Next Article in Journal
SAR-HUB: Pre-Training, Fine-Tuning, and Explaining
Previous Article in Journal
Spaceborne SAR Time-Series Images Change Detection Based on SAR-SIFT-Logarithm Background Subtraction
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Remote Sensing
Volume 15
Issue 23
10.3390/rs15235532
remotesensing-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Application of a Multifractal Model for Identification of Lithology and Hydrothermal Alteration in the Dasuji Porphyry Mo Deposit in Inner Mongolia, China

by
Mingjie Xi
1,2,3,
Wanchang Zhang
2,
Jiakui Tang
3,4,*,
Huiran Gao
5 and
Masoud Jafari Shalamzari
2,6
1
Development and Research Center of China Geological Survey, Beijing 100037, China
2
Key Laboratory of Digital Earth Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China
3
University of Chinese Academy of Sciences, Beijing 100049, China
4
Yanshan Earth Key Zone and Surface Flux Observation and Research Station, University of Chinese Academy of Sciences, Beijing 101408, China
5
National Institute of Natural Hazards, Ministry of Emergency Management of China, Beijing 100085, China
6
Iran’s Department of Environment, Tabas Branch, Tabas 9791735618, Iran
*
Author to whom correspondence should be addressed.
Remote Sens. 2023, 15(23), 5532; https://doi.org/10.3390/rs15235532
Submission received: 28 October 2023 / Revised: 10 November 2023 / Accepted: 12 November 2023 / Published: 28 November 2023
(This article belongs to the Section Remote Sensing in Geology, Geomorphology and Hydrology)
Browse Figures
Versions Notes

Abstract

:
The Dasuji giant porphyry molybdenum deposit is one of the largest ore deposits recently discovered along the Yinshan–Yanshan–Liaoning molybdenum belt in China. Using this deposit along the Yinshan–Yanshan–Liaoning molybdenum belt as the study area, the present study proposed a two-stage approach aimed at marking out the hydrothermally altered anomalies in the study area for the guidance of future prospecting in other regions. First of all, the Principal Component Analysis (PCA) and specific Band Ratio methods were applied to the ASTER images from different acquisition dates to extract ferric oxides and hydroxyl alterations related to the porphyry molybdenum deposit. Then, the Fractal-Aided Anomaly-Overlaying Selection model was adopted to recognize two ferric and hydroxyl alteration layers for separating anomalies from the interferences caused by geology and random noise from the data. Furthermore, for lithological differentiation in the previously marked off area, the Random Forest Classifier (RFC) was applied to the composite data obtained via the ASTER, ETM, and DEM, and it is demonstrated that the DEM can significantly improve lithological mapping in areas with complex vegetation cover and topography. Based on field verification and comparison with geological maps, the research revealed that the suggested two-stage approach may effectively reduce erroneously recognized anomalies produced during the first stage while retaining ore-related anomalies for gigantic porphyry molybdenum deposit prospecting in the Dasuji area, which showed the good application potential of the proposed model to extract actual hydrothermally altered anomalies adopted for lithological discrimination and mapping.

1. Introduction

As a well-known, cost-effective and viable technology, remote sensing has been widely employed in a variety of mining applications, including identifying lithologies and mapping hydrothermal alteration zones in metallogenic provinces (e.g., [1,2,3,4,5,6,7,8,9,10,11]). The variation in the reflectance or emittance radiating from materials with respect to the wavelength is used to delineate lithologies and map hydrothermal alteration zones. They distinguish between hydrothermally altered and unaltered rocks in the visible and near-infrared (0.3–2.5 µm) bands [12,13]. As a result, spectral characteristics have been widely used to pinpoint hydrothermal alteration zones using both spaceborne and aerial images for decades (e.g., [3,14,15,16,17,18,19]). Since the 1970s, various approaches for distinguishing hydrothermal alteration zones based on the spectral properties of minerals or rocks have been proposed, such as Principal Component Analysis (PCA) [1,20], Band Ratio (BR) [20], Spectral Angle Mapper (SAM) [2,6] and Constrained Energy Minimization (CEM) [21,22,23]. As a multivariate statistical technique, PCA has been used to eliminate redundancy in multispectral data and produce uncorrelated components. This method has become one of the most commonly used techniques in mineral extraction and exploration (e.g., [16,24,25,26,27]). Furthermore, lithological mapping by employing remote sensing data is commonly used in arid–semiarid regions as an economical and effective approach for geological inquiry (e.g., [28,29,30]) thanks to the alleviated influence of the less vegetated environments in such regions.
Multispectral data are more extensively used in geological explorations since they are less expensive and more accessible than hyperspectral data (e.g., [3,4,5,25,31]). However, due to the coarse spectral resolution of multispectral images, spectral differences between rocks may be greatly attenuated, and the different characteristics of spectral signatures may refer to the same mineral composition, and vice versa. Furthermore, even though the same processing methods are applied, multispectral satellite images acquired at various times may produce different results for a variety of reasons [32]. In addition, the interferences of extensive regolith, carmeloite, and vegetation coverage pose huge obstacles to differentiating lithological units and hydrothermal alteration zones from remotely sensed images in the research area. Thus, separating regional anomalies or extracting lithological units and hydrothermal alteration zones from multispectral imagery has been a difficult challenge in terms of the interpretation and application of remote sensing data in geology.
Fractal/multifractal theory has been extensively applied in geoscience [33,34,35,36,37]. It is thought of as an effective approach to separate geological anomalies from complicated backgrounds [38,39]. Fractal geometry describes irregular objects that Euclidian geometry cannot [40,41]. A multifractal is a spatially interwoven fractal with a continuous spectrum of fractal dimensions (continuous multifractals) or discrete spectra (discrete multifractals) [42,43,44]. Fractal/multifractal models, such as the Number–Size (N–Z) model [45,46], the Concentration–Area (C–A) model [47], the Spectrum–Area (S–A) model [48,49], the Concentration–Distance (C–D) model [50], the Concentration-Volume model [51,52], and the Singularity Index [53,54,55], have been frequently proved robust for separating geological anomalies, such as geochemical, geophysical and remote sensing anomalies from a complicated background. As an efficient classification method, the Concentration–Area (C–A) fractal model was proved to be useful in determining the anomaly threshold in geological exploration. PCA and the C–A fractal model can be combined to identify hydrothermal alteration zones from remotely sensed imageries (e.g., [15,25,31,56]).
In this research, a hybrid method, called the Fractal-Aided Anomaly-Overlaying Selection (FAAOS) model, is proposed based-on ASTER data for hydrothermal anomaly selection. In this model, the C–A fractal model is used to determine the anomaly threshold of hydrothermal alterations obtained using PCA and BR methods. Likewise, the anomalies at different acquisition dates are overlaid to eliminate interfering false anomalies. A detailed description of the data processing for the hydrothermal alteration mapping of the Dasuji molybdenum deposit in Inner Mongolia is presented at first. Then, the capability of satellite imagery from ETM+ and ASTER in association with DEM is confirmed for lithological discrimination using the Random Forest Classifier in the regolith covered area. The results of the mapping procedure are eventually verified using the geological maps and fieldwork.

2. Geological Setting

Since the Precambrian period, the North China Craton (NCC) area, which is a broad term for the Chinese portion of the Sino–Korea Craton, and its nearby areas have endured a long and complex tectonic–magmatic evolutionary history [57,58,59]. This evolution has been responsible for the abundance of mineral resources found in this region [60,61]. As shown in Figure 1a, a number of porphyry deposits with molybdenum as the principal metal resource have been reported along the region’s southern and northern edges. The Mo deposits in the northern NCC are magmatic–hypothermal systems, with similar wall rock alterations from inside out, including hypothermal assemblages (potassic alteration and silicification), mesothermal (phyllic and potassic alteration), and low temperature (carbonate and argillic) [62]. Furthermore, the EW-trending Yinshan–Yanshan–Liaoning Mo belt (YYLMB), China’s third biggest Mo province, was discovered at the northern boundary of the NCC [63].
The Dasuji porphyry Mo deposit, which was recently identified in the western YYLMB, is one of the largest deposits on the NCC’s northern boundary. The deposit is located in Zhuozi County, Inner Mongolia, spreading geographically between 112°42′30″~112°44′00″E and 40°43′00″~40°44′30″N. As illustrated by the simplified geological map of the study area in Figure 1b, the deposit is located in the southwest of the region. More than half of the study area is covered by deeply weathered regolith of Quaternary sediments and Cenozoic carmeloite, with the remainder mostly underlain by a sequence of Archean to Mesozoic acid intrusive rocks.
As shown in Figure 1c, the exposed formations in the Dasuji deposit are Precambrian metamorphic rocks, Mesozoic magmatic rocks, Neoproterozoic mafic dykes and Quaternary unconsolidated sediments. The Precambrian metamorphic rocks are composed of gneiss, migmatite, and granulite from the Jining group, which are the exposed base strata in the form of xenoliths and cataclastic plagiogranite. The Mesozoic magmatic rocks are mainly composed of granite porphyry, quartz porphyry, syenogranite, and diabase, which are the primary ore-bearing granitoid intrusions. Molybdenum mineralization in both Archean metamorphic rocks and granitoid porphyry stocks occurs mainly as veins, veinlets and disseminated blocks [65]. The oxidized ore minerals in the deposit mainly include molybdite and limonite, with molybdenite and pyrite being as the sulfide ore minerals. The gangue minerals mainly include quartz, plagioclase, kaolinite, orthoclase, sericite, and epidote.
The wall rock alteration in the deposit is intense and widespread, primarily occurring in the quartz porphyry and syenogranite. The main types of hydrothermal alteration include silicification, sericitization, k-feldspathization, biotitization, chloritization, uralitization, and carbonatization, among which silicification and sericitization are well developed and have a close relationship with molybdenum mineralization [65]. Overlay occurs in each step of the hydrothermal alteration in space, while zonation has not yet been developed.

