Integrated Prospectivity Mapping for Copper Mineralization in the Koldar Massif, Kazakhstan

Abstract

This study developed a copper mineral prospectivity map for the Koldar massif, Kazakhstan, using an integrated approach combining geophysical and satellite methods. A strong spatialgenetic link was identified between faults and hydrothermal mineralization, with faults acting as key conduits for ore-bearing fluids. Lineament analysis and density mapping confirmed the high permeability of the Koldar massif, indicating its structural prospectivity. Hyperspectral and multispectral data (ASTER, PRISMA, WorldView-3) were applied for detailed mapping of hydrothermal alteration (phyllic, propylitic, argillic zones), which are critical for discovering porphyry copper deposits. In particular, WorldView-3 imagery facilitated the identification of new prospective zones. The transformation of magnetic and gravity data successfully delineated geological features and structural boundaries, confirming the fractured nature of the massif, a key structural factor for mineralization. The resulting map of prospective zones, created by normalizing and integrating four evidential layers (lineament density, PRISMA-derived hydrothermal alteration, magnetic, and gravity anomalies), is thoroughly validated, successfully outlining the known Aktogay, Aidarly, and Kyzylkiya deposits. Furthermore, new, previously underestimated prospective areas were identified. This work fills a significant knowledge gap concerning the Koldar massif, which had not been extensively studied using satellite methods previously. The key advantage of this research lies in its comprehensive approach and the successful application of high-quality hyperspectral imagery for mapping new prospective zones, offering a cost-effective and efficient alternative to traditional ground-based investigations.

1. Introduction

The global mining industry faces a serious challenge: with 250 active copper mines across 40 countries and estimated reserves of 870 million tons, resource depletion is inevitable. S&P Global forecasts a copper deficit of 10 million tons by as early as 2035 [1], demanding immediate and decisive action.

Traditional approaches to increasing production, such as modernizing metal extraction processes or re-exploring the flanks of existing deposits, while important, cannot fully offset the anticipated deficit. These methods primarily optimize known resources but fail to uncover fundamentally new sources.

Therefore, there is an urgent need to attract and actively implement entirely new, advanced exploration methods and technologies. Only the comprehensive application of innovative approaches, including modern remote sensing, geophysics, artificial intelligence, and machine learning, will enable the efficient and timely discovery and commissioning of new, hidden deposits capable of meeting the growing global demand for copper. Without a radical shift in the exploration paradigm, a global copper deficit will become a reality.

Kazakhstan holds a significant position in the global copper market, accounting for approximately 4% of worldwide production. As of 2024, the country ranks among the top 11 global leaders in copper reserves, with proven reserves of 40.8 million tons (estimated at an additional 20 million tons). Over 77% of these reserves are concentrated in Central and Eastern Kazakhstan.

In Northeast Balkhash region, the primary prospects for copper are linked to the porphyry copper ore formation, with the most significant and frequent copper accumulations known within the Koldar ore cluster (Aktogay ore field). Copper occurrences here are predominantly of hydrothermal origin, less commonly skarn or sedimentary, and belong to various ore formations. However, practical exploration interest is exclusively focused on the porphyry copper vein-disseminated formation [2,3,4].

Despite its undeniable significance and the presence of known deposits, the Aktogay ore region remains critically underexplored. The majority of geological investigations in this key area were conducted over 40 years ago. At that time, these efforts relied on traditional exploration methods which, by modern standards, are insufficiently effective and representative to fully unlock the region’s true potential.

While the fundamental geological framework of the area remains relevant, its vast yet largely concealed mineral potential demands urgent and comprehensive re-evaluation. This is achievable only through the active and widespread adoption of advanced remote sensing technologies and modern digital data interpretation. Employing these innovative approaches will not only allow for more comprehensive and precise mapping of known anomalies but also for the identification of previously undetected, deep-seated, or subtly expressed prospective structures. Such a comprehensive and in-depth re-evaluation is absolutely critical for the future development of Kazakhstan’s copper industry and for ensuring its long-term sustainability amidst rising global demand.

Global experience demonstrates that the application of remote sensing data plays a crucial role in copper exploration worldwide [5,6,7,8,9,10,11]. Currently, both traditional and modern technological approaches are employed in mineral exploration. Traditional methods include geological [12,13,14,15,16,17,18], geophysical [19,20,21,22,23,24,25], and geochemical investigations [26,27,28,29]. Modern approaches increasingly integrate remote sensing technologies [30,31,32,33], Geographic Information Systems (GISs) [34,35,36,37], numerical modeling [38], and advanced computational tools like artificial intelligence (AI) and machine learning (ML) algorithms [39,40,41,42,43,44,45].

Recent advances in multi- and hyperspectral remote sensing have significantly improved the ability to detect and map hydrothermal alteration minerals associated with ore deposits. These techniques provide detailed spectral information that enables discrimination of mineral assemblages at high spatial and spectral resolution, supporting more effective geological mapping and mineral exploration in both arid and vegetated terrains [46].

To enhance the accuracy and reliability of data interpretation, the integration of multiple methods and technologies has become widespread [47,48,49,50,51,52]. This integrated approach allows for more effective assessment of mineral prospectivity and ensures reliable results.

In our study, a comprehensive set of geophysical and satellite data was utilized to map geological features associated with copper mineralization zones. Geophysical fields were employed to delineate lithological units and define geological boundaries. Specifically, gravity data allowed for the differentiation of various lithological formations based on density variations of rocks [53,54]. Concurrently, magnetic data were applied to identify magnetic anomalies, indicating the presence or absence of magnetic minerals linked to the heterogeneous lithology of the study area [55,56,57]. While both gravity and magnetic methods are effective individually, their joint application significantly enhances interpretation reliability. Each method compensates for the limitations of the other, thereby providing a more complete understanding of the regional geological structure [58,59,60,61].

Fault systems within the Earth’s crust play a crucial role as conduits for the migration of ore-bearing hydrothermal fluids and organic compounds, thus participating in or creating the necessary conditions for the formation of mineral deposits [62,63]. In this regard, fault-related structures are considered a key control on mineralization, particularly of hydrothermal origin. The hypothesis of a spatial and genetic link between hydrothermal ore bodies and fault zones is widely regarded as one of the most promising for mineral deposit prediction. In this context, lineaments, which represent surface expressions of deep-seated tectonic structures, can serve as indicators of mineral mineralization, especially copper.

We applied lineament analysis to study the influence of tectonic deformation on rock permeability as a critical factor controlling ore-forming processes. Although this analysis does not directly identify mineral deposits, it plays a decisive role in identifying potentially mineralized zones, as well as the geometry and development mechanisms of ore-controlling structures.

