing Satellite Data. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e (15), 3708 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAmani, S., Shafizadeh-Moghadam, H. \u0026amp; Morid, S. Integrating sentinel-2 and sentinel-3 for actual evapotranspiration estimation across diverse climate zones using the sen-ET plugin and machine learning models. \u003cem\u003eEarth Sci. Inf.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e (4), 338 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAsadzadeh, S., Chabrillat, S., Cudahy, T., Rashidi, B. \u0026amp; de Souza Filho, C. R. Alteration Mineral Mapping of the Shadan Porphyry Cu-Au Deposit (Iran) Using Airborne Imaging Spectroscopic Data: Implications for Exploration Drilling. \u003cem\u003eEcon. Geol.\u003c/em\u003e \u003cb\u003e119\u003c/b\u003e (1), 139\u0026ndash;160 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBai, L. Y. et al. Predictive Prospecting Using Remote Sensing in a Mountainous Terrestrial Volcanic Area, in Western Bangongco\u0026ndash;Nujiang Mineralization Belt, Tibet. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e (19), 4851 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaid, S. et al. Integrating geochemical insights and remote sensing for enhanced identification of hydrothermal alterations in the Igoudrane region, Anti-Atlas, Morocco. \u003cem\u003eJ. Afr. Earth Sc.\u003c/em\u003e \u003cb\u003e218\u003c/b\u003e, 105368 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBhakthan, S. M. \u0026amp; Loganathan, A. A hyperspectral unmixing model using convolutional vision transformer. \u003cem\u003eEarth Sci. Inf.\u003c/em\u003e \u003cb\u003e17\u003c/b\u003e (3), 2255\u0026ndash;2273 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoumahdi, M., Garc\u0026iacute;a-Pedrero, A., Lillo-Saavedra, M. \u0026amp; Gonzalo-Martin, C. A new benchmark for spatiotemporal fusion of Sentinel-2 and Sentinel-3 OLCI images. \u003cem\u003eEarth Sci. Inf.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e (2), 349 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarvalho, M., Cardoso-Fernandes, J., Gonz\u0026aacute;lez, F. J. \u0026amp; Teodoro, A. C. Comparative Performance of Sentinel-2 and Landsat-9 DataforRawMaterials\u0026rsquo; Exploration OnshoreandinCoastal Areas. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e17\u003c/b\u003e (2), 305 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, J. et al. Using geochemical data for prospecting target areas by the sequential maximum angle convex cone method in the Manzhouli area, China. \u003cem\u003eGeochem. J.\u003c/em\u003e \u003cb\u003e52\u003c/b\u003e (1), 13\u0026ndash;27 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, Q. et al. ASTER and GF-5 Satellite Data for Mapping Hydrothermal Alteration Minerals in the Longtoushan Pb-Zn Deposit, SW China. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e (5), 1253 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, J. et al. Metallogenic prediction based on fractal theory and machine learning in Duobaoshan Area, Heilongjiang Province. (2024). Ore Geol Rev 106030.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, Q. et al. Remote sensing identification of hydrothermal alteration minerals in the Duobuza porphyry copper mining area in Tibet using WorldView-3 and GF-5 data: The impact of spatial and spectral resolution. \u003cem\u003eOre Geol. Rev.\u003c/em\u003e \u003cb\u003e180\u003c/b\u003e, 106573 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheng, Q. M. The perimeter-area fractal model and its application to geology. \u003cem\u003eMath. Geol.\u003c/em\u003e \u003cb\u003e27\u003c/b\u003e, 69\u0026ndash;82 (1995).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheng, L. et al. Self-similarity study based on the particle sizes of coal-series diatomite. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e (1), 7502 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCorr\u0026ecirc;a da Costa, M. A., Carneiro Naleto, J. L., Perrotta, M. M., Soares Monteiro, L. V. \u0026amp; de Souza Filho, C. R. Visible-Near Infrared\u0026ndash;Short-Wave Infrared Spectroscopy and Mineral Mapping of Hydrothermal Alteration Zones at the Breju\u0026iacute; W-Mo Skarn Deposit, Serid\u0026oacute; Mobile Belt, Borborema Province, Brazil. \u003cem\u003eEcon. Geol.\u003c/em\u003e \u003cb\u003e119\u003c/b\u003e (5), 1171\u0026ndash;1198 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCrowley, J. K. Near-infrared reflectance of zunyite; implications for field mapping and remote-sensing detection of hydrothermally altered high alumina rocks. \u003cem\u003eEcon. Geol.\u003c/em\u003e \u003cb\u003e79\u003c/b\u003e (3), 553\u0026ndash;557 (1984).