3. Materials and Methods

3.1. Image Data and Data Preprocessing

The ASTER is an advanced multispectral sensor that was launched on board the Terra spacecraft in December 1999. It covers a broad spectral range with 14 spectral bands, including three visible and near-infrared (VNIR) bands with 15 m spatial resolution, six shortwave infrared (SWIR) bands with 30 m spatial resolution, and five thermal infrared (TIR) bands with 90 m spatial resolution. In addition, a rearview stereoscopic telescope is used in the near-infrared spectral band (band 3B) [66]. In this study, two ASTER level-1B scenes geometrically corrected and respectively acquired on 23 March 2003 and 22 March 2006 (both collections in spring and cloud-free) were selected and downloaded from the Earth Explorer platform (http://glovis.usgs.gov/ (accessed on 10 May 2019)).
The VNIR and SWIR bands of the ASTER images were imported to and calibrated in the ENVI 5.2 software package. All the VNIR and SWIR bands of each image were bicubically resampled to a 30 m cell size, and then stacked to form a 9-band image. Afterwards, atmosphere correction was carried out using the FLAASH (Fast Line-of sight Atmospherics Analysis to Hypercubes) function to eliminate the atmospheric effects and to convert the digital numbers (DNs) of the image to surface reflectance. A cross-track illumination correction was applied to the ASTER images to remove the effects of energy overspill from band 4 into bands 5 and 9. The resulting image was projected to Universal Transverse Mercator (UTM) zone 49 N using the World Geodetic System 1984 (WGS-84) datum. Moreover, a Landsat 7 (ETM+) and a Digital Elevation Model (DEM) scene with 30 m spatial resolution were obtained from the USGS (US Geological Survey, SRTM1 dataset).
In the study area, the VNIR and SWIR spectral bands of the ASTER are supposedly capable of extracting the main hydrothermal alteration minerals. Hydrothermal alteration in the Dasuji deposit is well developed, mainly including biotitization, chloritization, phyllic (sericitization), and argillic alteration. In order to detect these latter hydrothermal alterations from satellite data, the spectral signatures of the corresponding altered minerals were used as the reference to find suitable analogue data. Here, we used the laboratory spectral signatures of the minerals from the USGS mineral spectral library available in the ENVI 5.2 software package.
Phyllic minerals (e.g., sericite, muscovite), argillic minerals (e.g., kaolinite, alunite), and iron oxide minerals (e.g., limonite, goethite) are common hydrothermal alteration minerals exposed in porphyry deposits. In the visible to shortwave infrared (VNIR–SWIR) wavelengths, the hydroxyl (OH) alteration (phyllic and argillic minerals) exhibits a robust diagnostic spectral absorption feature at 2.2 µm (6th ASTER band) and strong reflectance features at 2.27 µm (7th ASTER band) due to the absorption and reflectance features of the Al–OH bond. In addition, the argillic minerals also exhibit the Al–OH absorption features at 2.17 µm (5th ASTER band) [3,13], as shown in Figure 2a.
Ferric oxide is the common constituent of the alteration zones associated with hydrothermal sulfide deposits. Ferric iron (Fe3+) has a high absorption ratio at approximately 0.65 and 0.87 µm [67]. Due to Fe–OH and OH stretching and bending in the SWIR wavelength, jarosite displays diagnostic vibrational absorption features at near 1.6 µm (4th ASTER band) [13]. Goethite and limonite display a broad ferric iron absorption feature near 0.9 µm and two additional features near 0.5 µm (1th ASTER band) and 0.66 µm (2th ASTER band), with a peak near 0.75 µm (3th ASTER band) [13], as shown in Figure 2b.

3.2. Methods

3.2.1. Concentration–Area Model

The Concentration–Area (C–A) model is one of the most widely applied multifractal models in geoscience (e.g., [15,38,47,68]). It was developed by Cheng et al. [38] for separating geochemically anomalous areas from the background. This model separates anomalies from complex geological patterns via a power-law function of their area in the spatial domain. The fractal relationship between the concentration (ρ) and the area (A) can be expressed as follows:
A ( ρ ≥ υ ) ∝ ρ − β .
where A(ρ) is the area enclosed by contours with values greater than or equal to a certain threshold υ, and β denotes the fractal dimension of the C–A model [38,47]. The C–A fractal model has been proven useful in remote sensing classification [47,69], and recently, it has been successfully employed in extracting anomalies from remote sensing imagery (e.g., [15,25,31,47]). A log–log plot of A(ρ) against ρ can develop a multifractal form consisting of a series of straight lines or segments. Each segment represents a population (e.g., background, lowly intensive anomalies and highly intensive anomalies or different rock units), which is self-similar or self-affine for certain intervals of ρ values.

3.2.2. Band Ratio

The Band Ratio is perhaps the most popular and simplest method for enhancing spectral features in geological mapping and mineral exploration using remote sensing data (e.g., [30,70,71,72,73,74]). This method is effective in highlighting certain features that would not be easily noticeable in a single band. Based on the spectral characteristics of phyllic and argillic alterations in the deposit area, the Band Ratio of (band 4 * 3)/(band 5 + band 6 + band 7) (corresponding to the widely used TM5/TM7) [75] was utilized to delineate hydroxyl alterations using ASTER data in this research.

3.2.3. Principal Component Analysis

Principal Component Analysis (PCA) is a multivariate technique extensively utilized in geosciences (e.g., [24,26,31,56,76,77]). PCA calculates the principal component images based on the relationships between the spectral responses of the target materials and numeric values extracted from the eigenvector matrix [1,76,77]. PCA can effectively decorrelate multi-variable images and concentrate the maximum information into a few uncorrelated principal components, allowing identification of the principal components (PCs) containing spectral information about the target minerals. In this research, PCA was performed on ASTER bands 1–9 to extract ferric oxide alterations.

3.2.4. Fractal-Aided Anomaly-Overlaying Model

The anomaly-overlaying selection method proposed by Liu et al. [32] suggests that the real alteration anomalies exist in the selected PCA obtained images with different acquisition dates. Simultaneously, the noise interference-caused false anomalies are randomly distributed in each individual image and do not coincide with others [32,78]. The utilization of the anomaly-overlaying selection method can eliminate random false anomalies caused by the data noise [78]. However, due to the complex geological background, the disturbances of large and deep regoliths of Quaternary and Cenozoic carmeloite may weaken the alteration anomalies derived from remote sensing data. In addition, uncertainty may arise where the surrounding rocks contain the same minerals as the altered rocks. For example, granites that contain muscovite may be confused with phyllic-altered rocks. As a result, distinguishing the altered minerals from their equivalents in country rocks is a critical aspect of the investigation [79].
Based on the multifractal model theory, anomalies (altered rocks) and backgrounds (surrounding rocks) belong to different populations and power-law relationships. Hence, multifractal models can be used to provide a visual representation of the variance in an image based on its pixel values and its frequency distribution as well as the spatial and geometrical properties of the image patterns [54]. In this study, instead of using the mean value and standard deviation of the relevant principle components proposed by Liu et al., the C–A fractal model was used to determine the threshold of anomalies, taking into account the different self-similar relationships between altered minerals and their analogues in country rocks [32].
Combining the advantages of conventional image processing approaches (PCA and BR), the anomaly-overlaying selection method and the C–A model, the Fractal-Aided Anomaly-Overlaying Selection model was proposed and applied in order to separate real alteration anomalies from false anomalies caused by data noise for a complicated geological background. The method is comprised of three steps, as follows:
  • Applying the PCA and specific Band Ratio (band 4 * 3)/(band 5 + band 6 + band 7) to the ASTER images from different acquisition dates to extract ferric oxides and hydroxyl alteration.
  • Applying the C–A model to each resulting alteration image to separate the alteration anomalies from the complicated geological background in ENVI software.
  • Applying the anomaly-overlaying selection method to the resulting anomaly layers to eliminate the random interference-caused false anomalies. In this step, a pixel is classified as a real anomaly if, and only if, it exists as an anomaly in both anomaly layers.