Spectral analysis of satellite imagery, varying in sensor type and acquisition scale, represents a key outcome of this research and is an indispensable tool in modern geological exploration. Through a comprehensive analysis of ASTER, WorldView-3, and PRISMA optical images, zones of hydrothermal alteration associated with copper mineralization were successfully identified. This approach has long demonstrated its effectiveness in delineating prospective ore-bearing areas [64,65,66,67,68].

The results of this study offer an additional tool for integrating ground-based geological–geophysical and satellite methods to identify prospective porphyry copper structures across Kazakhstan. Detailed and comprehensive field testing, as well as further validation and development of advanced machine learning algorithms, are planned.

2. Geographic and Geologic Settings of the Study Area

2.1. Geography

The study area is in the central part of the UralMongolian fold belt, within the DzungarianBalkhash fold system (Balkhash crustal segment) in Northeast Balkhash. Administratively, the area is located in the Ayagoz district of the East Kazakhstan region of the Republic of Kazakhstan (Figure 1).

The landscape features gently undulating sandy plains and low hills, part of the Koldar Mountains, with a relative relief ranging from 5 to 30 m and absolute elevations from 390 to 470 m above sea level. The region experiences a sharply continental climate, characterized by significant diurnal and seasonal temperature fluctuations. Winters are cold and harsh, while summers are typically hot and arid. The river network, including the Ayagoz River, is poorly developed.

The study area is primarily an agricultural region, dominated by transhumant livestock farming (sheep breeding). To a lesser extent, it also supports dairy production and crop cultivation, including wheat, corn, sunflowers, and fodder grasses.

Vegetation is sparse, characteristic of semi-desert and dry steppe zones, consisting mainly of drought-resistant grasses and shrubs such as feather grass (Stipa spp.) and wormwood (Artemisia spp.). Due to the low vegetation cover and limited surface biomass, extensive outcrops of bedrock are present. These conditions are highly favorable for remote sensing, as they enhance the visibility and spectral detection of hydrothermal alteration zones associated with mineralization.

2.2. Geology

The generalization of geological and structural features of the Koldar massif presented in this section is made based on scientific publications [70,71,72,73,74,75] and materials of specialized geological reports (1961–1997).

The Koldar massif, located in the Eastern Balkhash region, is a key geological object, whose structure and evolution are closely related to the regional tectonics and metallogeny of the DzungarBalkhash fold system. In a broader structural context, it is an integral part of the Bakanas synclinorium, which, in turn, is part of the extensive DzungarBalkhash folded system. This system, occupying the southeast of Kazakhstan, is characterized by a long-term eugeosynclinal development regime, covering the Baikal and Caledonian tectonic epochs.

The Koldar massif is also located in the eastern part of the deep volcanic belt extending along the northern part of Lake Balkhash, which emphasizes its connection with large porphyry copper deposits of the region, such as Kounrad and Aktogay. It is the Koldar massif that forms the core of the Koldar horst-anticline, which makes it a central element in understanding the geological structure and ore content of the Aktogay deposit.

In the modern erosion section, the Koldar massif has a characteristic lenticular shape, elongated in the sublatitudinal direction, with dimensions of 17 by 4 ÷ 8 km and a total area of about 75 square kilometers. The axial line of the massif is slightly curved in the northern direction. The deep structure of the massif, revealed by gravity, magnetic and seismic exploration (GM) data, indicates its laccolithic form. This means that the magma intruded to a relatively shallow depth, spreading between the layers of host rocks, and causing the uplift of the overlying strata. A distinctive feature of the deep structure is a strongly lowered southern wing and a clearly defined feeder channel located on the side of the hanging wall of the deep suture fault of the Aktogay direction. The neck of the laccolith can be traced to a significant depth of 4.7 km. Such morphology indicates a close relationship between magmatism and tectonics, where deep faults not only served as paths for the rise of magma, but also actively participated in the deformation of the massif both during and after the intrusion.

Stratigraphically, the Koldar massif intrudes volcanic formations of the Middle-Upper Carboniferous Keregetas suite, which are represented by intensively reworked effusive-pyroclastic and volcanomictic terrigenous deposits of medium and acidic com-position. These rocks lie monoclinally with a dip to the southwest. The area also contains younger volcanogenic-sedimentary deposits of the Koldar suite, located in the southeast of the deposit. The composition of the Koldar suite includes conglomerates with granitoid pebbles, which indicates active erosion of the already exposed parts of the Koldar massif during its formation. The Koldar intrusive massif itself is composed of diorites and gran-odiorites and is attributed to the first phase of the Aktogay intrusive complex of Upper Carboniferous age. It is composed of four subphases of the intrusion, reflecting the differ-entiation of the magmatic melt: from early gabbro-diorites and various types of diorites to later medium-grained diorites, granodiorites and a “porphyry” subphase represented by small bodies and dikes of granodiorite porphyry and plagiogranite porphyry. This sequence from more basic to more acidic rocks is typical for intrusive complexes formed under conditions of a long magmatic process.

Fault tectonics plays a decisive role in the structure of the Koldar massif and the Aktogay ore field. The location of the massif was controlled by the Aktogay fault zone of sublatitudinal strike (290 degrees). The structure of the ore field is significantly complicated by large, long-lived disjunctive dislocations of three main directions: Aktogay (sublatitudinal), Koldar (northeastern) and northwestern. The Koldar fault zone is younger than the Aktogay and northwestern faults. Of particular importance is the Southern Koldar fault, which borders the box-shaped brachysyncline filled with Koldar suite deposits from the north. This fault extends NE 250 and dips south at an angle of about 80 degrees. Its nature as a normal fault with a vertical displacement amplitude of more than 500 m is con-firmed by the absence of granitoids in the southern downthrown block to significant depths. This fact suggests that extensional stresses predominated during certain periods, possibly under conditions of post-orogenic collapse or rifting.

The main Koldar fault also shows a significant displacement amplitude, exceeding 140 m. These faults are long-lived structures that have been repeatedly activated, con-trolling both magmatism and subsequent deformation, and playing a key role in the localization of ore bodies.

The significant vertical displacements along the faults indicate that deep faults not only served as pathways for magma ascent, but also actively participated in the formation and subsequent deformation of the intrusive body.

The neotectonic stage of the region’s development is confirmed by the distinct manifestation of the Southern Koldar and Main Koldar faults in the modern relief. Moreover, these faults displace Paleogene deposits, which is direct evidence of their activity in the Cenozoic, up to the modern stage. This indicates ongoing vertical and, possibly, horizontal movements of the Earth’s crust, which can be associated with regional stress fields inherited from the Hercynian orogeny or associated with modern tectonic processes in the Central Asian mobile belt.