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDing, H. F. et al. Research on Scale Improvement of Geochemical Exploration Based on Remote Sensing Image Fusion. Remote Sens 15(8): 1993 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDin\u0026ccedil; G\u0026ouml;ğ\u0026uuml;ş, \u0026Ouml;., Avşar, E., Develi, K. \u0026amp; \u0026Ccedil;alık, A. Quantifying the Rock Damage Intensity Controlled by Mineral Compositions: Insights from Fractal Analyses. \u003cem\u003eFractal Fract.\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e (5), 383 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong, X. F., Gan, F. P., Li, N., Zhang, S. F. \u0026amp; Li, T. T. Mineral mapping in the Duolong porphyry and epithermal ore district, Tibet, using the Gaofen-5 satellite hyperspectral remote sensing data. \u003cem\u003eOre Geol. Rev.\u003c/em\u003e \u003cb\u003e151\u003c/b\u003e, 105222 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFern\u0026aacute;ndez-Mart\u0026iacute;nez, M. \u0026amp; S\u0026aacute;nchez-Granero, M. A. Fractal dimension for fractal structures: A Hausdorff approach revisited. \u003cem\u003eJ. Math. Anal. Appl.\u003c/em\u003e \u003cb\u003e409\u003c/b\u003e (1), 321\u0026ndash;330 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFilippi, A. M. et al. Comparison of Image Endmember- and Object-Based Classification of Very-High-Spatial-Resolution Unmanned Aircraft System (UAS) Narrow-Band Images for Mapping Riparian Forests and Other Land Covers. \u003cem\u003eLand\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e (2), 246 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eForouzan, M. \u0026amp; Arfania, R. Integration of the bands of ASTER, OLI, MSI remote sensing sensors for detection of hydrothermal alterations in southwestern area of the Ardestan, Isfahan Province, Central Iran. \u003cem\u003eEgypt. J. Remote Sens. Space Sci.\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e (2), 145\u0026ndash;157 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFu, Y. F., Cheng, Q. M., Jing, L. H., Ye, B. \u0026amp; Fu, H. Z. Mineral Prospectivity Mapping of Porphyry Copper Deposits Based on Remote Sensing Imagery and Geochemical Data in the Duolong Ore District, Tibet. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e (2), 439 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGhoneim, S. M., Ali, H. F., Abdelrahman, K., Fnais, M. S. \u0026amp; Lala, A. M. S. Integrating remote sensing and geophysical data for mapping potential gold mineralization localities at Abu Marawat area, central Eastern Desert, Egypt. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e (1), 18273 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHabashi, J., Oskouei, M. M., Moghadam, H. J. \u0026amp; Pour, A. B. Optimizing alteration mineral detection: A fusion of multispectral and hyperspectral remote sensing techniques in the Sar-e-Chah-e Shur, Iran. Remote Sensing Applications: Society and Environment 35: 101249. (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHamedianfar, A. et al. Leveraging High-Resolution Long-Wave Infrared Hyperspectral Laboratory Imaging Data for Mineral Identification Using Machine Learning Methods. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e (19), 4806 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHegab, M. A. E. R. A multi-disciplinary approach for uranium exploration using remote sensing and airborne gamma-ray spectrometry data in the Gebel Duwi area, Central Eastern Desert, Egypt. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e (1), 19739 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHosseinjani, M. \u0026amp; Tangestani, M. H. Mapping alteration minerals using sub-pixel unmixing of ASTER data in the Sarduiyeh area, SE Kerman, Iran. \u003cem\u003eInt. J. Digit. Earth\u003c/em\u003e. \u003cb\u003e4\u003c/b\u003e (6), 487\u0026ndash;504 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, M. L. et al. Linking Porphyry Cu Formation to Tectonic Change in Postsubduction Settings: A Case Study from the Giant Yulong Belt, Eastern Tibet. \u003cem\u003eEcon. Geol.\u003c/em\u003e \u003cb\u003e119\u003c/b\u003e (2), 279\u0026ndash;304 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHummer, D. R. Fractal distribution of mineral species among the crystallographic point groups. \u003cem\u003eAm. Mineral.\u003c/em\u003e \u003cb\u003e106\u003c/b\u003e (10), 1574\u0026ndash;1579 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHussain, M., Li, H., Beshr, A. M. \u0026amp; Memon, F. A. Remote sensing-based structural and lithological mapping for prospecting polymetallic mineralization at Xiaoshan region, China. \u003cem\u003eEarth Sci. Inf.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e (1), 82 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIslam, N. U. et al. Mineralogical mapping and lithological discrimination by using ASTER remote sensing data in the Chitral region, Khyber Pakhtunkhwa, Northern Pakistan. \u003cem\u003eEarth Sci. Inf.\u003c/em\u003e \u003cb\u003e17\u003c/b\u003e (6), 6075\u0026ndash;6094 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhashaba, S. M. A., El-Shibiny, N. H., Hassan, S. M., Takazawa, E. \u0026amp; Khedr, M. Z. Application of remote sensing data integration in detecting mineralized granitic zones: A case study of the Gabal Al-Ijlah Al-Hamra, Central Eastern Desert, Egypt. \u003cem\u003eJ. Afr. Earth Sc.\u003c/em\u003e \u003cb\u003e200\u003c/b\u003e, 104855 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKong, F. C. \u0026amp; Ding, G. Y. The implications the fractal dimension values of lineaments. \u003cem\u003eEarthquake\u003c/em\u003e \u003cb\u003e5\u003c/b\u003e, 33\u0026ndash;37 (1991).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin, B. et al. Multipulsed Magmatism and Duration of the Hydrothermal System of the Giant Jiama Porphyry Cu System, Tibet, China. \u003cem\u003eEcon. Geol.\u003c/em\u003e \u003cb\u003e119\u003c/b\u003e (1), 201\u0026ndash;217 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, Z. H. et al. Fractal and Multifractal Characteristics of Lineaments in the Qianhe Graben and Its Tectonic Significance Using Remote Sensing Images. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e (4), 587 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, H. et al. Fractal dimension used as a proxy to understand the spatial distribution for carlin-type gold deposits. \u003cem\u003eOre Geol. Rev.\u003c/em\u003e \u003cb\u003e158\u003c/b\u003e, 105534 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, Y. et al. Fractal-Based Pattern Quantification of Mineral Grains: A Case Study of Yichun Rare-Metal Granite. \u003cem\u003eFractal Fract.\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e (1), 49 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLyu, C., Cheng, Q. M., Zuo, R. G. \u0026amp; Wang, X. P. Mapping Spatial Distribution Characteristics of Lineaments Extracted from Remote Sensing Image Using Fractal and Multifractal Models. \u003cem\u003eJ. Earth Sci.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e, 507\u0026ndash;515 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaghsoudi Moud, F., van Ruitenbeek, F., Hewson, R. \u0026amp; van der Meijde, M. An Approach to Accuracy Assessment of ASTER Derived Mineral Maps. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e (13), 2499 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMandelbrot, B. B. \u003cem\u003eThe Fractal Geometry of Nature\u003c/em\u003epp. 17\u0026ndash;179 (WH Freeman, 1983).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMbianya, G. N. et al. Remote sensing satellite-based structural/alteration mapping for gold exploration in the Kett\u0026eacute; goldfield, Eastern Cameroon. \u003cem\u003eJ. Afr. Earth Sc.\u003c/em\u003e \u003cb\u003e184\u003c/b\u003e, 104386 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiao, L. D., Qi, H. R. \u0026amp; Szu, H. A maximum entropy approach to unsupervised mixed-pixel decomposition. \u003cem\u003eIEEE Trans. Image Process.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e (4), 1008\u0026ndash;1021 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNabatian, G., Songjian, A., Pour, A. B., Abdollahi, F. \u0026amp; Habashi, J. Integration of ASTER imagery and field data for chromite exploration in the Eastern Khoy Ophiolite Complex, NW Iran. \u003cem\u003eEarth Sci. Inf.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e (2), 245 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNaceur, M. S., Loghmari, M. A. \u0026amp; Boussema, M. R. The contribution of the sources separation method in the decomposition of mixed pixels. \u003cem\u003eIEEE Trans. Geosci. Remote Sens.\u003c/em\u003e \u003cb\u003e42\u003c/b\u003e (11), 2642\u0026ndash;2653 (2004).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNascimento, J. M. P. \u0026amp; Dias, J. M. B. Vertex component analysis: a fast algorithm to unmix hyper-spectral data. \u003cem\u003eIEEE Trans. Geosci. Rem. Sens.\u003c/em\u003e \u003cb\u003e43\u003c/b\u003e (4), 898\u0026ndash;910 (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNi, C. Z., Zhang, S. T., Chen, Z., Yan, Y. F. \u0026amp; Li, Y. J. Mapping the spatial distribution and characteristics of lineaments using fractal and multifractal models: A case study from Northeastern Yunnan Province, China. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e (1), 10511 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQiu, J. T., Zhang, C. \u0026amp; Hu, X. Integration of Concentration-Area Fractal Modeling and Spectral Angle Mapper for Ferric Iron Alteration Mapping and Uranium Exploration in the Xiemisitan Area, NW China. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e (10), 13878\u0026ndash;13894 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRen, Z. L., Zhai, Q. P. \u0026amp; Sun, L. A Novel Method for Hyperspectral Mineral Mapping Based on Clustering-Matching and Nonnegative Matrix Factorization. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e (4), 1042 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSharifi, A. \u0026amp; Hosseingholizadeh, M. H. Validation of extracted endmembers from hyperspectral images. \u003cem\u003eInt. Archives Photogrammetry Remote Sens. Spat. Inform. Sci.\u003c/em\u003e \u003cb\u003e42\u003c/b\u003e, 989\u0026ndash;992 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShi, R. \u0026amp; Xiao, H. G. Particle Flow Simulation of the Mechanical Properties and Fracture Behavior of Multi-Mineral Rock Models with Different Fractal Dimensions. \u003cem\u003eFractal Fract.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e (1), 13 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSikakwe, G. U. Mineral exploration employing drones, contemporary geological satellite remote sensing and geographical information system (GIS) procedures: A review. \u003cem\u003eRemote Sens. Applications: Soc. Environ.\u003c/em\u003e \u003cb\u003e31\u003c/b\u003e, 100988 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun, T. et al. Fractal and multifractal characteristics of regional fractures in Tongling metallogenic Area. \u003cem\u003eNonferrous Met. Eng.\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e, 111\u0026ndash;115 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun, T. et al. Fractal-Based Multi-Criteria Feature Selection to Enhance Predictive Capability of AI-Driven Mineral Prospectivity Mapping. \u003cem\u003eFractal Fract.\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e (4), 224 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThompson, D. R., Mandrake, L., Gilmore, M. S. \u0026amp; Castano, R. Superpixel endmember detection. \u003cem\u003eIEEE Trans. Geosci. Remote Sens.\u003c/em\u003e \u003cb\u003e48\u003c/b\u003e (11), 4023\u0026ndash;4033 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang, C., Jia, H. K., Dong, L. F., Zhao, H. S. \u0026amp; Zhao, M. H. Selection of Landsat-8 Operational Land Imager (OLI) Optimal Band Combinations for Mapping Alteration Zones. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e (2), 392 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang, S. H. \u0026amp; Tian, S. F. Extraction of Alteration Information from Hyperspectral Data Base on Kernel Extreme Learning Machine. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e (19), 3646 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang, D., Shao, H., Chang, H., Zhang, R. \u0026amp; Ding, B. Landsat 8 and ASTER remote sensing data for ore prospecting in the postmasburg manganese ore field, South Africa. \u003cem\u003eEarth Sci. Inf.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e (1), 1\u0026ndash;16 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYao, F. J. et al. Review on the development of multi- and hyperspectral remote sensing technology for exploration of copper-gold deposits. \u003cem\u003eOre Geol. Rev.\u003c/em\u003e \u003cb\u003e162\u003c/b\u003e, 105732 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYuan, C. J., Zhao, J., Liu, Z. B., Liu, X. F. \u0026amp; Tang, J. Detecting mineralization via white mica spectral footprints: From field-based sampling to satellite hyperspectral remote sensing. \u003cem\u003eOre Geol. Rev.\u003c/em\u003e \u003cb\u003e181\u003c/b\u003e, 106588 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, H. C., Zhang, Y. N. \u0026amp; Duan, F. A robust and fast geometry based unmixing algorithm for hyperspectral imagery. \u003cem\u003eInt. J. Remote Sens.\u003c/em\u003e \u003cb\u003e31\u003c/b\u003e (20), 5481\u0026ndash;5493 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, G. et al. Comprehensive Multi-Source remote sensing data integration for enhanced mineralization alteration extraction and geological structure interpretation in the Lala region of Sichuan Province. \u003cem\u003eOre Geol. Rev.\u003c/em\u003e \u003cb\u003e168\u003c/b\u003e, 106032 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao, H. Q. \u0026amp; Zhao, X. S. Nonlinear unmixing of minerals based on the log and continuum removal model. \u003cem\u003eEur. J. Remote Sens.\u003c/em\u003e \u003cb\u003e52\u003c/b\u003e (1), 277\u0026ndash;293 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao, F. W., Hu, J. H., Yang, Y. N., Xiao, H. X. \u0026amp; Ma, F. C. Cross-Scale Study on Lime Modified Phosphogypsum Cemented Backfill by Fractal Theory. \u003cem\u003eMinerals\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e (4), 403 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao, Z., Qiao, K., Liu, Y. W., Chen, J. \u0026amp; Li, C. L. Geochemical Data Mining by Integrated Multivariate Component Data Analysis: The Heilongjiang Duobaoshan Area (China) Case Study. \u003cem\u003eMinerals\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e (8), 1035 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao, Z. H. et al. Spatial Analysis of Structure and Metal Mineralization Based on Fractal Theory and Fry Analysis: A Case Study in Nenjiang\u0026thinsp;\u0026ndash;\u0026thinsp;Heihe. \u003cem\u003eMetallogenic Belt Minerals\u003c/em\u003e. \u003cb\u003e13\u003c/b\u003e (3), 313 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao, Q., Yan, Y., Tonai, S., Dilek, Y. \u0026amp; Zhu, Z. F. Timing of India-Asia collision and significant coupling between them around 51 Ma: Insights from the activation history of the Zhongba-Gyangze thrust in southern Tibet. \u003cem\u003eGeology\u003c/em\u003e \u003cb\u003e52\u003c/b\u003e (1), 61\u0026ndash;66 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZuo, R. G. \u0026amp; Wang, J. L. ArcFractal: An ArcGIS add-in for processing geoscience data using fractal/ multifractal models. \u003cem\u003eNat. Resour. Res.\u003c/em\u003e \u003cb\u003e29\u003c/b\u003e, 3\u0026ndash;12 (2020).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Tibet, Sentinel-2, spectral mixture analysis, Fractal Theory, Alteration Information","lastPublishedDoi":"10.21203/rs.3.rs-6598339/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6598339/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAs a key target area in China's new round of strategic mineral exploration initiatives, Tibet possesses favorable metallogenic conditions shaped by its unique geological evolution and tectonic setting. In this paper, the Saga region of Tibet is the research object, and Level-2A Sentinel-2 imagery is utilized. By applying mixed pixel decomposition, interfering endmembers were identified, and spectral unmixing and reconstruction were performed, effectively avoiding the drawback of traditional methods that tend to remove mineral alteration signals and masking interference. Combined with band ratio analysis and principal component analysis (PCA), various types of remote sensing alteration anomalies in the region were extracted. Furthermore, the fractal box-counting method was employed to quantify the fractal dimensions of the different alteration anomalies, thereby delineating their spatial distribution and fractal structural characteristics. Based on these results, two prospective mineralization zones were identified. The results indicate that:(1) In areas of Tibet with low vegetation cover, applying spectral mixture analysis (SMA) effectively removes substantial background interference, thereby enabling the extraction of subtle remote sensing alteration anomalies. (2) The fractal dimensions of various remote sensing alteration anomalies were calculated using the fractal box-counting method over a spatial scale range of 0.765 to 6.123 km. These values quantitatively characterize the spatial fractal properties of the anomalies, and the differences in fractal dimensions among alteration types reflect the spatiotemporal heterogeneity of the mineralization system. (3) The high-potential mineralization zones identified in the composite contour map of fractal dimensions of alteration anomalies show strong spatial agreement with known mineralization sites. Additionally, two new prospective mineralization zones were delineated in their periphery, providing theoretical support and exploration targets for future prospecting in the study area.\u003c/p\u003e","manuscriptTitle":"Extraction of Remote Sensing Alteration Information Based on Integrated Spectral Mixture Analysis and Fractal Analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-13 18:43:11","doi":"10.21203/rs.3.rs-6598339/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"56371446-d3ec-4a95-a68b-3058e1a42864","owner":[],"postedDate":"May 13th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":48357532,"name":"Earth and environmental sciences/Solid earth sciences"},{"id":48357533,"name":"Earth and environmental sciences/Space physics"},{"id":48357534,"name":"Physical sciences/Engineering"}],"tags":[],"updatedAt":"2025-06-06T11:23:12+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-13 18:43:11","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6598339","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6598339","identity":"rs-6598339","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}