3.2.5. Random Forest Classifier

The Random Forest Classifier (RFC),as a favorable and efficient classifier, has been widely applied in recent decades in many fields, such as geology [80,81] and ecology [82,83,84]. The RFC method is a supervised machine learning approach that contains multiple decision trees and combines the predictions from all the trees [82]. With this classifier, each decision tree offers a classification and the random forest decides which class an object is to be classified into according to a majority rule: the class with the highest votes across all the trees is determined as the final class [85]. In present study, the RFC algorithm was adopted for lithological discrimination and mapping based on a combination of ASTER, ETM and DEM data.

3.2.6. Fieldwork Verification

The hydrothermal alteration recognition and lithological mapping in this study were verified using geological maps and field reconnaissance results. The mapping results were compared with the exposed rocks and mining regions on the geological maps. Field reconnaissance was carried out two times during June to July 2016 and December 2017. Geological locations were measured using a portable GPS with an average accuracy of less than 10 m. Lithological units and hydrothermally altered areas from the image processing results were carefully checked on site. Field photographs of the lithological units, hydrothermally altered rocks and ore rocks have been taken to obtain comprehensive information about the study area. Samples were collected from the open-pit quarry of the Dasuji mines and surrounding areas for analysis.

3.2.7. The Total Research Flowchart

Mineralization alteration information and lithology distribution can be mapped from remote sensing data [86]. The integration of alteration information and lithology information is conducive to delineating prospecting potential area [87]. Reported research [88] has shown that both ASTER and ETM+ data can be used to perform lithological identification and mapping. However, combining ASTER and ETM+ data for lithology distribution can obtain a higher identification accuracy than using either data alone, indicating that the two kinds of data are complementary in terms of the spectral characteristics of lithology. Therefore, in this study, the ASTER was solely chosen for remote sensing extraction of alteration anomaly information, although the strategy of combining ASTER and ETM data was chosen for the lithology identification. Moreover, the influence of DEM data applied in the C–A model on lithology mapping was further compared. Please see the research flowchart in Figure 3.

4. Results

4.1. Principal Component Analysis and Band Ratio

Figure 4 displays the first four results of the PCA on nine bands of two multi-temporal ASTER images for mapping the alteration minerals. In the results of the ASTER images, the PC1 components in two PC images account for more than 80% of the variance among the nine bands, which mainly provide information about albedo and topography. PC2 probably describes the difference between visible channels and infrared channels. PC3 is related to enhancing ferric oxide minerals as bright pixels because of the positive contribution from band 4 and the negative contributions from band 1 and band 2, respectively [89]. PC4 shows ferric oxides as bright pixels with opposite signs in bands 4 and 3 [90,91]. Due to the limited fractions (lower than 0.88% or 0.59%), the rest of the PCs produce only noise, which cannot provide useful information for hydrothermal alteration mapping.
Figure 5a,b illustrate the results of ferric oxides extracted as bright pixels in the study area. There are no obvious bright pixels that exist in both images, indicating no ferric oxides alteration in the PC3. Figure 5c–f show the four respective results of alteration (ferric and hydroxyl minerals) obtained via PCA and the specific Band Ratio (band 4 * 3)/(band 5 + band 6 + band 7) in the study area, respectively. The primary distributions of the multi-temporal ferric and hydroxyl alteration mapping results (Figure 5c,d) in Figure 5e,f are similar while concentrating around the Dasuji deposit. However, there are still many punctate pixels that exist as alterations only in one date’s image because of the interferences of random noise caused by the instability of the satellite data themselves.

4.2. Fractal-Aided Anomaly-Overlaying Selection Model

Firstly, based on the raster map of the hydrothermal alteration, the C–A plots consisting of the concentrations of alterations (ρ, see Section 3.2.1 for details) versus the number of cells with concentrations of alterations greater than or equal to ρ were obtained (Figure 6). Four straight lines (Figure 6a,d) or three straight lines (Figure 6b,c) can be fitted to the C–A relationship using the LS method. Each straight line is an indicator of the population (e.g., anomalies or background). The left-hand side of the green line represents the background, and the right side indicates the anomalies.
Secondly, according to the C–A relationship of alteration, anomaly images for ferric and hydroxyl alteration were determined using the threshold values: (1) the threshold values of ferric oxides were 89 and 143 for images obtained from the ASTER acquired on 23 March 2003 and 22 March 2006 via the PCA method, respectively; and (2) the threshold values of hydroxyl alteration were 1.385 and 1.36 for images derived from the ASTER received on 23 March 2003 and 22 March 2006 via the Band Ratio (BR), respectively. The ferric anomaly images obtained from the ASTER are given in Figure 7a,b, and the hydroxyl anomaly images are given in Figure 8a,b. The anomalies mainly concentrate in the mine pit. However, due to the interference of random noise, there are many punctate anomalies in both images, albeit with different locations, so they can be eliminated via overlaying selection.
Thirdly, the two anomaly images were overlaid. The random anomalies, which only exist in one image, were considered to be noise interference-caused false anomalies and were excluded. Figure 7c and Figure 8c show the final extracted anomaly results of the ferric and hydroxyl alterations in the deposit. All the punctate random anomalies caused by noise were eliminated and the real anomalies were retained. The comparison between the alterations and resulting anomalies in Figure 7 and Figure 8 demonstrates that the overlaying selection is effective in removing the disturbances caused by topography and data instability.
Figure 9 presents the results of the ferric and hydroxyl alteration anomaly mapping superimposed on the ASTER band 1 image. The distribution of ferric alterations in the study area is shown as red pixels, whereas the hydroxyl alterations are in blue. Figure 9a exhibits the partly enlarged results of the anomalies in the deposit. The occurrences of the two kinds of anomalies are both mainly in the granites and potash–feldspar granites. In addition, the resulting anomalies show noticeable consistency with the molybdenum deposit.

4.3. Lithological Discrimination

Due to the large area of regolith cover in the study area, the lithological units are difficult to identify via visual interpretation in the original image, as shown in Figure 10. The field pictures also indicated that there are large areas of regolith and vegetation cover in the study area, which makes it hard to distinguish rock units from remote sensing images via visual interpretation. Figure 11a presents the C–A log–log plot of the DEM data of the study area, with each segment representing a type of rock, and Figure 11b shows the lithological population distribution by the DEM based on the C–A model. In spite of the large and deep coverage of regolith in the study area, the DEM data could distinguish between different lithologies because of the diverse topography caused by different weathering rates.
Figure 12a,b show lithological discrimination by the RFC using the combined data from the ETM + ASTER and ETM + ASTER + DEM, respectively. The mapping results in Figure 12b show a quite higher accuracy than the discrimination results presented in Figure 12a, which again demonstrates that the DEM plays an important role in lithological identification with the RFC method in the study area. The Quaternary, Cenozoic carmeloite and Archean granite in the study area were well discriminated using the RFC method due to significant differences in topography and texture. However, the Archean garnet leptite and garnet gneiss in the results were poorly distinguished because of the similar compositions in the field, as shown in Figure 12a,b. In addition, due to the deep and considerable coverage of regolith in the study area, discriminating the dykes from the images using the RFC method is difficult, which led to some errors in the final lithological classification. Moreover, due to the similar topography and texture, some ridges of carmeloite and granite were misclassified as norite, garnet leptite and garnet gneiss.

4.4. Field Validation and Analysis

Mixtures of ferric oxides and hydroxyl alterative mineral outcrops are shown in Figure 13 as typical examples from the study area. In Figure 13a,b, it is indicated that the Archaeozoic granulite is intruded by diabase dyke and granite, while Figure 13c,d show the hydroxyl (OH)-bearing alteration in the study area, Figure 13e,f show the ferritization, kaolinization and slight chloritization in the Archean garnet leptite and garnet granite gneiss, and Figure 13g,h show the ore rocks in the deposit. The field observations are consistent with the anomalies obtained via the ASTER, which directly support the mapping results using remote sensing.
Field observations were conducted to verify the predictive power of the proposed lithological mapping method. Figure 14 shows the results of the lithological analogy and verification. The Archean garnet leptite and garnet gneiss contain similar minerals such as garnet, feldspar and quartz, which makes their identification difficult, as shown in Figure 14a,b, corresponding to the lithological classification using the RFC method described above. Figure 14c,d show the Cenozoic carmeloite, which occupies a large area of this expedition site and was well recognized by the RFC method. In the field, the carmeloite was interfered with by vegetation and regolith, which makes it difficult to be correctly classified using a single dataset of ASTER + ETM. The biggest difference observed in the field between the Cenozoic carmeloite and Quaternary sediments is their topography, which also demonstrates the importance of the DEM in lithological discrimination when implementing the RFC technique.