Structural parageneses of the Koldar massif are closely related to its intrusion and subsequent tectonic evolution. The massif forms the core of the Koldar horst-anticline, the flanks of which are composed of younger volcanogenic-sedimentary suites. The Aktogay deposit associated with the Koldar massif is a stockwork type characteristic of a porphyry copper formation. Its ring geological formations and ore-bearing stockwork are deter-mined by the tectonic structure of the territory, which is confirmed by geophysical and lithogeochemical data. The formation of ores occurred in two main stages: the early alka-line stage, including potassium feldspathization, silicification, biotitization, chloritization, epidotization and prenitization, and the later acidic stage, characterized by silicified, sericitized, chloritized and carbonated rocks. Most of the ore substance is associated with the first, the alkaline stage.

The zones of development of hydrothermally altered rocks with mineralization are often recorded by local depressions of the magnetic and gravitational fields. Ore-bearing granitoids and volcanics are distinguished by large positive anomalies of the magnetic and gravimetric fields. The formation age of the Koldar massif (Upper Carboniferous) be-longs to the Neo-Variscian phase of the Hercynian orogeny, which covered the period from the Late Devonian to the Late Carboniferous (380–290 million years ago). This phase was characterized by intense thrust tectonics, significant crustal thickening, thermal disturbances and large-scale plutonism, including granite formation.

The Koldar ore cluster has predicted P2 + P3 resources of up to 15 million tons of cop-per, 276 thousand tons of molybdenum, 60 tons of gold. Thus, the Koldar massif is a striking example of an intrusive body whose formation, morphology, internal structure and ore content are a direct consequence of the multi-stage Hercynian orogeny and sub-sequent neotectonic activity, which makes it an important object for studying tectono-magmatic processes and metallogeny of the region.

3. Materials and Methods

3.1. Data

In this study, satellite and geophysical datasets were processed and integrated to delineate prospective zones of copper mineralization. The satellite data included a digital elevation model (DEM), ASTER multispectral imagery, WorldView-3 high-resolution multispectral imagery, and PRISMA hyperspectral imagery.

The DEM was derived from ALOS-PALSAR FBS (Fine Beam Single polarization mode, HH) RT1 (Radiometrically Terrain Corrected) data, with a spatial resolution of 12.5 m, obtained from the Alaska Satellite Facility (https://asf.alaska.edu/ (accessed on 27 March 2024)).

The ASTER image was selected based on two primary criteria. First, its 60 km × 60 km scene size provided complete coverage of the study area within a single frame. Second, to ensure the capture of the full spectral range, data acquired prior to 2008 were necessary due to the failure of ASTER’s shortwave infrared (SWIR) sensor in that year. Given that commercial mining at the Aktogay deposit commenced in 2015, a scene from 30 August 2004 (AST_L1B_0030830200200404054250_20240619115630_2680975) was chosen. This allowed for analysis unaffected by surface alterations resulting from subsequent mining activities. This image was downloaded from the official NASA Earthdata platform (https://search.earthdata.nasa.gov/ (accessed on 5 February 2024)).

The PRISMA image, acquired by the Italian Space Agency’s hyperspectral satellite (https://prisma.asi.it/ (accessed on 10 April 2024)) launched in 2019, provides data across 239 spectral channels within the 400–2500 nm range, encompassing visible, near-infrared, and shortwave infrared bands [76,77]. This extensive spectral resolution offers significant advantages for various geological exploration applications. The PRISMA scene from 2 April 2024 (PRS_L2D_STD_2024040202053906_2024040202053911_0001) was selected due to minimal cloud cover and moderate vegetation development. Despite being the early growing season, the semi-desert landscape’s sparse vegetation allowed for effective detection of spectral features indicative of hydrothermal alteration.

All aforementioned ASTER and PRISMA remote sensing data are publicly available and can be downloaded free of charge after user registration.

A WorldView-3 image covering 25.4 km² of the Aidarly deposit was acquired commercially on 11 May 2019. WorldView-3 is a commercial high-resolution multispectral satellite equipped with multiple sensors. Operating in a 617 km orbit, it provides imagery with a spatial resolution of 31 cm in panchromatic mode, 1.24 m in multispectral mode, 3.7 m in SWIR, and 30 m with the CAVIS (Clouds, Aerosols, Vapors, Ice, and Snow) instrument. The satellite features a revisit time of less than one day. Owing to its ultra-high spatial resolution, WorldView-3 imagery complemented ASTER and PRISMA data, enabling detailed mapping of mineralized zones and refinement of geological structures.

Together, ASTER, PRISMA, and WorldView-3 provide complementary datasets that, when integrated, offer a robust basis for mineral exploration. A detailed characterization of these datasets is presented in

Table 1. Characteristics of space data.

Space Images	Spectral Bands	Spectral Region (μm)	Terrain Coverage (km)	Ground Resolution Cell (m)
ASTER	14	VNIR 0.52–0.86 SWIR 1.6–2.43 TIR 8.125–11.65	60	VNIR 15 by 15 SWIR 30 by 30 TIR 90 by 90
WorldView-3	28	VNIR 0.4–1.04 SWIR 1.195–2.365 CAVIS 0.405–2.245 Pan 0.45–0.8	13.1	VNIR 1.24 SWIR 3.7 CAVI 30 Pan 0.31
PRISMA	239	0.4–2.5	30	30

.

Geophysical data for this study included magnetic and gravity fields. Magnetic data were obtained from the Republican Geological Fund of Kazakhstan. The averaged map of magnetic field anomalies (∆Ta) for the Bakanas synclinorium at a scale of 1:200,000, com-piled by the South Kazakhstan Geophysical Expedition in 1978, was used as a source of magnetic information [78]. Gravity data were represented by a Bouguer anomaly map calculated using a reduction density of 2.67 g/cm³.

3.2. Methods

Our research methodology was an integrated approach combining various types of analysis of the initial satellite and ground-based geophysical data. In general, this study was conducted in several stages: database formation, preliminary processing and preparation for in-depth analysis, extended analysis, post-processing of results and preparation of the final report. The satellite database included DEM (ALOS DEM), multispectral data (ASTER, WorldView-3) and hyperspectral data (PRISMA). Ground-based geophysical data were represented by digitized maps of gravimetric and magnetic surveys of various scales. At the preliminary processing stage, correction of primary materials was carried out (noise removal, relief washing, atmospheric corrections, radiometric calibration, etc.). The main part of the processing involved the use of specialized software methods for each type of data: automatic lineament analysis, False-Color Composite images (FCC), spectral angle method (SAM), analytical signal (AS), first vertical derivatives (1VD), tilt derivatives (TDR) and total horizontal derivatives (HD TDR). At the post-processing stage, filtering and expert analysis of the obtained results were performed to form final conclusions. The complete data-processing workflow is illustrated in Figure 2.

3.2.1. Lineament Analysis

To identify ore-controlling decompressed and fractured structures (lineaments), we used remote sensing methods. Known porphyry copper deposits of the Koldar massif served as reference objects for verifying our results.