5. Discussion

As previously stated, relying on the analysis of the hydrothermal alterations and the results of lithological mapping, it is obvious that the multifractal C–A model played a significant role in the alteration and lithological recognition in this study. Considering different minerals (such as altered and background minerals) and lithological units (such as Quaternary sediments and Cenozoic carmeloite) belonging to different populations, the application of the C–A model to alteration images and DEM data separates hydrothermal anomalies from the background and lithological units from each other. In addition, random punctate points caused by the instability of the satellite itself were efficiently eliminated by the overlaying method. The Fractal-Aided Anomaly-Overlaying Selection model effectively reduced the false anomalies caused by the geological background and satellite itself, which selected the hydrothermal anomalies associated with the porphyry molybdenum deposit, as shown in Figure 9.
The significant difference between the two lithological mapping approaches utilizing the RFC method can be explained by the DEM because of the diverse terrain of the different lithological units, as shown in Figure 11. Unlike semi-arid and arid regions [72,92], there are no significant color differences among the lithologies in the RGB compositions of multispectral data due to the interference of the large regolith and vegetation coverage, as displayed in Figure 10 and Figure 14. Therefore, it is difficult to distinguish between lithological units via visual interpretation in this study. Given the complex topography of the study area, the utilization of the DEM combined with multispectral images enabled us to classify lithological units by implementing the RFC method. However, due to similar terrain and compositions of the Archean garnet leptite and garnet gneiss, they were misinterpreted in the lithological results. The small scales and characterless terrain of almost all the dykes in the study area made them difficult to recognize. Likewise, the observed hydrothermal alterations in the field are mainly attributed to ferric iron and kaolinite alteration because of the interferences of severe weathering in the study area, as shown in Figure 13e,f. Other hydrothermal alterations, such as chlorite, gypsum, and pyrite, are all distributed in the Dasuji deposit mining area.
The multifractal Concentration–Area (C–A) model performed well in terms of the lithology and hydrothermal alteration extraction for the Dasuji porphyry Mo deposit based on these results. The C–A model can identify alteration anomalies within a region with a complex geological setting, topography and regolith or vegetation coverage when combined with the overlaying selection method from different dates’ remote sensing images. This indicates that the C–A model has the probability of separating alteration anomalies from the background and the overlaying selection method has the advantage of weakening the disturbances of large and deep regolith coverage. The proposed fusion of C–A model and the overlaying selection method showed the potential to discover new Mo deposits in this study area. The multifractal Concentration–Area (C–A) model [35] was initially proposed for separating geochemical anomalies from backgrounds with complicated geochemical surfaces and complex tectonic settings; therefore, the proposed method theoretically may be used for other relative deposits with alteration anomaly features, and more application research should be carried out with in-depth field validation in the future.

6. Conclusions

Identification of lithology units and hydrothermal alterations via remote sensing data is commonly used in arid–semiarid regions as an economical and effective approach for geological mining applications. However, for places with complex vegetation and terrain settings, the formations deeply covered with regolith might lead to serious uncertainty, so that it is difficult to appropriately differentiate lithological units from remotely sensed data using a simple band combination.
According to field observations, traditional hydrothermal alteration extraction methods such as PCA and the BR, when applied to ASTER data, can reveal the ferric oxides and hydroxyl alteration zones that form the exposed mining pit of the studied region. However, there are large areas covered with deep regolith or vegetation, which makes it hard to distinguish rock units and hydrothermal alterations from remote sensing data. Single PCA or the BR method does not work well in differentiating actual anomalies from the interference of the complex geological background.
In this study, a new model, the Fractal-Aided Anomaly-Overlaying Selection model, which is based on the different power-law relationships between altered rocks and country rocks, was proposed and applied to hydrothermal alteration extracted via PCA and the Band Ratio to successfully separate real anomalies from a complex geo-background. Satisfactory results in terms of lithological unit discrimination were obtained and verified, which shows the good application potential of the proposed model to extract actual hydrothermally altered anomalies adopted for lithological discrimination and mapping.
Furthermore, it was demonstrated in this study that the RFC method in conjunction with the DEM significantly improves lithological mapping in areas with complex vegetation cover and topography. Based on the field verification and comparison with geological maps, we also found that the RFC method applied to the data combination of ETM + ASTER + DEM performed better in discriminating the lithologies in the Dasuji area than the data combination without the DEM, which showed that the utilization of the DEM is helpful in classifying lithological units from remote sensing data by implementing the RFC method.
However, there were still some misinterpreted units in the lithological mapping results due to the similar terrain and multi-spectral indistinguishable compositions, such as Archean garnet leptite and garnet gneiss in the study region. In addition, almost all the dykes with small scales and characterless terrains were not successfully identified. All these misinterpretation or misidentification problems should be addressed in future research.

Author Contributions

Conceptualization: M.X. and W.Z.; methodology: J.T. and W.Z; validation: M.X. and H.G.; formal analysis: M.X.; investigation: M.X. and J.T.; data processing: M.X.; writing—original draft preparation: M.X.; writing—review and editing: M.J.S., W.Z. and J.T. All authors have read and agreed to the published version of the manuscript.

Funding

This work was jointly supported by the Science and Technology Fundamental Resources Investigation Program (Grant No. 2022FY101701 and Grant No. 2022FY100100) funded by Ministry of Science and Technology of the People’s Republic of China, the Key R&D and Transformation Program of Qinghai Province (Grant No. 2020-SF-C37) funded by Science and Technology Department of Qinghai Province, the geological survey of mineral resources in Heiying Mountain area, Inner Mongolia (Grant No. DD20160040) funded by the China Geological Survey, and the National Key Research and Development Program of China (Grant No. 2018YFE0208300) funded by Ministry of Science and Technology of the People’s Republic of China.

Data Availability Statement

Due to the nature of this research, the participants in this study did not agree to their geological map and field validation data being shared publicly, so the supporting data are not available. The remote sensing data are included.

Acknowledgments

The authors are grateful to Linhai Jing and Haifeng Ding for their constructive comments after reviewing the manuscript. Gratitude is extended to the United States Geological Survey (USGS) (https://glovis.usgs.gov/ (accessed on 10 May 2019)) for providing the ASTER data. We highly appreciate the comments from the anonymous reviewers and editors that helped to improve the quality and content of this manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Crósta, A.P.; Moore, J. Enhancement of Landsat Thematic Mapper Imagery for Residual Soil Mapping in Sw Minas Gerais State, Brazil- a Prospecting Case History in Greenstone Belt Terrain. In Proceedings of the Seventh Thematic Conference on Remote Sensing for Exploration Geology: Methods, Integration, Solutions, Calgary, AB, Canada, 2–6 October 1989. [Google Scholar]
  2. Debba, P.; Van Ruitenbeek, F.; Van Der Meer, F.; Carranza, E.; Stein, A. Optimal Field Sampling for Targeting Minerals Using Hyperspectral Data. Remote Sens. Environ. 2005, 99, 373–386. [Google Scholar] [CrossRef]
  3. Alimohammadi, M.; Alirezaei, S.; Kontak, D.J. Application of Aster Data for Exploration of Porphyry Copper Deposits: A Case Study of Daraloo–Sarmeshk Area, Southern Part of the Kerman Copper Belt, Iran. Ore Geol. Rev. 2015, 70, 290–304. [Google Scholar] [CrossRef]
  4. Carrino, T.A.; Crósta, A.P.; Toledo, C.L.B.; Silva, A.M.; Silva, J.L. Geology and Hydrothermal Alteration of the Chapi Chiara Prospect and Nearby Targets, Southern Peru, Using Aster Data and Reflectance Spectroscopy. Econ. Geol. 2015, 110, 73–90. [Google Scholar] [CrossRef]
  5. Amer, R.; El Mezayen, A.; Hasanein, M. Aster Spectral Analysis for Alteration Minerals Associated with Gold Mineralization. Ore Geol. Rev. 2016, 75, 239–251. [Google Scholar] [CrossRef]
  6. Ge, W.; Cheng, Q.; Jing, L.; Chen, Y.; Guo, X.; Ding, H.; Liu, Q. Mineral Mapping in the Western Kunlun Mountains Using Tiangong-1 Hyperspectral Imagery. In IOP Conference Series: Earth and Environmental Science; IOP Publishing: Bristol, UK, 2016. [Google Scholar]
  7. Salem, S.; El Sharkawi, M.; El-Alfy, Z.; Soliman, N.; Ahmed, S. Exploration of Gold Occurrences in Alteration Zones at Dungash District, Southeastern Desert of Egypt Using Aster Data and Geochemical Analyses. J. Afr. Earth Sci. 2016, 117, 389–400. [Google Scholar] [CrossRef]
  8. He, L.; Lyu, P.; He, Z.; Zhou, J.; Hui, B.; Ye, Y.; Hu, H.; Zeng, Y.; Xu, L. Identification of Radioactive Mineralized Lithology and Mineral Prospectivity Mapping Based on Remote Sensing in High-Latitude Regions: A Case Study on the Narsaq Region of Greenland. Minerals 2022, 12, 692. [Google Scholar] [CrossRef]
  9. Saha, S. Remote Sensing and Geographic Information System Applications in Hydrocarbon Exploration: A Review. J. Indian Soc. Remote Sens. 2022, 8, 1457–1475. [Google Scholar] [CrossRef]
  10. Chen, Q.; Zhao, Z.; Zhou, J.; Zhu, R.; Xia, J.; Sun, T.; Zhao, X.; Chao, J. ASTER and GF-5 Satellite Data for Mapping Hydrothermal Alteration Minerals in the Longtoushan Pb-Zn Deposit, SW China. Remote Sens. 2022, 14, 1253. [Google Scholar] [CrossRef]
  11. Ali, H.; Ghoneim, S. Satellite-based silica mapping as an essential mineral for clean energy transition: Remote sensing mineral exploration as a climate change adaptation approach. J. Afr. Earth Sci. 2022, 196, 104683. [Google Scholar] [CrossRef]
  12. Hunt, G.R. Near-Infrared (1.3–2.4) Μm Spectra of Alteration Minerals—Potential for Use in Remote Sensing. Geophysics 1979, 44, 1974–1986. [Google Scholar] [CrossRef]
  13. Hunt, G.R.; Ashley, R.P. Spectra of Altered Rocks in the Visible and near Infrared. Econ. Geol. 1979, 74, 1613–1629. [Google Scholar] [CrossRef]
  14. Kruse, F.A.; Boardman, J.W.; Huntington, J.F. Comparison of Airborne Hyperspectral Data and EO-1 Hyperion for Mineral Mapping. IEEE Trans. Geosci. Remote Sens. 2003, 41, 1388–1400. [Google Scholar] [CrossRef]
  15. Shahriari, H.; Ranjbar, H.; Honarmand, M. Image Segmentation for Hydrothermal Alteration Mapping Using PCA and Concentration–Area Fractal Model. Nat. Resour. Res. 2013, 22, 191–206. [Google Scholar] [CrossRef]
  16. Sojdehee, M.; Rasa, I.; Nezafati, N.; Abedini, M.V. Application of Spectral Analysis to Discriminate Hydrothermal Alteration Zones at Daralu Copper Deposit, SE Iran. Arab. J. Geosci. 2016, 9, 41. [Google Scholar] [CrossRef]
  17. Belgrano, T.; Diamond, L.; Novakovic, N.; Hewson, R.; Hecker, C.; Wolf, R.; de Doliwa Zieliński, L.; Kuhn, R.; Gilgen, S. Multispectral discrimination of spectrally similar hydrothermal minerals in mafic crust: A 5000 km2 ASTER alteration map of the Oman–UAE ophiolite. Remote Sens. Environ. 2022, 280, 113211. [Google Scholar] [CrossRef]
  18. Zuo, L.; Wang, G.; Carranza, E.; Zhai, D.; Pang, Z.; Cao, K.; Mou, N.; Huang, L. Short-Wavelength Infrared Spectral Analysis and 3D Vector Modeling for Deep Exploration in the Weilasituo Magmatic–Hydrothermal Li–Sn Polymetallic Deposit, Inner Mongolia, NE China. Nat. Resour. Res. 2022, 25, 1–33. [Google Scholar] [CrossRef]
  19. Mahanta, P.; Maiti, S. Remote Detection of Hydrothermal Alteration Zones Using WorldView-3 VNIR-SWIR Reflectance Data: A Study from Lawa Gold Mines, India. J. Indian Soc. Remote Sens. 2022, 50, 1979–1993. [Google Scholar] [CrossRef]
  20. Fraser, S.; Green, A. A Software Defoliant for Geological Analysis of Band Ratios. Int. J. Remote Sens. 1987, 8, 525–532. [Google Scholar] [CrossRef]
  21. Harsanyi, J.C. Detection and Classification of Subpixel Spectral Signatures in Hyperspectral Image Sequences; University of Maryland: College Park, MD, USA, 1994. [Google Scholar]
  22. Farrand, W.H.; Harsanyi, J.C. Mapping the Distribution of Mine Tailings in the Coeur D’alene River Valley, Idaho, through the Use of a Constrained Energy Minimization Technique. Remote Sens. Environ. 1997, 59, 64–76. [Google Scholar] [CrossRef]
  23. Samani, P.; Prizomwala, S. Rajawat AS. Assessing the mineral alteration in Ambaji–Deri Region (Northwestern India) using hyperspectral remote sensing. J. Indian Soc. Remote Sens. 2021, 49, 249–257. [Google Scholar] [CrossRef]
  24. Ranjbar, H.; Honarmand, M.; Moezifar, Z. Application of the Crosta Technique for Porphyry Copper Alteration Mapping, Using ETM+ Data in the Southern Part of the Iranian Volcanic Sedimentary Belt. J. Asian Earth Sci. 2004, 24, 237–243. [Google Scholar] [CrossRef]
  25. Ahmadfaraj, M.; Mirmohammadi, M.; Afzal, P. Application of Fractal Modeling and PCA Method for Hydrothermal Alteration Mapping in the Saveh Area (Central Iran) Based on Aster Multispectral Data. Int. J. Min. Geo-Eng. 2016, 50, 37–48. [Google Scholar]
  26. Masoumi, F.; Eslamkish, T.; Honarmand, M.; Abkar, A.A. A Comparative Study of Landsat-7 and Landsat-8 Data Using Image Processing Methods for Hydrothermal Alteration Mapping. Resour. Geol. 2017, 67, 72–88. [Google Scholar] [CrossRef]
  27. Mulja, T.; Heriawan, M. The Miwah high sulphidation epithermal Au–Ag deposit, Aceh, Indonesia: Dynamics of hydrothermal alteration and mineralisation interpreted from principal component analysis of lithogeochemical data. Ore Geol. Rev. 2022, 147, 104988. [Google Scholar] [CrossRef]
  28. Gad, S.; Kusky, T. Aster Spectral Ratioing for Lithological Mapping in the Arabian–Nubian Shield, the Neoproterozoic Wadi Kid Area, Sinai, Egypt. Gondwana Res. 2007, 11, 326–335. [Google Scholar] [CrossRef]
  29. Zhang, X.; Li, P. Lithological Mapping from Hyperspectral Data by Improved Use of Spectral Angle Mapper. Int. J. Appl. Earth Obs. Geoinf. 2014, 31, 95–109. [Google Scholar] [CrossRef]