The primary input data were the ALOS DEM. The DEM was chosen due to its superior accuracy in identifying lineaments compared to optical imagery [79,80]. This is because many DEMs are generated from radar data, which offers significant advantages over optical data: they can acquire images at any time of day and in any weather, and they can detect objects invisible in the optical or infrared spectrum, including those hidden beneath snow, vegetation, or even below the Earth’s surface.

The methodological framework for lineament extraction comprised six sequential stages:

(1)

Geological and tectonic analysis of the study area;

(2)

Preprocessing of DEM data (hillshading);

(3)

Automatic lineament extraction;

(4)

Post-processing and manual filtering;

(5)

Calculation of quantitative indicators, including rose diagrams and lineament density;

(6)

Validation using ground-truth data and production of final outputs.

The initial step involved reviewing geological maps and reports to understand the regional tectonic framework. Hillshade images were generated to enhance topographic contrast and facilitate better visualization of structural features. Following methodologies reported by [81,82,83], hillshading was applied using eight azimuth directions (0°, 45°, 90°, 135°, 180°, 225°, 270°, and 315°), with a fixed altitude angle of 45°, which proved effective for the local terrain. Hillshades were produced using the 3D Analyst tool in ArcMap.

Subsequently, automatic lineament extraction was carried out using the LINE mod-ule in PCI Geomatica. This module employs a three-step algorithm involving edge detection, thresholding, and curve extraction. The algorithm relies on six configurable parameters [79,80,81,82,83,84]:

• RADI: filter radius;
• GTHR: edge gradient threshold;
• LTHR: minimum curve length;
• FTHR: line fitting tolerance;
• ATHR: angular difference threshold;
• DTHR: link distance threshold.

The optimal parameter values were determined through experimental tuning and literature references [80,81,82,83,84], taking into account terrain characteristics, image resolution, and the geological context (see

Table 2. LINE module parameter values.

Parameters	Description	Unit of Measurement	Default Values	Thresholds Applied
RADI	For the detection of edges	pixel	10	40
GTHR	For the detection of edges	-	100	200
LTHR	Minimum length of a curve taken as lineament	pixel	30	160
FTHR	Allowed tolerance in curve fitting to form a polyline	pixel	3	3
ATHR	Maximum angle between two linked polylines	degrees	30	30
DTHR	Maximum distance between two linked polylines	pixel	20	20

).

After automated extraction, vectorized lineaments were manually filtered to eliminate false positives such as roads, railways, and anthropogenic features. While algorithmic filtering can sort features based on length, shape, and orientation, visual inspection remains essential to ensure geological relevance. This involved comparing extracted features with known infrastructure and topographic maps.

To analyze the structural orientation, rose diagrams were generated using Rock-Works software Version 2023.06.06. Additionally, a lineament density map was created in ArcMap using the Line Density tool, which calculates density as the total lineament length per unit area (km/km²).

3.2.2. Spectral Analysis

At copper porphyry ore deposits, characteristic zonal complexes of hydrothermal alteration minerals are often observed [85]. These minerals possess unique spectral absorption features across the visible–near-infrared (VNIR), shortwave infrared (SWIR; 0.4–2.5 µm) and thermal infrared (TIR; 8.0–14.0 µm) wavelength ranges [86,87].

To remotely map and identify these hydrothermally altered rocks using ASTER, PRISMA, and WorldView-3 data, we compared the spectral signatures of image pixels with library spectral datasets of these minerals [5,88,89].

The spectral reflectance curves of indicator minerals for porphyry copper deposits exhibit characteristic absorption features that can be effectively detected by the ASTER, WorldView-3, and PRISMA sensors (Figure 3). ASTER, with its shortwave infrared (SWIR) bands, is particularly sensitive to absorption features associated with Al–OH and Fe–OH bearing minerals such as muscovite, kaolinite, and chlorite, typically around 1.6–2.4 µm. WorldView-3, due to its high spatial resolution and 8 SWIR bands, captures subtle variations in mineral spectra, enabling detailed mapping of alteration zones with diagnostic absorption features in the 1.2–2.3 µm range. PRISMA, as a hyperspectral sensor, provides continuous spectral coverage across the visible to SWIR range (0.4–2.5 µm), allowing for precise identification of diagnostic absorption features of key alteration minerals, including both phyllosilicates and carbonates. Together, these sensors enable the recognition of mineral-specific spectral patterns critical for discriminating hydrothermal alteration zones associated with porphyry copper systems.

The zonation model of Lowell [90] and later Lowell and Gilbert [85] shows that porphyry copper deposits exhibit distinct alteration zonations, with each zone defined by a unique set of minerals. These compositional changes are due to the participation and mixing of different types of fluids (Figure 4). During deposit formation, two main zones initially develop: in the very center, potassic alteration is formed by ascending magmatic fluid and is immiscible with groundwater. Heat from this magma flow causes ground-water convection, creating a propylitic zone around the core. Later, at the intersection of potassic alteration, phyllitic alteration appears, often associated with mafic ores. Clayey alteration, such as argillic, is usually located above the core of the deposit and extends around its periphery.

Propylitic alteration is characterized by the development of chlorite, epidote, and calcite, which exhibit diagnostic absorption features primarily in the shortwave infrared (SWIR) region. Chlorite shows a broad absorption feature centered around 2.29–2.35 µm (Figure 5), associated with Fe-OH and Mg-OH vibrational modes. This feature is partially captured by ASTER Band 8 and is more clearly resolved in WorldView-3’s SWIR Bands 7 and 8. PRISMA’s high spectral resolution allows for accurate identification and separation of chlorite from spectrally similar minerals. Epidote exhibits an absorption feature near 2.34 µm (Figure 5), which may overlap with chlorite but remains distinguishable using WorldView-3 and PRISMA data due to subtle shifts in reflectance minima. Calcite is recognized by a shallow absorption around 2.33–2.35 µm (Figure 5). These spectral responses enable the delineation of propylitic zones in the peripheral parts of porphyry systems.

Phyllic alteration is typically associated with illite, muscovite, quartz, and pyrite, with illite and muscovite providing the most distinctive spectral signatures in the SWIR region. Both minerals exhibit a sharp absorption feature near 2.20 µm related to Al–OH bonds (Figure 5), which is captured by ASTER Band 6. Illite tends to show a slightly shallower and broader absorption feature than muscovite. This spectral difference can be resolved using the dense spectral sampling of PRISMA, which allows for detailed discrimination of absorption feature shape and depth. Additionally, the high spatial resolution of WorldView-3 enables the detection of small and spatially localized alteration zones, even when the spectral differences are subtle or partially mixed at coarser resolutions. Quartz and pyrite are spectrally neutral in the SWIR region, exhibiting minimal absorption features, but their presence can be inferred indirectly through associated mineral assemblages. The distinct Al–OH absorption around 2.20 µm allows for reliable mapping of phyllic alteration zones (Figure 5), particularly in the central parts of porphyry systems.