  30. Pour, A.B.; Hashim, M.; Park, Y.; Hong, J.K. Mapping Alteration Mineral Zones and Lithological Units in Antarctic Regions Using Spectral Bands of Aster Remote Sensing Data. Geocarto Int. 2017, 33, 1281–1306. [Google Scholar] [CrossRef]
  31. Asl, R.A.; Afzal, P.; Adib, A.; Yasrebi, A.B. Application of Multifractal Modeling for the Identification of Alteration Zones and Major Faults Based on ETM+ Multispectral Data. Arab. J. Geosci. 2015, 8, 2997–3006. [Google Scholar] [CrossRef]
  32. Liu, L.; Zhou, J.; Jiang, D.; Zhuang, D.; Mansaray, L.R.; Hu, Z.; Ji, Z. Mineral Resources Prospecting by Synthetic Application of TM/ETM+, Quickbird and Hyperion Data in the Hatu Area, West Junggar, Xinjiang, China. Sci. Rep. 2016, 6, 21851. [Google Scholar] [CrossRef]
  33. Cheng, Q. The Perimeter-Area Fractal Model and Its Application to Geology. Math. Geol. 1995, 27, 69–82. [Google Scholar] [CrossRef]
  34. Wang, Z.; Cheng, Q.; Cao, L.; Xia, Q.; Chen, Z. Fractal Modelling of the Microstructure Property of Quartz Mylonite During Deformation Process. Math. Geol. 2007, 39, 53–68. [Google Scholar] [CrossRef]
  35. Zuo, R.; Cheng, Q.; Xia, Q.; Agterberg, F. Application of Fractal Models to Distinguish between Different Mineral Phases. Math. Geosci. 2009, 41, 71–80. [Google Scholar] [CrossRef]
  36. Zuo, R. Identifying Geochemical Anomalies Associated with Cu and Pb–Zn Skarn Mineralization Using Principal Component Analysis and Spectrum–Area Fractal Modeling in the Gangdese Belt, Tibet (China). J. Geochem. Explor. 2011, 111, 13–22. [Google Scholar] [CrossRef]
  37. Nazarpour, A.; Sadeghi, B.; Sadeghi, M. Application of Fractal Models to Characterization and Evaluation of Vertical Distribution of Geochemical Data in Zarshuran Gold Deposit, NW Iran. J. Geochem. Explor. 2015, 148, 60–70. [Google Scholar] [CrossRef]
  38. Cheng, Q.; Agterberg, F.; Ballantyne, S. The Separation of Geochemical Anomalies from Background by Fractal Methods. J. Geochem. Explor. 1994, 51, 109–130. [Google Scholar] [CrossRef]
  39. Mokhtari, Z.; Sadeghi, B. Geochemical anomaly definition using multifractal modeling, validated by geological field observations: Siah Jangal area, SE Iran. Geochemistry 2021, 4, 125774. [Google Scholar] [CrossRef]
  40. Mandelbrot, B.B. Fractals: Form, Chance and Dimension; WH Freeman & Co.: San Francisco, CA, USA, 1979; pp. 16, 365. [Google Scholar]
  41. Mandelbrot, B.B.; Pignoni, R. The Fractal Geometry of Nature; WH Freeman: New York, NY, USA, 1983. [Google Scholar]
  42. Evertsz, C.J. Multifractal Measures. Chaos Fractals New Front. Sci. 1992, 1992, 921–953. [Google Scholar]
  43. Cheng, Q. Discrete Multifractals. Math. Geol. 1997, 29, 245–266. [Google Scholar] [CrossRef]
  44. Mandelbrot, B.B.; Fisher, A.J.; Calvet, L.E.; A Multifractal Model of Asset Returns. No 1164, Cowles Foundation Discussion Papers, Cowles Foundation for Research in Economics, Yale University. 1997. Available online: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=78588 (accessed on 23 November 2022).
  45. Wang, Q.; Wan, L.; Zhang, Y.; Zhao, J.; Liu, H. Number-Average Size Model for Geological Systems and Its Application in Economic Geology. Nonlinear Process. Geophys. 2011, 18, 447. [Google Scholar] [CrossRef]
  46. Hashemi, M.; Afzal, P. Identification of Geochemical Anomalies by Using of Number–Size (N–S) Fractal Model in Bardaskan Area, NE Iran. Arab. J. Geosci. 2013, 6, 4785–4794. [Google Scholar] [CrossRef]
  47. Cheng, Q.; Li, Q. A Fractal Concentration–Area Method for Assigning a Color Palette for Image Representation. Comput. Geosci. 2002, 28, 567–575. [Google Scholar] [CrossRef]
  48. Cheng, Q.; Xu, Y.; Grunsky, E. Integrated Spatial and Spectrum Method for Geochemical Anomaly Separation. Nat. Resour. Res. 2000, 9, 43–52. [Google Scholar] [CrossRef]
  49. Xu, Y.; Cheng, Q. A Fractal Filtering Technique for Processing Regional Geochemical Maps for Mineral Exploration. Geochem. Explor. Environ. Anal. 2001, 1, 147–156. [Google Scholar] [CrossRef]
  50. Li, C.; Ma, T.; Shi, J. Application of a Fractal Method Relating Concentrations and Distances for Separation of Geochemical Anomalies from Background. J. Geochem. Explor. 2003, 77, 167–175. [Google Scholar] [CrossRef]
  51. Rahmati, A.; Afzal, P.; Abrishamifar, S.A.; Sadeghi, B. Application of Concentration–Number and Concentration–Volume Fractal Models to Delineate Mineralized Zones in the Sheytoor Iron Deposit, Central Iran. Arab. J. Geosci. 2015, 8, 2953–2965. [Google Scholar] [CrossRef]
  52. Afzal, P.; Alghalandis, Y.F.; Khakzad, A.; Moarefvand, P.; Omran, N.R. Delineation of Mineralization Zones in Porphyry Cu Deposits by Fractal Concentration–Volume Modeling. J. Geochem. Explor. 2011, 108, 220–232. [Google Scholar] [CrossRef]
  53. Chhabra, A.B.; Meneveau, C.; Jensen, R.V.; Sreenivasan, K. Direct Determination of the f(α) Singularity Spectrum and Its Application to Fully Developed Turbulence. Phys. Rev. A 1989, 40, 5284. [Google Scholar] [CrossRef] [PubMed]
  54. Cheng, Q. Mapping Singularities with Stream Sediment Geochemical Data for Prediction of Undiscovered Mineral Deposits in Gejiu, Yunnan Province, China. Ore Geol. Rev. 2007, 32, 314–324. [Google Scholar] [CrossRef]
  55. Cheng, Q.; Zhao, P. Singularity Theories and Methods for Characterizing Mineralization Processes and Mapping Geo-Anomalies for Mineral Deposit Prediction. Geosci. Front. 2011, 2, 67–79. [Google Scholar] [CrossRef]
  56. Shahriari, H.; Honarmand, M.; Ranjbar, H. Comparison of Multi-Temporal Aster Images for Hydrothermal Alteration Mapping Using a Fractal-Aided Sam Method. Int. J. Remote Sens. 2015, 36, 1271–1289. [Google Scholar] [CrossRef]
  57. Chen, Y.J.; Zhang, C.; Li, N.; Yang, Y.F.; Deng, K. Geology of the Mo Deposits in Northeast China. J. Jilin Univ. 2012, 42, 46. [Google Scholar]
  58. Wang, P.; Chen, Y.-J.; Wang, C.-M.; Zhu, X.-F.; Wang, S.-X. Genesis and Tectonic Setting of the Giant Diyanqin’amu Porphyry Mo Deposit in Great Hingan Range, NE China: Constraints from U–Pb and Re–Os Geochronology and Hf Isotopic Geochemistry. Ore Geol. Rev. 2017, 81, 760–779. [Google Scholar] [CrossRef]
  59. Wu, G.; Li, X.; Xu, L.; Wang, G.; Liu, J.; Zhang, T.; Quan, Z.; Wu, H.; Li, T.; Zeng, Q. Age, Geochemistry, and Sr–Nd–Hf–Pb Isotopes of the Caosiyao Porphyry Mo Deposit in Inner Mongolia, China. Ore Geol. Rev. 2017, 81, 706–727. [Google Scholar] [CrossRef]
  60. Huang, F.; Chen, Y.; Wang, D.; Yuan, Z.; Chen, Z. A Discussion on the Major Molybdenum Ore Concentration Areas in China and Their Resource Potential. Geol. China 2011, 5, 002. [Google Scholar]
  61. Wu, H.; Zhang, L.; Pirajno, F.; Shu, Q.; Zhang, M.; Zhu, M.; Xiang, P. The Mesozoic Caosiyao Giant Porphyry Mo Deposit in Inner Mongolia, North China and Paleo-Pacific Subduction-Related Magmatism in the Northern North China Craton. J. Asian Earth Sci. 2016, 127, 281–299. [Google Scholar] [CrossRef]
  62. Chen, Y. Mesozoic Mo Deposits in Northern North China Craton. In Main Tectonic Events and Metallogeny of the North China Craton; Springer: Berlin/Heidelberg, Germany, 2016; pp. 487–510. [Google Scholar]
  63. Shen, C.; Zhang, M.; Yu, X.; Chen, W.; Gao, W.; Zhou, W. New Progresses in Exploration of Molybdenum Deposits and Analysis of Mineralization Prospect in Inner Mongolia. Geol. Explor. 2010, 46, 561–575. [Google Scholar]
  64. Yu, X.; Chen, W.; Li, W. Discovery and Prospecting Significance of Dasuji Porphyry Molybdenum Deposit, Inner Mongolia. Geol. Prospect. 2008, 2, 006. [Google Scholar]
  65. Nie, F.; Liu, Y.; Zhao, Y.; Cao, Y. Discovery of Dasuji and Caosiyao Large-Size Mo Deposits in Central Inner Mongolia and Its Geological Significances. Miner. Depos. 2012, 31, 930–940. [Google Scholar]
  66. Fujisada, H.; Iwasaki, A.; Hara, S. Aster Stereo System Performance. In Sensors, Systems, and Next-Generation Satellites V; SPIE: Bellingham, WA, USA, 2001. [Google Scholar]
  67. Abrams, M.J.; Rothery, D.A.; Pontual, A. Mapping in the Oman Ophiolite Using Enhanced Landsat Thematic Mapper Images. Tectonophysics 1988, 151, 387–401. [Google Scholar] [CrossRef]
  68. Zuo, R.; Xia, Q.; Zhang, D. A Comparison Study of the C–A and S–A Models with Singularity Analysis to Identify Geochemical Anomalies in Covered Areas. Appl. Geochem. 2013, 33, 165–172. [Google Scholar] [CrossRef]
  69. Cheng, Q. Multifractality and Spatial Statistics. Comput. Geosci. 1999, 25, 949–961. [Google Scholar] [CrossRef]
  70. Khan, S.D.; Mahmood, K.; Casey, J.F. Mapping of Muslim Bagh Ophiolite Complex (Pakistan) Using New Remote Sensing, and Field Data. J. Asian Earth Sci. 2007, 30, 333–343. [Google Scholar] [CrossRef]
  71. Rajendran, S.; Al-Khirbash, S.; Pracejus, B.; Nasir, S.; Al-Abri, A.H.; Kusky, T.M.; Ghulam, A. Aster Detection of Chromite Bearing Mineralized Zones in Semail Ophiolite Massifs of the Northern Oman Mountains: Exploration Strategy. Ore Geol. Rev. 2012, 44, 121–135. [Google Scholar] [CrossRef]
  72. Pournamdari, M.; Hashim, M.; Pour, A.B. Spectral Transformation of Aster and Landsat TM Bands for Lithological Mapping of Soghan Ophiolite Complex, South Iran. Adv. Space Res. 2014, 54, 694–709. [Google Scholar] [CrossRef]
  73. Van der Werff, H.; van der Meer, F. Sentinel-2 for Mapping Iron Absorption Feature Parameters. Remote Sens. 2015, 7, 12635–12653. [Google Scholar] [CrossRef]
  74. Emam, A.; Zoheir, B.; Johnson, P. Aster-Based Mapping of Ophiolitic Rocks: Examples from the Allaqi–Heiani Suture, Se Egypt. Int. Geol. Rev. 2016, 58, 525–539. [Google Scholar] [CrossRef]
  75. Gupta, R.P. Remote Sensing Geology; Springer: Berlin/Heidelberg, Germany, 2017; ISBN 3662558769. [Google Scholar]
  76. Loughlin, W. Principal Component Analysis for Alteration Mapping. Photogramm. Eng. Remote Sens. 1991, 57, 1163–1169. [Google Scholar]
  77. Crosta, A.; De Souza Filho, C.; Azevedo, F.; Brodie, C. Targeting Key Alteration Minerals in Epithermal Deposits in Patagonia, Argentina, Using Aster Imagery and Principal Component Analysis. Int. J. Remote Sens. 2003, 24, 4233–4240. [Google Scholar] [CrossRef]
  78. Liu, L.; Zhou, J.; Han, L.; Xu, X. Mineral Mapping and Ore Prospecting Using Landsat TM and Hyperion Data, Wushitala, Xinjiang, Northwestern China. Ore Geol. Rev. 2017, 81, 280–295. [Google Scholar] [CrossRef]
  79. Mars, J.C.; Rowan, L.C. Regional Mapping of Phyllic-and Argillic-Altered Rocks in the Zagros Magmatic Arc, Iran, Using Advanced Spaceborne Thermal Emission and Reflection Radiometer (Aster) Data and Logical Operator Algorithms. Geosphere 2006, 2, 161–186. [Google Scholar] [CrossRef]
  80. Matin, S.; Hower, J.C.; Farahzadi, L.; Chelgani, S.C. Explaining Relationships among Various Coal Analyses with Coal Grindability Index by Random Forest. Int. J. Miner. Process. 2016, 155, 140–146. [Google Scholar] [CrossRef]
  81. Li, L.; Solana, C.; Canters, F.; Kervyn, M. Testing Random Forest Classification for Identifying Lava Flows and Mapping Age Groups on a Single Landsat 8 Image. J. Volcanol. Geotherm. Res. 2017, 345, 109–124. [Google Scholar] [CrossRef]
  82. Cutler, D.R.; Edwards, T.C.; Beard, K.H.; Cutler, A.; Hess, K.T.; Gibson, J.; Lawler, J.J. Random Forests for Classification in Ecology. Ecology 2007, 88, 2783–2792. [Google Scholar] [CrossRef]
  83. Oliveira, S.; Oehler, F.; San-Miguel-Ayanz, J.; Camia, A.; Pereira, J.M. Modeling Spatial Patterns of Fire Occurrence in Mediterranean Europe Using Multiple Regression and Random Forest. For. Ecol. Manag. 2012, 275, 117–129. [Google Scholar] [CrossRef]
  84. Rodriguez-Galiano, V.F.; Ghimire, B.; Rogan, J.; Chica-Olmo, M.; Rigol-Sanchez, J.P. An Assessment of the Effectiveness of a Random Forest Classifier for Land-Cover Classification. ISPRS J. Photogramm. Remote Sens. 2012, 67, 93–104. [Google Scholar] [CrossRef]
  85. Breiman, L. Random Forests. Mach. Learn. 2001, 45, 5–32. [Google Scholar] [CrossRef]
  86. 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. 2023; in press. [Google Scholar] [CrossRef]
  87. Yao, F.; Liu, S.; Wang, D.; Geng, X.; Wang, C.; Jiang, N.; Wang, Y. Review on the development of multi- and hyperspectral Remote sensing technology for exploration of copper-gold deposits. Ore Geol. Rev. 2023, 162, 105732. [Google Scholar] [CrossRef]
  88. Yu, H.; Li, P. Lithologic mapping using LANDSAT ETM + and ASTER data. Acta Petrol. Sin. 2010, 26, 345–351. [Google Scholar]
  89. Pour, A.B.; Hashim, M. Identification of Hydrothermal Alteration Minerals for Exploring of Porphyry Copper Deposit Using Aster Data, Se Iran. J. Asian Earth Sci. 2011, 42, 1309–1323. [Google Scholar] [CrossRef]
  90. Van der Meer, F.; Van der Werff, H.; Van Ruitenbeek, F. Potential of Esa’s Sentinel-2 for Geological Applications. Remote Sens. Environ. 2014, 148, 124–133. [Google Scholar] [CrossRef]
  91. Zhang, T.; Yi, G.; Li, H.; Wang, Z.; Tang, J.; Zhong, K.; Li, Y.; Wang, Q.; Bie, X. Integrating Data of Aster and Landsat-8 OLI (Ao) for Hydrothermal Alteration Mineral Mapping in Duolong Porphyry Cu-Au Deposit, Tibetan Plateau, China. Remote Sens. 2016, 8, 890. [Google Scholar] [CrossRef]
  92. Gad, S.; Kusky, T. Lithological Mapping in the Eastern Desert of Egypt, the Barramiya Area, Using Landsat Thematic Mapper (TM). J. Afr. Earth Sci. 2006, 44, 196–202. [Google Scholar] [CrossRef]
Remotesensing 15 05532 g001aRemotesensing 15 05532 g001b
Figure 1. (a) Simplified geological map of Northeast China, showing the distribution of Mo deposits. The Dasuji Mo deposit is located in the Northeast of China, modified after Chen et al., [57] and Wang et al., [58]. (b) Regional geological map. (c) Ore deposit geological map of the Dasuji porphyry Mo deposit, (b) and (c) modified after Shen et al., [63], Yu et al., [64] and Nie et al., [65].
Figure 1. (a) Simplified geological map of Northeast China, showing the distribution of Mo deposits. The Dasuji Mo deposit is located in the Northeast of China, modified after Chen et al., [57] and Wang et al., [58]. (b) Regional geological map. (c) Ore deposit geological map of the Dasuji porphyry Mo deposit, (b) and (c) modified after Shen et al., [63], Yu et al., [64] and Nie et al., [65].
Remotesensing 15 05532 g001aRemotesensing 15 05532 g001b
Remotesensing 15 05532 g002
Figure 2. (a) Comparison of spectral features between phyllic minerals (sericite, muscovite) and argillic minerals (kaolinite, alunite) from the USGS mineral spectral library. (b) Spectral features of ferric iron minerals resampled to the ASTER bands. The arrows indicate absorption peaks and valleys.
Figure 2. (a) Comparison of spectral features between phyllic minerals (sericite, muscovite) and argillic minerals (kaolinite, alunite) from the USGS mineral spectral library. (b) Spectral features of ferric iron minerals resampled to the ASTER bands. The arrows indicate absorption peaks and valleys.
Remotesensing 15 05532 g002
Remotesensing 15 05532 g003
Figure 3. The total research flowchart for the alteration anomaly and lithological discrimination.
Figure 3. The total research flowchart for the alteration anomaly and lithological discrimination.
Remotesensing 15 05532 g003
Remotesensing 15 05532 g004
Figure 4. First four PCs of both multi-temporal ASTER images: (a) PCs of image on 23 March 2003 and (b) PCs of image on 22 March 2006, respectively.
Figure 4. First four PCs of both multi-temporal ASTER images: (a) PCs of image on 23 March 2003 and (b) PCs of image on 22 March 2006, respectively.
Remotesensing 15 05532 g004
Remotesensing 15 05532 g005