Argillic alteration is dominated by kaolinite and alunite, both of which exhibit strong and diagnostic spectral features in the SWIR region. Kaolinite is characterized by a doublet absorption near 2.16 and 2.20 µm due to Al–OH vibrations (Figure 5), and a shoulder around 1.4 µm associated with OH stretching (Figure 5). These features are partially resolved by ASTER (Bands 5 and 6), better separated by WorldView-3, and most precisely defined in PRISMA hyperspectral imagery. Alunite displays a sharp absorption near 2.17 µm and secondary features at 1.48 and 1.77 µm (Figure 5), enabling confident identification in high-resolution datasets. PRISMA, with its continuous spectral coverage, offers the most effective discrimination between kaolinite, alunite, and other clay minerals. The clear and narrow absorption features of these minerals make them reliable indicators of advanced argillic alteration, often associated with high-sulfidation epithermal environments and the upper zones of porphyry systems.

Pre-processing: All satellite images were selected to be cloud-free and with minimal vegetation cover to enhance the spectral contrast of lithological units. The ASTER (Level 1T) and WorldView-3 datasets were obtained as orthorectified images, but without radiometric or atmospheric corrections. Therefore, as part of the preprocessing workflow, both datasets were first subjected to radiometric correction by converting digital numbers (DN) into radiance values, accounting for sensor-specific calibration coefficients and acquisition geometry.

Subsequently, atmospheric correction was performed using the FLAASH (Fast Line-of-sight Atmospheric Analysis of Spectral Hypercubes) module in the ENVI software Version 5.3 environment. The Mid-Latitude Summer (MLS) atmospheric model was selected to match the latitude of the study area (~46° N), while the “Rural” aerosol model was used to represent the natural, low-pollution characteristics of the region. FLAASH converts radiance data to surface reflectance, enabling reliable spectral analysis.

The PRISMA hyperspectral image was downloaded at Level 2D from the official ASI-PRISMA catalog (https://prisma.asi.it/ (accessed on 10 April 2024)), which provides atmospherically corrected, geocoded reflectance products using MODTRAN-based look-up tables (LUTs). Preprocessing of the PRISMA 2D data involved additional refinement steps to enhance spectral reliability. Specifically, noisy spectral bands located at the edges of the spectral range, unstable or dead detector channels, and regions heavily affected by atmospheric water vapor absorption (e.g., 1340–1450 nm and 1800–1950 nm) were excluded due to low signal-to-noise ratios.

Further, all pixels with invalid values (NaN), resulting from residual cloud contamination, instrument noise, or sensor limitations, were removed from the dataset. Since Level 2D data are already atmospherically corrected and expressed as surface reflectance, no additional atmospheric processing was required. To further enhance spectral stability, a Savitzky–Golay filter was applied to the reflectance spectra. This smoothing technique efficiently reduces high-frequency noise while preserving key spectral absorption features critical for mineralogical interpretation.

Thematic processing: To facilitate visual interpretation of hydrothermal alteration zones, FCC (False-Color Composite) imagery was generated from ASTER and PRISMA data using combinations of spectral bands sensitive to specific alteration minerals. Unlike standard RGB composites, FCCs incorporate infrared and SWIR bands, which enhance mineralogical contrasts by exploiting differences in surface reflectance. For example, secondary clays, ferruginous compounds, and silica exhibit distinct spectral characteristics that, in specific band combinations, appear as diagnostic colors (e.g., red, purple, or blue), effectively distinguishing altered from unaltered lithologies. This method supports rapid identification of alteration zones, geologic unit discrimination, detection of linear and ring structures, and preliminary localization of mineralized areas. FCC visualization has been widely recognized as a robust tool for the initial interpretation of both multispectral and hyperspectral data in geological studies [46,92].

To map hydrothermal alteration types, such as propylitic and argillic alterations, at the Aidarly copper porphyry deposit, the SAM (Spectral Angle Mapper) algorithm was applied. SAM is a supervised classification technique that quantifies the spectral similarity between image pixels and reference spectra in n-dimensional reflectance space, making it well-suited for mineral mapping applications [93,94]. This analysis utilized WorldView-3 data, which offers high spatial resolution (1.24 m) and moderate spectral resolution (8 VNIR (near-infrared) and 8 SWIR bands), enabling detailed delineation of mineralized zones.

For the classification, only the eight shortwave infrared (SWIR) bands of the WorldView-3 sensor were used, covering the spectral range from 1195 to 2365 nm. These bands were specifically selected due to their diagnostic absorption features that are particularly effective in distinguishing hydrothermal alteration minerals such as kaolinite, illite, montmorillonite, muscovite, chlorite, and epidote. The spatial resolution of the resulting classification map corresponds to the native resolution of the SWIR bands, which is 3.7 m.

SAM angle thresholds were empirically determined through iterative classification and validation against known alteration zones. Optimal classification performance was achieved using a spectral angle threshold of α = 0.85 for phyllic alteration and α = 0.75 for argillic alteration, balancing classification sensitivity with exclusion of background noise.

All preprocessing and thematic analyses, including atmospheric correction, spectral normalization, and supervised classification, were conducted using ENVI software Version 5.3. Visualization, spatial mapping, and further analysis of classified outputs were carried out in the ArcGIS environment.

3.2.3. Geophysical Processing

The raw geophysical field data were processed using Geosoft Oasis Montaj. Initially, the total magnetic intensity (TMI) data were corrected by subtracting the International Geomagnetic Reference Field (IGRF) to obtain the residual magnetic intensity (RMI). Subsequently, a reduction-to-the-pole (RTP) filter was applied to the RMI data to reposition anomalies directly over their sources, thus simplifying interpretation in areas with moderate to high magnetic inclinations.

Gravity data, represented by the Bouguer anomaly map (calculated using a reduction density of 2.67 g/cm³), were also transformed into a residual gravity anomaly map to isolate local variations by removing regional trends.

To enhance the interpretability of potential field data, a suite of derivative-based filters was applied.

The AS highlights the edges of magnetic and gravity sources and enhances local anomalies by combining the horizontal and vertical gradients. It is particularly effective for identifying geological contacts and the boundaries of magnetized bodies. The AS is calculated using the following expression:

$$\mathrm{A}\mathrm{S}=\sqrt{{\left({\displaystyle \frac{\partial M}{\partial x}}\right)}^2+{\left({\displaystyle \frac{\partial M}{\partial y}}\right)}^2+{\left({\displaystyle \frac{\partial M}{\partial z}}\right)}^2}$$

(1)

where M is the magnetic field, and the partial derivatives represent the gradients in the x, y, and z directions [95].

The 1VD of magnetic data is widely used in geophysical interpretation to enhance near-surface features while suppressing the effect of deep-seated sources. This technique increases the resolution of shallow anomalies, making it particularly valuable for identifying surface and sub-surface geological structures such as faults, fractures, dikes, and small ore bodies [96,97].

The 1VD filter emphasizes the rate of change of the magnetic field with depth, effectively highlighting discontinuities and abrupt variations in the magnetic field associated with structural and lithological boundaries.

The 1VD derivative of the magnetic field is calculated using Equation (2):

$$1\mathrm{V}\mathrm{D}={\displaystyle \frac{\partial M}{\partial z}}$$

(2)

where M is the magnetic field intensity and z is the vertical direction (depth). This derivative is typically implemented using frequency-domain techniques such as the Fourier-transform, where vertical derivatives are calculated by multiplying the frequency spectrum of the magnetic data by a factor proportional to the wavenumber.

By applying the 1VD filter, geoscientists can obtain a clearer image of near-surface features, thereby improving the delineation of ore-controlling structures and enhancing the accuracy of geological interpretations.

The TDR is one of the most widely used edge detection filters in potential field data interpretation due to its ability to enhance shallow subsurface features without amplifying the high-frequency noise typically associated with higher-order vertical derivatives [98]. It is especially useful for mapping subtle magnetic lineaments related to faults, dikes, and lithological contacts.

TDR is calculated using Equation (3):

$$\mathrm{T}\mathrm{D}\mathrm{R}={tan}^{-1}\left({\displaystyle \frac{1\mathrm{V}\mathrm{D}}{\mathrm{T}\mathrm{H}\mathrm{D}\mathrm{R}}}\right)$$

(3)

where:

1VD is the first vertical derivative of the magnetic field. THDR is total horizontal derivatives. It is defined as follows:

$$\mathrm{T}\mathrm{H}\mathrm{D}\mathrm{R}=\sqrt{{\left({\displaystyle \frac{dM}{dx}}\right)}^2+{\left({\displaystyle \frac{dM}{dy}}\right)}^2}$$

(4)

This arctangent relationship provides a normalized result in radians or degrees, allowing for consistent interpretation of shallow and deep magnetic sources across varying geological conditions.

The HD_TDR is a compound filter that enhances the edges of magnetic anomalies by detecting the maximum values located at the boundaries or corners of magnetic bodies [99]. This technique is particularly effective for delineating the edges of large and complex geological features. The HD_TDR is calculated using the following Equation (5):

$${\mathrm{H}\mathrm{D}}_{\mathrm{T}\mathrm{D}\mathrm{R}}=\sqrt{{\left({\displaystyle \frac{\partial TDR}{\partial x}}\right)}^2+{\left({\displaystyle \frac{\partial TDR}{\partial y}}\right)}^2}$$

(5)

where TDR is the tilt derivative, and x, y are the horizontal spatial directions. This filter measures the spatial gradient of the tilt derivative, thereby highlighting structural boundaries with high precision.

These derivatives significantly enhance the interpretability of geophysical data by emphasizing subtle variations in the magnetic and gravity fields. This, in turn, facilitates the identification of critical geological features such as mineralized zones, fault systems, lithological boundaries, and igneous intrusions.

4. Results

4.1. Analysis of Lineaments and Structural Density

This study analyzed eight hillshaded images generated from ALOS DEM data, covering an area of approximately 1215 km². It was observed that the azimuthal direction of illumination during hillshading directly influences the orientation of extracted lineaments. For instance, hillshades with azimuths of 90° and 270° emphasize lineaments with a predominant N–S orientation, while those at 135° and 315° highlight NE–SW-oriented features. To ensure an objective and reliable result, all eight directions were processed and analyzed. The lineament field was extracted from the hillshaded images using the LINE module with the parameter set presented in Table 2.

The automatic extraction followed by manual filtering resulted in a total of 593 lineaments, ranging in length from 600 to 6152 m. These lineaments are interpreted as crustal faults within the study area. They exhibit a linear geometry and possess key characteristics typical of tectonic structures in the region.

The next stage involved the calculation of quantitative indicators for the lineament field, including lineament density and a rose-diagram construction (Figure 6). Rose diagrams were generated within the boundaries of the Koldar massif for both the extracted lineaments and faults derived from a 1:50,000-scale geological map (Figure 1). The lineaments presented in the rose diagram (Figure 6c) demonstrate three dominant directions, which correspond exactly to the trends of geological structures revealed by ground mapping (Figure 6d). Geological data confirm that tectonics is controlled by large fault zones of three main directions: Aktogay (sublatitudinal), Koldar (northeastern) and northwest-ern. Statistical analysis of lineaments also showed the predominance of sublatitudinal faults, as well as northwestern and northeastern directions, which confirms the high reliability of the results of lineament analysis. Most lineaments have lengths in the range of 0,6–1 km (Figure 6a).

Lineament density values range from 0 to 3.5 km/km² (Figure 6b). However, the most geologically significant range in terms of ore mineralization potential lies between 2 and 3.5 km/km². These zones exhibit a high concentration of tectonic features that likely enhance rock permeability and could serve as pathways for hydrothermal fluid migration. Notably, areas with this level of lineament density spatially coincide with known porphyry copper deposits such as Aidarly and Aktogay, supporting their tectonic and hydrogeological significance during ore formation.

This correlation suggests that elevated lineament density is not merely a geomorphological characteristic, but a potential indirect predictor of mineralization. High-density zones may indicate regions of tectonic preparation conducive to porphyry copper deposition, especially at intersections or junctions of structures of different genetic origins. Therefore, areas with a lineament density exceeding 2 km/km² should be considered high-priority targets for further exploration activities.

4.2. False-Color Composites

To effectively map hydrothermal alteration zones within the Koldar massif, data from two different satellite systems were utilized, each providing results at varying levels of spatial and spectral detail. In this study, two FCCs were constructed using ASTER and PRISMA imagery. For ASTER, bands SWIR 4, 6, and 8 were used, while for PRISMA, the SWIR bands 77, 133, and 150 were selected (Figure 7 and Figure 8). Specific band combinations were chosen based on the analytical objectives and the spectral properties of the target materials.

Experimental analyses of ASTER imagery demonstrated that the RGB combination 4–6–8 was the most effective for identifying alteration zones in porphyry copper deposits. Depending on the ore deposit type, the presence of alteration zones can be a strong indicator of potential mineral resources in the region [100]. For instance, the occurrence of chlorite is particularly important for this deposit type. In Figure 7, zones of propylitic alteration are marked in green, while zones of phyllic alteration appear in pink hues. This color distinction is attributed to the high reflectance of alunite, kaolinite, and muscovite in band 4 compared to bands 6 and 8 [101].

Phyllic alteration can be recognized by the presence of sericite (or muscovite), which exhibits strong spectral absorption in the Al-OH feature, especially in ASTER band 6, with a weaker response in band 8 [102].

Figure 8 presents a hydrothermal alteration map generated using the FCC method applied to PRISMA hyperspectral imagery for the study area. Bright pink pixels within specific lithological units delineate the spatial distribution of phyllic alteration zones, while green pixels correspond to propylitic alterations. The resulting image provides valuable insight into the identification of hydrothermal alteration zones potentially associated with copper mineralization.

In the Koldar massif, the propylitic alteration zones are characterized by significant concentrations of chlorite and epidote, indicative of low-temperature metamorphic processes. The phyllic alteration zones, represented by fine-layered metamorphic rocks, contain sericite, quartz, and pyrite, which further supports evidence of hydrothermal activity. These results enable more precise delineation of mineralized zones, thus informing targets for future detailed exploration.

Although only three bands were selected from ASTER and PRISMA imagery to construct the FCC maps for visual representation, the full spectral information available from both sensors—particularly the hyperspectral range of PRISMA—was utilized during data analysis. Specifically, the complete set of SWIR bands was examined to identify diagnostic absorption features associated with alteration minerals. The selection of bands for FCC display was based on their sensitivity to key spectral signatures.

4.3. Spectral Angle Mapper

Hydrothermal porphyry deposits are characterized by various mineralogical alteration zones, such as phyllic, argillic, and propylitic. The FCC method enabled the identification of phyllic and propylitic alterations, while the SAM method applied to WorldView-3 imagery revealed argillic and propylitic alterations at the Aidarly deposit. The processing of WorldView-3 multispectral data resulted in clearly delineated zones of hydrothermally altered minerals at Aidarly, which correlate well with the known geological structure of the deposit. Moreover, a similar hydrothermally altered zone was identified north of the Aidarly deposit, exhibiting comparable structural characteristics (Figure 9).

4.4. Magnetic Conversions

The TMI map of the study area is dominated by elongated positive anomalies with intensities ranging from 6 to 11 nT. These zones are associated with volcanogenic formations of terrigenous andesitedacite, andesitebasaltrhyolite, and andesiterhyolite assemblages. Smaller negative magnetic anomalies, reaching values of –3 nT in the northern part of the area, are related to the volcanogenic-sedimentary rocks of the Koldar Suite (Figure 10a). The maximum value of the AS exceeds 0.006 nT/m, while values within the Koldar massif vary from 0.000016 to 0.002 nT/m, indicating a weak magnetic signature (Figure 10b). Two local minima can be distinguished: one in the northern part of the massif with an isometric rounded shape, and another in the northwest with an irregular shape. It is important to note that these values are relative and should not be interpreted as definitive boundaries of ore bodies.

The maps presented in Figure 11 correspond to (a) 1VD, (b) TDR, and (c) HD TDR. The 1VD filter effectively suppressed regional magnetic anomalies, thereby enhancing shallow anomalies that reflect structural features of the Koldar massif. The TDR method facilitated the delineation of structural units that may correspond to faults and lineaments within the study area. Figure 11b clearly reveals elongated linear anomalies, predominantly trending north and northwest. The HD TDR highlighted magnetic anomalies with greater intensity, which may indicate the presence of concealed geological bodies. Although the general outlines resemble those produced by the TDR, the HD TDR revealed subtler variations in the magnetic field that might not have been detected through conventional data processing, making it a valuable tool for refining geological models (Figure 11c).

4.5. Gravic Conversions

The gravity field, converted into the Bouguer anomaly for the study area, reveals two prominent features: the Koldar positive anomaly and the Kyzylkaynar negative anomaly. The Koldar anomaly is located in the area of the eponymous mountains and occupies the central portion of the study area. It extends approximately 46 km in length and 33 km in width, with a predominant northwest orientation and a northeast-trending apophysis. The anomaly is characterized by medium intensity and is complicated by a series of superimposed local anomalies of higher orders, which correspond to intermediate and mafic intrusive bodies of the Koldar massif.

The Kyzylkaynar negative anomaly is situated in the northeastern part of the study area and also trends northwest. It corresponds to the surface exposure of intrusive rocks of granite–granodiorite composition (Figure 12a). While the Bouguer anomaly map can provide insights into the geometry and lithology of subsurface rocks, the observed anomalies are influenced by regional effects caused by deep-seated sources. To improve the interpretation of near-surface features, mathematical transformations were applied to the gravity data.

The TDR is a transformation that provides a more balanced enhancement of gravity field data, particularly useful for detecting the position and orientation of density contrasts (anomalies). It is especially effective for delineating the boundaries and spatial extent of anomalous bodies, offering a more robust approach than using the 1VD derivative alone. The TDR technique helps clearly define anomaly edges, as the tilt angle is typically zero over the center of an anomaly and changes markedly at its boundaries (Figure 12b).

The HD TDR is a combined derivative measure that captures variations in all horizontal directions. It enhances the detection of edges or boundaries of gravity anomalies in the horizontal plane. As HD TDR applies a horizontal derivative to the tilt derivative, it improves the resolution of anomaly boundaries and localization of geological features. This transformation yielded clearer identification of local maxima and the delineation of geological contacts (Figure 12c).

Figure 11. (a) 1VD, (b) TDR, and (c) HD TDR maps derived from magnetic data of the Koldar massif.

Figure 11. (a) 1VD, (b) TDR, and (c) HD TDR maps derived from magnetic data of the Koldar massif.

Figure 12. (a) Bouguer anomaly map, (b) TDR, and (c) HD TDR maps based on gravity data of the Koldar massif.

Figure 12. (a) Bouguer anomaly map, (b) TDR, and (c) HD TDR maps based on gravity data of the Koldar massif.

5. Discussion

This study develops the approaches laid out in our previous work [103]. Previously, we developed a remotesensing-based mineral exploration method aimed at delineating hydrothermally altered zones associated with porphyry copper mineralization at the Aktogay deposit in eastern Kazakhstan using Landsat-8 and ASTER satellite data. The reliability of the remote sensing results was confirmed by field studies, including analysis of the spatial distribution of geological lineaments and petrographic analysis of samples from alteration zones and ore mineralization. ASTER data effectively identified key hydrothermal zones (potassic, propylitic, argillite and iron oxide), which serve as indicators of potential mineralization zones. To quantitatively assess the reliability of the results based on spectral characteristics, 45 control points in the borehole core were studied. The kappa coefficient calculated to assess the agreement was 0.615, indicating a significant level of confidence beyond chance. Of the 45 control points, 32 (71.11%) confirmed the presence of the identified hydrothermal alteration [103]. The developed and validated methodology provided a reproducible workflow, which is applied in the present study. The key difference and novelty of the current work is the hyperspectral processing of hydrothermal alteration data (based on PRISMA data), the use of the SAM (Spectral Angle Mapper) method on WorldView-3 multispectral images, and the integration of four evidentiary layers: lineament density, hydrothermal alteration, magnetic and gravity anomalies. This made it possible to create a comprehensive map of promising copper mineralization zones in the Koldar massif.

The spatialgenetic relationship between hydrothermal ore bodies and faults is considered one of the most promising hypotheses in geology for mineral deposit prediction. The Earth’s crustal fault systems act as conduits for the migration of ore-bearing fluids and organic compounds, playing a crucial role in forming deposits or creating necessary conditions for mineral accumulation [104,105,106]. Thus, faults are a primary factor in mineralization, especially of hydrothermal origin.

The morphology of the Koldar intrusive massif is structurally controlled by fault systems associated with hydrothermal activity. The delineated lineaments (Figure 6a) reflect permeable structures that facilitated fluid migration from magmatic sources to ore-bearing structures in the upper crust. The Aktogay and Aidarly deposits formed under the influence of the sublatitudinal Aktogay fault system (280–310°), while the Kyzylkiya deposit was influenced by northwesterly (azimuth 70–80°) and northeasterly (azimuth 280–310°) trending faults [71,107]. Lineament analysis within the study area revealed three dominant orientations: sublatitudinal (270–280°), northwesterly (320–330°), and northeasterly (60–70°), as shown in the rose diagram (Figure 6c).

The lineament density map (Figure 6b) demonstrates a high concentration of faults and fractures within the Koldar massif. This indicates the presence of highly permeable zones and confirms the genetic link between fault tectonics and copper mineralization [108,109].

High-temperature hydrothermal fluids, enriched in potassium and hydrogen, cause significant mineralogical changes, forming characteristic potassic, phyllic, and propylitic zones in porphyry copper occurrences [85,110,111]. Minerals in these zones exhibit distinct spectral absorption features across the visible, VNIR, SWIR, and TIR (thermal infrared) electromagnetic spectrum ranges. These diagnostic features enable the effective detection and mapping of hydrothermal alteration zones using remote sensing data [102,112,113,114].

Considering the diverse spectral signatures of hydrothermal zones associated with porphyry copper systems, ASTER, PRISMA, and WorldView-3 data were utilized in this study. ASTER imagery showed strong correlation with petrographic data, while PRISMA imagery provided a more comprehensive view of hydrothermal alteration zones. As depicted in Figure 8, phyllic alteration, associated with sericite and muscovite (minerals typically formed by weakly acidic hydrothermal fluids), was identified in the central parts of the Aktogay, Aidarly, and Kyzylkiya deposits. These phyllic zones are encircled by propylitic alteration, linked to epidote, chlorite, and calcite minerals, indicating the influence of lower-temperature hydrothermal fluids.

WorldView-3 data processing, owing to its higher spectral and spatial resolution in the VNIR and SWIR bands, allowed for the identification of argillic and propylitic alteration zones at the Aidarly deposit and revealed a new prospective zone to the north (Figure 9). Argillic alteration is associated with clay minerals such as kaolinite and montmorillonite, indicative of intense hydrothermal activity, whereas propylitic alteration is characterized by elevated chlorite and epidote content, typically resulting from low-temperature metamorphic processes.

To suppress regional anomalies and enhance near-surface signals related to various structural features, as well as to delineate geological boundaries between different structural domains, geophysical data transformation was performed. Magnetic maps showed moderate magnetic susceptibility over the deposits due to the absence of strongly magnetic minerals. According to [115], alteration zones at the Aktogay deposit consist of potassium feldspar, biotite, albite, and quartzsericite, all of which exhibit weak magnetization. The results of magnetic and gravity TDR and HD TDR are in good agreement, delineating altered zones in the southern part of the Koldar massif and a linear anomaly extending from north to northwest. These results reflect the fractured nature of the massif, a key indicator of structural prospectivity for mineralization.

The main outcome of this study is a map of prospective zones for copper mineralization in the Koldar massif (Figure 13), generated through a comprehensive application of methods. Four evidential layers were used for its creation: lineament density, hydrothermal alteration from PRISMA data, and magnetic and gravity anomalies. All layers were normalized to a scale of 0 to 1 and integrated into ArcGIS. Each input layer was divided into ten classes and assigned an equal weight of 25% in the final analysis.

Highly prospective areas, totaling approximately 25 km² or 20% of the massif’s total area, were delineated (Figure 13a). The Kyzylkiya deposit in the northeastern part of the massif was identified as a prospective zone. The Aktogay deposit is characterized by a broad outer halo favorable for potential mineralization. An irregularly shaped prospective zone was also identified around the Aidarly deposit. Moreover, an anomalous zone located to the north (Figure 13b) is distinctly visible on the hydrothermal alteration map derived from WorldView-3 data using the SAM method (Figure 9).

Some highly prospective areas identified in the southern, northern, and western parts of the massif, given their geological setting, fall within Quaternary deposits. This is attributed to false-positive results obtained from PRISMA data, caused by anthropogenic factors: in the south, by a drying water body, and in the west, by sand dunes and salt flats due to their proximity to the eastern shore of Lake Balkhash.

Considering that this massif had not been thoroughly studied using satellite methods previously, the obtained results fill a significant knowledge gap, offering a comprehensive approach by integrating ground-based and satellite data. Unlike costly and labor-intensive ground surveys, satellite data can be acquired remotely, significantly reducing expenses and effort. Furthermore, most regional studies in this area were conducted in the 1990s using the technologies available at that time. Modern satellite data allow for the actualization and detailed refinement of previously obtained results, complementing them with highly precise and large-scale remote sensing.

The main advantages of this study lie in its comprehensive approach, combining ground-based and remote sensing data, and in the successful processing of high-quality WorldView-3 multispectral imagery, which enabled the mapping of a new prospective zone.

6. Conclusions

In this study, geophysical and satellite methods were utilized to construct a mineral prospectivity map for copper. Hyperspectral processing data expanded our understanding of hydrothermal alteration zones associated with porphyry copper deposits, building upon information obtained via ASTER. The SAM method, applied to WorldView-3 imagery, effectively mapped areas of argillic and propylitic hydrothermal alteration at the Aidarly deposit and identified a new prospective area to its north. Lineament density analysis confirmed the theory of fault tectonics as a controlling factor in copper deposit formation. The transformation of geophysical data, specifically magnetics and gravity, allowed for the removal of regional background noise and the clear delineation of geological features and structural unit boundaries.

The resulting map of prospective zones for porphyry copper mineralization successfully highlighted the Aktogay, Aidarly, and Kyzylkiya deposits, indicating the map’s high validation. Furthermore, an area adjacent to the Aidarly and Aktogay deposits was identified, suggesting the underestimated potential of these territories. This may indicate the need for more thorough investigation of these areas to refine geological characteristics and assess potential mineral reserves.

Future research will focus on deepening and detailing the obtained results through the use of high-resolution hyperspectral imagery, the implementation of machine learning to improve data clustering and determine the relevance of each method, the conduct of local geochemical and geophysical surveys to confirm the results of high-quality spectral analysis, and the creation of 3D models of prospective territories for a more comprehensive understanding of their geological structure.