Figure 5. Six respective results of argillic alteration extracted from the ASTER data by using PCA and the Band Ratio (b4 + b7)/(b5 + b6) in the deposit area: (a) PC3 of 23 March 2003 using PCA, (b) PC3 of 22 March 2006 using PCA, (c) PC4 of 23 March 2003 using PCA, (d) PC4 of 22 March 2006 using PCA, (e) result of 23 March 2003 using BR, and (f) result of 22 March 2006 using BR.
Figure 5. Six respective results of argillic alteration extracted from the ASTER data by using PCA and the Band Ratio (b4 + b7)/(b5 + b6) in the deposit area: (a) PC3 of 23 March 2003 using PCA, (b) PC3 of 22 March 2006 using PCA, (c) PC4 of 23 March 2003 using PCA, (d) PC4 of 22 March 2006 using PCA, (e) result of 23 March 2003 using BR, and (f) result of 22 March 2006 using BR.
Remotesensing 15 05532 g005
Remotesensing 15 05532 g006
Figure 6. Log–log plot of concentrations versus areas of (a) ferric iron of 23 March 2003, (b) ferric iron of 22 March 2006, (c) hydroxyl alteration of 23 March 2003, and (d) hydroxyl alteration of 22 March 2006. The colorful lines indicate the turning point of the fitted line segment.
Figure 6. Log–log plot of concentrations versus areas of (a) ferric iron of 23 March 2003, (b) ferric iron of 22 March 2006, (c) hydroxyl alteration of 23 March 2003, and (d) hydroxyl alteration of 22 March 2006. The colorful lines indicate the turning point of the fitted line segment.
Remotesensing 15 05532 g006
Remotesensing 15 05532 g007aRemotesensing 15 05532 g007b
Figure 7. The results of the ferric alterations obtained from the ASTER data, shown as the blue pixels: (a) anomalies extracted by the C–A model of 23 March 2003, (b) anomalies extracted by the C–A model of 22 March 2006 and (c) results of anomalies derived via anomaly-overlaying selection.
Figure 7. The results of the ferric alterations obtained from the ASTER data, shown as the blue pixels: (a) anomalies extracted by the C–A model of 23 March 2003, (b) anomalies extracted by the C–A model of 22 March 2006 and (c) results of anomalies derived via anomaly-overlaying selection.
Remotesensing 15 05532 g007aRemotesensing 15 05532 g007b
Remotesensing 15 05532 g008aRemotesensing 15 05532 g008b
Figure 8. The results of the hydroxyl alterations obtained from the ASTER data, shown as the purple pixels: (a) anomalies extracted by the C–A model of 23 March 2003, (b) anomalies extracted by the C–A model of 22 March 2006, and (c) results of anomalies derived via anomaly-overlaying selection.
Figure 8. The results of the hydroxyl alterations obtained from the ASTER data, shown as the purple pixels: (a) anomalies extracted by the C–A model of 23 March 2003, (b) anomalies extracted by the C–A model of 22 March 2006, and (c) results of anomalies derived via anomaly-overlaying selection.
Remotesensing 15 05532 g008aRemotesensing 15 05532 g008b
Remotesensing 15 05532 g009
Figure 9. (a) Ferric and hydroxyl alteration anomaly mapping superimposed on the ASTER band 1 image: ferric iron (red) and hydroxyl alteration (blue); (b) the zoom image that shows the distribution of the alteration anomalies in the Dasuji region; and (c) the image that shows the distribution of the alteration anomalies in the Dasuji deposit.
Figure 9. (a) Ferric and hydroxyl alteration anomaly mapping superimposed on the ASTER band 1 image: ferric iron (red) and hydroxyl alteration (blue); (b) the zoom image that shows the distribution of the alteration anomalies in the Dasuji region; and (c) the image that shows the distribution of the alteration anomalies in the Dasuji deposit.
Remotesensing 15 05532 g009
Remotesensing 15 05532 g010
Figure 10. RGB composition image of the ASTER (R: band 3; G: band 2; B: band 1) and field pictures in the study area.
Figure 10. RGB composition image of the ASTER (R: band 3; G: band 2; B: band 1) and field pictures in the study area.
Remotesensing 15 05532 g010
Remotesensing 15 05532 g011
Figure 11. Log–log plot of the concentrations versus areas of the DEM data (a), and the lithological population distribution by the DEM based on the C–A model (b).
Figure 11. Log–log plot of the concentrations versus areas of the DEM data (a), and the lithological population distribution by the DEM based on the C–A model (b).
Remotesensing 15 05532 g011
Remotesensing 15 05532 g012
Figure 12. Lithological mapping results using the RFC method in the study area: (a) mapping results using the ASTER + ETM data combination, and (b) mapping results using the ASTER + ETM + DEM data combination.
Figure 12. Lithological mapping results using the RFC method in the study area: (a) mapping results using the ASTER + ETM data combination, and (b) mapping results using the ASTER + ETM + DEM data combination.
Remotesensing 15 05532 g012
Remotesensing 15 05532 g013
Figure 13. Field verifications of the hydrothermal alterations in the study area. Examples include (a) granulite intruded by the diabase dyke; (b) granulite intruded by granite; (c) hydroxyl (OH)-bearing alteration, such as chlorite, sericite, kaolinite, and limonite; (d) kaolinite alteration; (e) garnet leptite, showing kaolinite, ferric iron and chlorite; (f) garnet granite gneiss, showing kaolinite and ferric iron; (g) molybdenum ore, showing stockwork molybdenite and pyrite; and (h) quartz porphyry, showing feldspar, quartz, molybdenite, pyrite, kaolinite and biotite.
Figure 13. Field verifications of the hydrothermal alterations in the study area. Examples include (a) granulite intruded by the diabase dyke; (b) granulite intruded by granite; (c) hydroxyl (OH)-bearing alteration, such as chlorite, sericite, kaolinite, and limonite; (d) kaolinite alteration; (e) garnet leptite, showing kaolinite, ferric iron and chlorite; (f) garnet granite gneiss, showing kaolinite and ferric iron; (g) molybdenum ore, showing stockwork molybdenite and pyrite; and (h) quartz porphyry, showing feldspar, quartz, molybdenite, pyrite, kaolinite and biotite.
Remotesensing 15 05532 g013
Remotesensing 15 05532 g014
Figure 14. Field observations of the lithological units in the study area, showing (a) the Archean garnet leptite, (b) the Archean garnet gneiss, and (c,d) the large outcrop of carmeloite.
Figure 14. Field observations of the lithological units in the study area, showing (a) the Archean garnet leptite, (b) the Archean garnet gneiss, and (c,d) the large outcrop of carmeloite.
Remotesensing 15 05532 g014
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Xi, M.; Zhang, W.; Tang, J.; Gao, H.; Shalamzari, M.J. Application of a Multifractal Model for Identification of Lithology and Hydrothermal Alteration in the Dasuji Porphyry Mo Deposit in Inner Mongolia, China. Remote Sens. 2023, 15, 5532. https://doi.org/10.3390/rs15235532

AMA Style

Xi M, Zhang W, Tang J, Gao H, Shalamzari MJ. Application of a Multifractal Model for Identification of Lithology and Hydrothermal Alteration in the Dasuji Porphyry Mo Deposit in Inner Mongolia, China. Remote Sensing. 2023; 15(23):5532. https://doi.org/10.3390/rs15235532

Chicago/Turabian Style

Xi, Mingjie, Wanchang Zhang, Jiakui Tang, Huiran Gao, and Masoud Jafari Shalamzari. 2023. "Application of a Multifractal Model for Identification of Lithology and Hydrothermal Alteration in the Dasuji Porphyry Mo Deposit in Inner Mongolia, China" Remote Sensing 15, no. 23: 5532. https://doi.org/10.3390/rs15235532

APA Style

Xi, M., Zhang, W., Tang, J., Gao, H., & Shalamzari, M. J. (2023). Application of a Multifractal Model for Identification of Lithology and Hydrothermal Alteration in the Dasuji Porphyry Mo Deposit in Inner Mongolia, China. Remote Sensing, 15(23), 5532. https://doi.org/10.3390/rs15235532

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop