A reconnaissance study for tracing and ordering new mineralisation zones using integrated remote sensing, GIS, and aeromagnetic techniques, west Allaqi-Heiani-Suture,Egypt

ABSTRACT

Reducing expenses is a critical aspect of natural resource explorations, especially in the reconnaissance stage. Accordingly, GIS algorithms using the free remote sensing data were used to rank the probable mineralised zones before any future implementations of the high-cost drilling and/or detailed geophysical programmes. Geological and tectonic settings, and being a historical mineral resource were motivations to choose the Allaqi-Heiani Suture (AHS), Upper Egypt, as a study area. Remote sensing data were processed for tracing the probably mineralised alteration zones, determining fractures and dominating structural systems, and mapping different geological units. Various alteration minerals including hydroxyls–carbonates, ferrous and ferric iron oxides were detected over different rock units using band ratio (BR) and Crosta techniques. Different geological units were distinguished using false colour composite (FCC) images. Moreover, an intense fractured system, demonstrating the area, was revealed by tracing lineaments using informative, spatially enhanced remote-sensed images. Fractures within the altered rocks represent 35% of the total fracturing system in the AHS to indicate a structural system controlling mineralisation occurrences. The processing of magnetic data suggested a deep/buried nature of these remotely detected lineaments/structures. Reduced to pole (RTP) and tilt derivative (TDR) maps detected the buried magnetic sources, whereas zero contours of TDR, Euler deconvolution, analytic signal and source parameter imaging (SPI) techniques traced buried contacts/structures and estimated their depths. A GIS strategy depending on the intensity of the fracturing system and the presence of more than one type of alteration was applied to potential-rank the prospective mineralised occurrences over AHS. The correlation between ranked zones with the old/known mines showed encouraging evaluation.

KEYWORDS:

• Remote sensing
• GIS
• geophysics
• alteration zones
• gold mineralisation in the Eastern Desert of Egypt
• Allaqi suture

1. Introduction

Upper Egypt and the Sudanese desert are called Nubian land; likewise, the rock belt along the western part of the Red Sea is called Nubian Shield. In the old Egyptian language, the word “Nubia” means gold (Habashi 2016), which indicates the importance of this desert as a gold source since ancient periods. Despite the knowledge (geological) simplicity of the ancient Egyptians, Romans, and early Arabs, they realised the close relationship between the presence of quartz veins and the narrow, rabid changes in lithology (alteration zones) with gold, silver and copper minerals occurrences. Therefore, gold production mainly depended on high-grade quartz veins and placer (alluvial) deposits through ancient eras. It would not be an exaggeration if we said that the ancients reached all the places that we consider today as sites for gold occurrences, but more than that, they reached places that are still rigid to reach today (for more, see Klemm and Klemm 2013). The historic mining activity is evidenced by the ancient house remains, stooped-out quartz veins, shafts, adits, mine tailings, hand hammers and crushed rocks (Klemm and Klemm 2013).

Recently and with the scientific progress, investigating mineralisations is conducted using satellite scenes interpretations and field geological studies to map the hydrothermal alteration zones and quartz veins, discriminating different rock units and determining the surface features/lineaments, while tracing the subsurface structures, estimating the depth and assessing the extension/rootedness and reserves of mineralised bodies are carried out using geophysical, and geochemical and drilling programmes (e.g. Ramadan and Sultan 2003; Akanbi and Ugodulunwa 2010; Mekkawi 2012; Zeinelabdein and El Nadi 2014; Pour and Hashim 2015; Wemegah et al. 2015; Sadek et al. 2015, 2020; El Arafy 2016; Hassan and Sadek 2017; Eldosouky et al. 2017; Gahlan and Ghrefat 2018; Amara et al. 2019; Salamba et al. 2019; Zoheir et al. 2019; Traore et al. 2020; Es-Sabbar et al., 2020; Saleh et al. 2020; Mekkawi et al. 2021; Gabr et al. 2022; ElGalladi et al. 2022; Araffa et al. 2022). The remote sensing and geophysical signal changes are subject to the spectral and physical properties that function in the mineralogical composition of various rocks. Therefore, the integration between the remote sensing and geophysical techniques was applied in this study by using Landsat 8 images and aeromagnetic data. The GIS algorithms were used to detect lineaments automatically and order the mineralisation potential zones. The west Allaqi-Heiani-Suture (AHS) represents the study area located downstream of Wadi Allaqi, South Eastern Desert, about 180 km southeast of Aswan city, Upper Egypt (Figure 1).

Figure 1. (a) Location of the study area. (b) A simplified geological map of the Egyptian Eastern Desert highlights the general geology and position of the study area (modified after Ramadan et al. 2001; Abdelsalam et al. 2003; Abdelsalam 2010; Johnson et al. 2011, 2017; Azer 2013; ElGalladi et al. 2022). (c) Detailed geological map of west Allaqi-Heiani Suture (AHS) (Kusky and Ramadan 2002; ElGalladi et al. 2022).

This manuscript tried to answer some addressed questions, like: Can we potential-ordering the probable mineralisation zones, within AHS, using the GIS algorithms and remote sensing data before any future drilling programmes? Does the surficial tracing of alterations and structures/lineaments, using geological surveys and remote sensing data, reflect the occurrence of deep mineralisations? Is the Allaqi district an encouraging future location for mineral extraction? Why? Where?

2. Tectonic, geological and drainage settings of west Allaqi-Heiani Suture (AHS)

One of the most obvious evidence of suturing is the presence of linear ophiolitic-decorated belts defining the position where the closure of oceans/basins took place (Abdelsalam and Stern 1996; Hamimi et al. 2019). These sutures/collisions are accompanied by plentiful granitic intrusions invading both sides of the accreted terranes/fragments (Kröner and Stern 2005) to represent a fertile environment for the hydrothermal solutions to deform/alter rock units and precipitate different ore bodies (ElGalladi et al. 2022) while the Egyptian Eastern Desert is a good example (Figure 1b). The Allaqi-Heiani Suture (AHS), study area, represents the western segment of the large, well-preserved arc–arc Allaqi-Heiani-Gerf-Onib-Sol-Hamed-Yanbu (AHGOSHY) suture within the northernmost of the Arabian-Nubian Shield (ANS) (Abdelsalam and Stern 1996; Abdelsalam et al. 2003; Johnson et al. 2011) (Figure 1b).

The basement rocks within west AHS are mainly represented by ophiolitic rocks, island-arc assemblages and granitic intrusions (Figure 1c) (CONOCO 1987; EGSMA 1996; Ramadan et al. 2001; Kusky and Ramadan 2002; Ramadan and Sultan 2003; Khalil et al. 2014; Eldosouky et al. 2017; ElGalladi et al. 2022). The ophiolitic rocks represent obducted remnants of the oceanic lithosphere and the oldest rock units within the AHS. Ophiolites are dismembered and distributed within the middle segment of West AHS. While Wadi Haimur, Gebel Shikeyit, Wadi Umm Araka, Wadi Um Ashira and Gabel Shilman are the main provinces of the recorded ophiolites. Ophiolitic rocks including serpentinites, talc-carbonates and metagabbros superimposed/thrusted over the island arc rocks (Kusky and Ramadan 2002). Metagabbros are very limited within the study area. Whereas the island-arc rocks are the most abundant units within the Wadi Allaqi area (study area) including considerable metasediment and metavolcanic rocks, while gabbro-to-diorite intrusions are less common. These intrusions are heterogeneous in composition and include gabbro, diorite, quartzdiorite and tonalite. The younger granites surround the study area in the north, east and south. The late orogenic granites are associated with mineralisation, like gold, radioactive and rare-metal mineralisations [e.g. Pb, Th and U in Wadi Um Geir–Wadi Allaqi (Saleh and Makroum 2003) and Pb, Ta, Zr, U, Nb and Y in the radioactive Umm Shilman granite (Bugrov 1972)]. Whereas Emam and Saad-Eldin (2011) suggested the source of Al, Pb, Zn, Ba and Rb elements, recorded in Wadi Allaqi samples, to be a result of weathered materials drained chiefly from the granites.

The drainage network reflects the topography and follows the elevation system over the study area. The stream flow is generally from the south-east (Wadi Allaqi), south-west (W. Gabgaba) and other Wadies in the north (such as Wadi Haimur, W. Um Ashira, W. Qulieb), which are combined at Khor-Allaqi to drain at Nasser-Lake (Figure 2a,b). Usually, placer deposits reflect the mineralised occurrences in different environments. Emam and Saad-Eldin (2011), Ramazanov and Ali (2011) and El-Makky and Sediek (2012) studied samples collected from different streams/Wadis within the study area. The geochemical analysis indicated the occurrence of different minerals and placer gold deposits within the collected samples. Thence, the drainage network in Figure 2a and 2b could be used to detect placer deposits, especially in zones where the stream becomes curved and at zones where various stream orders intersect. These zones represent places where the speed of the stream flow changes/decreases and represent a chance for the discharging or accumulation of the stream loads/minerals.

Figure 2. (a) Drainage system over west Allaqi district. (b) 3D view simulates the topography setting with the drainage network.

3. Material and methods

The gathering between (1) delineation of hydrothermally altered rocks that may be associated with mineralisation deposits, (2) detection of fracturing and weakness zones and (3) geological discrimination and mapping various rock units, is considered to be the best mineral exploration strategy (e.g. Sabins 1999; Gahlan and Ghrefat 2018; Araffa et al. 2022; ElGalladi et al. 2022). To achieve this optimum mineral exploration strategy, image processing techniques were applied for the corrected Landsat 8 OLI images (Figure 3), including

Figure 3. A flowchart scheme shows the processing techniques used for the exploration of new prospective mineralisation potential zones over the west Allaqi-Heiani Suture (AHS).

3.1. Delineation of hydrothermally altered rocks

3.1.1. Band ratio (BR)

Dividing or subtracting the digital numbers of two bands produces a new band (called; band ratio in the case of dividing) which highlights the spectral contrast and reduces the spectral similarity of the transformed/parent bands. Hydroxyl-bearing and carbonate minerals have strong absorption features in 2.08–2.35 μm corresponding to band 7 (2.1–2.3 μm) and intense reflectance in 1.55–1.75 μm parallel band 6 (1.56–1.66 μm) of Landsat 8. Therefore, SWIR bands of Landsat 8, or BR 6/7, are the most suitable ratio to discriminate hydroxyl-bearing alterations and carbonate minerals (Gupta 2003, 2018; Pour and Hashim 2015; El Arafy 2016; Banerjee et al. 2019; Zoheir et al. 2019; Traore et al. 2020; Araffa et al. 2022; ElGalladi et al. 2022).

Ferrous (Fe²⁺)-bearing minerals have absorption characteristics within B5 (0.845–0.885 μm) and high reflectance within B6 (1.560–1.660 μm). Thence, the OLI BR 6/5 is used for detecting the ferrous mafic rocks (El Arafy 2016; Banerjee et al. 2019; Zoheir et al. 2019; Traore et al. 2020; Araffa et al. 2022; ElGalladi et al. 2022). Moreover, BR 6/5 helps identify the weathered ferruginous rocks (gossans) produced by oxidation and leaching of sulphide mineralisation and represent a probably surface indicator of buried ore deposits (Sabins 1999; Haldar 2018) especially gold occurrences in the Arabian Nubian Shield, ANS (Ramadan and Sultan 2003; Zeinelabdein and El Nadi 2014; Gahlan and Ghrefat 2018; ElGalladi et al. 2022).

Ferric (Fe³⁺) irons (e.g. limonite, jarosite and haematite) tend to have high spectral absorption in the ultraviolet or blue region between 0.45 and 0.52 μm (corresponding OLI B2) and high reflectance characteristics near red zone between 0.63 and 0.69 μm (corresponding OLI B4) giving iron minerals its rusty reddish-brown colour (Gupta 2003, 2018; Pour and Hashim 2015; Araffa et al. 2022; ElGalladi et al. 2022). Thus, OLI BR 4/2 was used for mapping ferric-bearing minerals. For mapping, alteration minerals over west AHS, Sabins’ OLI BRs of 6/7, 6/5 and 4/2 were selected to sharpen hydroxyls and carbonates, ferrous- and ferric-iron oxides, respectively.

3.1.2. Thresholding and separating alteration zones

The threshold is an image segmentation technique that determines the inflection point or the threshold value where the spectral response critically changes. El Said et al. (2014), Fatima et al. (2017), Amara et al. (2019), Traore et al. (2020), Aboelkhair et al. (2020), Soliman et al. (2021) and Araffa et al. (2022) used threshold technique to separate and map alteration zones. Threshold has been applied for Sabins’ band ratios (Table 1) and for Crosta selected (of PCs and MNFs) eigenimages to extract alteration zones over west AHS.

Table 1. Statistical factors used for estimating the band ratios threshold values for mapping alteration minerals in Figure 4.

Alteration Minerals	Ratio	Min.	Max.	Mean	Standard Deviation	Threshold	Confidence %
Hydroxyls & Carbonates	BR 6/7	0	255	112.21	59.9	232.01	95%
Ferrous-Irons	BR 6/5	0	255	75.47	53.24	235.2	98%
Ferric-Irons	BR 4/2	0	255	96.55	70.65	237.85	95%

Figure 4. Threshold of band ratios 6/7, 6/5 and 4/2 to extract alteration zones: (a) Hydroxyl-bearing and carbonate minerals extraction, BR 6/7. (b) Extraction of ferrous-iron oxide minerals, BR 6/5. (c) Extraction of ferric-iron oxides minerals, BR 4/2. (d) The alteration minerals overlaid the natural colour image of Google Earth.

3.1.3. Feature-orientated principal component selection (FPCS) “Crosta” technique

The most optimum eigenvector (e.g. PC, MNF, ICA, etc.) for highlighting and mapping certain features (such as vegetation, alteration zones, etc.) is the one with the best fitting to the spectral features (bands) which are theoretically spectral sensitive for this target. The best-fitting eigenimage means the most spectral sensitive for the target, i.e. it is loaded by the highest contrast (positive and negative) eigenvalues for the reflectance and absorption bands of the target, to help in more sharpening the studied features. This technique is called FFPCS or “Crosta” (Crosta and Moore 1989). Crosta statistics were calculated for the PC and MNF components to sharpen the alteration zones within west AHS.

3.2. Lineament extraction and detection of fracturing and weakness zones

Detection of fracturing and weakness zones is one of the bases of the optimum triangle of mineral exploration strategy. Lineaments represent the weak zones and fractures which represent the path in which the hydrothermal fluids migrated/ascended through and deposit their loads (ore bodies) forming veins and alteration zones (El Arafy 2016; Eldosouky et al. 2017; Salamba et al. 2019; El-Desoky et al. 2021; Araffa et al. 2022; ElGalladi et al. 2022; Gabr et al. 2022).

3.2.1. Surface features, remote sensing data

3.2.1.1. Automatic lineament extraction

Lineament extraction automatically depends on variations in the brightness and colour tone, i.e. edge detection, of different features. So, as far as possible, increasing the spatial resolution of the used image will enhance the results of the lineament tracing process. Where most authors enhance the spatial resolution by pan-sharping and spatial-domain, edge detection, filtering of the used spectral images and hill-shading of the digital elevation models (e.g. El Arafy 2016; Amara et al. 2019; Es-Sabbar et al., 2020; ElGalladi et al. 2022; Araffa et al. 2022).

The automatic lineament process was carried to the most informative eigenimages including the PC1, which carries 95.8% of total variances of Landsat 8 OLI (1–7) bands (Table 2), the MNF1 image, the highest signal-to-noise image and the Global Digital Elevation Model (GDEM) of ASTER with 30 m spatial resolution. To improve the results of the lineament process, PC1 and MNF1 eigenimages were pan-sharpened and directional filtered, using a 3*3 neighbourhood convolution (kernel) matrix, in N-S (0°), NE-SW (45°), E-W (90°) and NW-SE (135°) directions, while the DEM image was hill-shaded in the same sun illumination angles (solar azimuth) using ENVI software. For extracting the geologic lineaments automatically, the LINE module algorithm of the PCI Geomatica software was used with the default parameters except for editing the threshold value for the edge gradient to become (50). The extracted lineaments, linear and curvilinear features, were exported from the LINE module algorithm as a vector layer to be statistically analysed and layout by the ArcGIS software. The rose diagram was plotted using RockWorks software to examine the orientation and identify the fracture trending system of the extracted lineaments.

Table 2. PCA eigenvector matrix of Landsat 8 used bands (B1–B7).

Landsat 8	Band 1	Band 2	Band 3	Band 4	Band 5	Band 6	Band 7	Variances in PCn (%)
PC 1	−0.0824	−0.1069	−0.2289	−0.4016	−0.4695	−0.55	−0.4952	95.8
PC 2	0.4632	0.498	0.5021	0.2524	0.0732	−0.3781	−0.2708	3.36
PC 3	−0.3814	−0.341	−0.0337	0.473	0.4997	−0.2144	−0.4666	0.61
PC 4	−0.1287	−0.0799	−0.0091	0.0289	0.1428	−0.7097	0.6724	0.19
PC 5	−0.1243	−0.1591	0.1218	0.6578	−0.7068	0.0252	0.1074	0.05
PC 6	−0.4358	−0.1833	0.8088	−0.3395	−0.0567	0.0623	−0.0018	0
PC 7	0.6394	−0.7484	0.1628	−0.0515	0.0368	−0.02	0.009	0

3.2.1.2. Lineament (line) density map

A density map calculates the concentration of lineaments per unit area (line frequency/unit area). It divides lineaments into low-density areas and high-density zones. The latter reveal highly fractured systems which are favourable media for the precipitation of the migrated/ascended hydrothermal solution loads and ore bodies deposition in the altered/fractured zones.

3.2.2. Subsurface structures, aeromagnetic data

Are these remotely detected lineaments/features reflecting the deep, buried, structures and tectonic settings or only relating to surface-weathered ones? The aeromagnetic data were processed to detect the magnetic (possible mineralised) sources and trace deep structures beneath the Allaqi-Heiani Belt to answer this question.

The aeromagnetic data of the Aero-Service (1984) survey covering the study area were corrected and processed to compute the total magnetic intensity (TMI) anomaly. TMI was reduced to the magnetic pole (RTP) to peak the magnetic anomalies directly over their subsurface causative sources. To highlight the edges of magnetic sources, the tilt derivative (TDR) technique was employed. TDR is positive/peak over the magnetic sources, negative outside and zero contoured at/near the edge of the source (Miller and Singh 1994). TDR was applied to the RTP map to detect the magnetic sources and trace the geological edges/contacts beneath the study area. Furthermore, the Euler deconvolution technique was utilised to follow the magnetic contacts using the structural index of zero (SI = 0) (Reid et al. 1990).

The analytic signal (AS) (Roest et al. 1992) is the square root of the sum of the first derivative squares in the vertical (dz) and horizontal (dx and dy) directions of the magnetic field (Equation 1(Equation. 1) $Analyticalsignal\left(AS\right)=\text{break}\sqrt{{\left(\frac{\mathrm{\partial }M}{\mathrm{\partial }x}\right)}^2+{\left(\frac{\mathrm{\partial }M}{\mathrm{\partial }y}\right)}^2+{\left(\frac{\mathrm{\partial }M}{\mathrm{\partial }z}\right)}^2}\left(nT/m\right)$(Equation. 1) ), so it is also known as the total gradient technique. AS is a helpful technique in mineral exploration as it highlights/peaks over the magnetic edges and the geological structures where the high and low magnetic susceptibility, of different rocks, get in contact and over fractures and shear zones where the magnetite-rich magma ascended and infilled these fractures (Roest et al. 1992; Akanbi and Ugodulunwa 2010; Mekkawi 2012; Wemegah et al. 2015; Mekkawi et al. 2021). While the analytic signal magnitude is a function of the magnetic susceptibility contrast of different rocks (Akanbi and Ugodulunwa 2010).

(Equation. 1) $Analyticalsignal\left(AS\right)=\text{break}\sqrt{{\left(\frac{\mathrm{\partial }M}{\mathrm{\partial }x}\right)}^2+{\left(\frac{\mathrm{\partial }M}{\mathrm{\partial }y}\right)}^2+{\left(\frac{\mathrm{\partial }M}{\mathrm{\partial }z}\right)}^2}\left(nT/m\right)$(Equation. 1)

where M is the total magnetic field, $\frac{\mathrm{\partial }M}{\mathrm{\partial }x}$: is the first horizontal derivative in the (x) direction, $\frac{\mathrm{\partial }M}{\mathrm{\partial }y}$: is the first horizontal derivative in the (y) direction, $\frac{\mathrm{\partial }M}{\mathrm{\partial }z}$: is the first derivative in the vertical (z) direction.

The analytic signal of the total magnetic anomaly (AS) was gridded according to (Equation 1(Equation. 1) $Analyticalsignal\left(AS\right)=\text{break}\sqrt{{\left(\frac{\mathrm{\partial }M}{\mathrm{\partial }x}\right)}^2+{\left(\frac{\mathrm{\partial }M}{\mathrm{\partial }y}\right)}^2+{\left(\frac{\mathrm{\partial }M}{\mathrm{\partial }z}\right)}^2}\left(nT/m\right)$(Equation. 1) ). Likewise, the analytic signal of the first vertical derivative (AS1) was calculated with (Equation 2(Equation. 2) $\left(AS1\right)=\sqrt{{\left(\frac{\mathrm{\partial }Mv}{\mathrm{\partial }x}\right)}^2+{\left(\frac{\mathrm{\partial }Mv}{\mathrm{\partial }y}\right)}^2+{\left(\frac{\mathrm{\partial }Mv}{\mathrm{\partial }z}\right)}^2}\left(nT/m2\right)$(Equation. 2) ).

(Equation. 2) $\left(AS1\right)=\sqrt{{\left(\frac{\mathrm{\partial }Mv}{\mathrm{\partial }x}\right)}^2+{\left(\frac{\mathrm{\partial }Mv}{\mathrm{\partial }y}\right)}^2+{\left(\frac{\mathrm{\partial }Mv}{\mathrm{\partial }z}\right)}^2}\left(nT/m2\right)$(Equation. 2)

where Mv is the first vertical derivative of the total magnetic field.

Depth (D) to the magnetic sources could be estimated using the ratio of the analytic signal of the total magnetic anomaly (AS) to the analytic signal of the first vertical derivative (AS1), under assumptions that anomalies are in two-dimension (2D), (Equation 3(Equation. 3) $D=\frac{AS}{AS1}\ast \left(N+1\right)\left(m\right)$(Equation. 3) ) (Saada 2015; Al-Badani and Al-Wathaf 2017; Mono et al. 2018; Saleh et al. 2020).

(Equation. 3) $D=\frac{AS}{AS1}\ast \left(N+1\right)\left(m\right)$(Equation. 3)

where D is the depth to the magnetic source, N is a structural index (Reid et al. 1990) which is directly related to a simple geometry of the magnetic sources. N = 0 is related to magnetic contacts, while N = 1 is for 2D narrow dikes or sills, pipes or horizontal cylinders have N = 2 and N = 3 is relative to spherical magnetic bodies.

Source parameter imaging (SPI) (Thurston and Smith 1997) is an automatic method that uses the complex analytic signal to estimate the depths to magnetic sources/edges without prior knowledge of the magnetisation direction. The accuracy of the SPI, based on a correlation between the calculated depths of SPI and real data from hole drilling, is ± 20% which is the same accuracy for the Euler technique (Aisabokhae et al. 2018; Meneisy and Al Deep 2021; Oguche et al. 2021). SPI tool of Geosoft software was used to compute the depths of geological contacts beneath the Allaqi district.

3.3. Geological discrimination and rock units mapping

Geological mapping is the third base of the triangle strategy used in this study for mineral exploration. While the band combinations of satellite bands are useful for this object. The true colour composite image (TCC) makes the recognition of the geological units easy as it enables the interpreter to see a familiar scene of rock units, as felsic rocks appear bright, while mafic appear dark, deep water is seen as blue, vegetation is green and so on. While false-colour composite (FCC) images can emphasise other features out/not only within the visible wavelength ranges of TCC image. The OLI (B2:B7) and TIR (B10 and B11) bands, in addition to OLI band ratios (BR), PCA (PC1:PC5) and MNF (MNF1:MNF6) eigenimages were included in the combination building process of different coloured FCC images to identify different rock units.

4. Results and discussion

4.1. Mapping alteration zones

4.1.1. Threshold of band ratio (BR)

The threshold technique was applied for the Sabins’ ratios 6/7, 6/5 and 4/2 to separate hydroxyl-bearing and carbonate minerals, ferrous mafic rocks and ferric iron oxides within west AHS (Figure 4a–c; Table 1) respectively. The threshold took place after clipping the Saharan cretaceous deposits (left zone of the study area, black in Figure 4) and masking the Wadi and quaternary deposits, inside the study area, to prevent conflation with alteration zones (Figure 4).

4.1.2. Application of FPCS “Crosta” technique to PCA images

For extracting hydroxyl-bearing alterations (e.g. clay minerals, mica) and carbonates (e.g. marbles, dolomites), Landsat 8 bands 6 (intense reflectance) and 7 (strong absorption) were selected to be examined overall PCs to get the optimal PC with the highest contrast between these two bands. The highest contrast means the best highlighting of the hydroxyl-bearing features. Principal component (PC4) was chosen to highlight hydroxyls as it has the highest contrast between bands 6 and 7 (Table 2). The PC4 is loaded by a negative eigenvalue (−0.7097) for band 6 (reflectance band) and a positive loading (+0.6724) for band 7 (absorption band). It is expected that PC4 will highlight hydroxyls with dark-tone colours. The PC4 image was inverted (by multiplying the PC4 band by −1) to reverse the brightness and highlight hydroxyls with bright colours to make the identification and thresholding process easier (Figure 5a). For mapping the hydroxyl–carbonate-bearing minerals, the thresholding process was applied to the inverted PC4 according to a confidence 98% (Figure 5a):

Figure 5. (a) Extraction of hydroxyl-bearing and carbonate minerals by thresholding the inverted PC4. (b) Extraction of iron oxide minerals by thresholding the inverted PC5. (c) Extraction of hydroxides and iron oxides by thresholding the inverted (PC4+ PC5). (d) Crosta’s FCC image of inverted PC4, inverted (PC4+ PC5) and inverted PC5 as RGB.

(Equation. 4) $\mathrm{T}\mathrm{h}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d}=\mathrm{M}\mathrm{e}\mathrm{a}\mathrm{n}+3\ast \left(\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{d}\mathrm{e}\mathrm{v}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\right)=\text{break}87.259+3\ast \left(53.197\right)=246.85\mathrm{a}\mathrm{t}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{f}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}98\mathrm{\%}$(Equation. 4)

For identifying iron oxide minerals, Landsat 8 band pairs of 6 and 5, 5 and 4 and 4 and 2 are the choices for applying the FPCS technique. PC5 was chosen to highlight ferrous-bearing minerals as it carries a negative loading of the absorption band (B5; −0.7068) and a positive eigenvalue of the reflector band (B6; +0.0252) with good contrast between bands 6 and 5 (Table 2). PC5 is also suitable for mapping bands B5 (−0.7068) and B4 (+0. 6578) with high contrast between them (Table 2). Moreover, PC5 is sensitive for bands B4 (+0.6578) and B2 (−0.1591), which sharpen the ferric-iron oxides (Table 2). Therefore, PC5 is the choice for sharpening iron oxides which expected to be highlighted with dark-coloured. The PC5 image was inverted to make the mineralised pixels appear with a bright tone to make the identification and threshold process easier (Figure 5b).

PC5 was thresholded to convert the image into two colours (binarization), the mineralised zones with colour and non-mineralised as a background (black or non-coloured), then we export the mineralised areas (pixels) as a shapefile (Figure 5b). The threshold equation was applied at confidence 98% (Figure 5b):

(Equation. 5) $\mathrm{T}\mathrm{h}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d}=\mathrm{M}\mathrm{e}\mathrm{a}\mathrm{n}+3\ast \left(\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{d}\mathrm{e}\mathrm{v}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\right)=\text{break}84.822+3\ast \left(52.613\right)=242.661\mathrm{a}\mathrm{t}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{f}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}98\mathrm{\%}$(Equation. 5)

The inverted PC4 (Hydroxyls image) and inverted PC5 (Iron image) were summed [(-PC4)+(-PC5)] to get the iron-hydroxyl (H + F) image (Figure 5c). The iron-hydroxyl image was also thresholded at confidence 95% (Figure 5c):

(Equation. 6) $\mathrm{T}\mathrm{h}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d}=\mathrm{M}\mathrm{e}\mathrm{a}\mathrm{n}+2\ast \left(\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{d}\mathrm{e}\mathrm{v}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\right)=99.136\text{break}+2\ast \left(66.093\right)=231.322\mathrm{a}\mathrm{t}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{f}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}95\mathrm{\%}$(Equation. 6)

The inverted PC4 (Hydroxyls image), [(-PC4)+(-PC5)] (H + F image) and inverted PC5 (Iron image), respectively, were used to build the Crosta alteration image which is a false-colour composite image of PCs that highlights the alteration zones over west AHS (Figure 5d).

4.1.3. Application of FPCS “Crosta” Technique to MNF images

Bands 6 and 7 were examined to map the hydroxyl-bearing and carbonate minerals as they are theoretically spectral responses for the hydroxyls and carbonates. MNF3 was chosen as it shows a high positive loading (+0.427) for B6 (reflectance band) and high negative loading (−0.0479) for B7 (absorption band) (Table 3). MNF3 highlights the carbonates and hydroxyls with a bright tone. MNF3 was thresholded at a confidence of 98% (Figure 6a):

Figure 6. (a) Extraction of hydroxyl-bearing and carbonate minerals by thresholding MNF3. (b) Extraction of iron oxide minerals by thresholding MNF4. (c) Extraction of hydroxides and iron oxide minerals by thresholding MNF3 + MNF4 image. (d) Crosta alteration image, FCC image of MNF3, (MNF3+ MNF4) and MNF4 respectively as RGB.

(Equation. 7) $\mathrm{T}\mathrm{h}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d}=\mathrm{M}\mathrm{e}\mathrm{a}\mathrm{n}+3\ast \left(\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{d}\mathrm{e}\mathrm{v}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\right)=143.\text{break}654+3\ast \left(35.84\right)=251.174\mathrm{a}\mathrm{t}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{f}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}98\mathrm{\%}$(Equation. 7)

Table 3. Eigenvectors of MNF components of Landsat 8 used bands (B1-B7).

Landsat 8	Band 1	Band 2	Band 3	Band 4	Band 5	Band 6	Band 7
MNF1	−0.3594	0.4405	0.7147	0.0006	0.3957	−0.0079	0.097
MNF2	0.8742	0.0444	0.3129	−0.2123	0.1724	−0.2471	0.0103
MNF3	0.2962	0.4777	0.0485	0.6225	−0.3312	0.427	−0.0479
MNF4	−0.0292	−0.6265	0.4381	0.5081	−0.0332	−0.1365	−0.3699
MNF5	0.133	−0.3098	−0.1413	0.1818	0.649	0.5465	0.3367
MNF6	−0.0079	−0.2864	0.3747	−0.1769	−0.5214	0.2105	0.6557
MNF7	0.0187	−0.0723	0.1915	−0.4949	−0.0992	0.6284	−0.555

Bands 6 and 4 were selected to be examined over all MNF images for mapping iron oxides. MNF4 was found to be suitable for mapping iron-bearing minerals as it carries a positive loading for band 4 (+0.5081) and a negative value for band 6 (−0.1365). Likewise, MNF4 was found to be helpful in highlighting the ferric iron minerals as it shows a high positive loading for the reflectance band (B4; +0.5081) and a high negative loading for the absorptive band (B2; −0.6265). The mineralised pixels will appear with bright pixels (Table 3). To extract the iron oxides, the MNF4 eigenimage was thresholded at confidence 95% (Figure 6b):

(Equation. 8) $\mathrm{T}\mathrm{h}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d}=\mathrm{M}\mathrm{e}\mathrm{a}\mathrm{n}+2\ast \left(\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{d}\mathrm{e}\mathrm{v}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\right)=137.\text{break}97+2\ast \left(50.175\right)=238.32\mathrm{a}\mathrm{t}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{f}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}95\mathrm{\%}$(Equation. 8)

The eigenimages of MNF3 (H-image) and MNF4 (F-image) were summed to obtain the possible occurrence (H + F) image of both iron-oxide and hydroxyl-bearing minerals. The mineralised areas (pixels) of the summed image were thresholded at a confidence of 95% (Figure 6c):

(Equation. 9) $\mathrm{T}\mathrm{h}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d}=\mathrm{M}\mathrm{e}\mathrm{a}\mathrm{n}+2\ast \left(\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{d}\mathrm{e}\mathrm{v}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\right)=144.\text{break}86+2\ast \left(39.736\right)rm=224.332\mathrm{a}\mathrm{t}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{f}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}95\mathrm{\%}$(Equation. 9)

The FCC, Crosta alteration image, of MNF3 (hydroxyls image), MNF3+ MNF4 (H + F image) and MNF4 (Iron image) as RGB, respectively, were used to highlight the hydrothermal alteration zones over the study area (Figure 6d). The hydroxyl-bearing minerals are delineated by yellow to golden, bright orange. Pixels of Wadi Allaqi and Wadi Gabgaba are highlighted by yellow, golden colour, to indicate that alluvial deposits cover these Wadis. While the iron oxide areas are distinguished by the cyan to bright blue pixels. Also, Wadis with quaternary deposits appeared in cyan, so they were masked in the threshold process (Figure 6a–c), to prevent conflation with iron-bearing alterations. White to creamy colours (highest digital numbers) characterised the altered zones with both iron and hydroxyl-bearing minerals which are the most promising areas for mineral exploration (Figure 6d).

4.2. Lineament extraction and detection of fractures and weakness zones

4.2.1. Surface features

4.2.1.1. Detection of the predominating surface trend systems

Lineaments of the pan-sharpened, directional filtered PC1 and MNF1, and the hill-shaded DEM images (in 0, 45, 90 and 135 directions) were combined into a single vector image (Figure 7a). A large number of lineaments was traced (106,542 features) with a total length 106,458.11 km to indicate a highly fractured, tectonized, system demonstrating the study area and refer a fertile zone for mineralisation occurrences.

Figure 7. (a) Remotely sensed combination of lineaments extracted automatically from the composition of enhanced PC1, MNF1 and DEM images. (b) Remote sensed lineament density map of west AHS. (c) Lineaments orientation rose diagram. (d) Lineaments length rose diagram. (e) Basic statistics of the total automatically extracted lineaments.

To examine the controlling structural trend/orientation system over the west AHS, the bearing rose diagram was plotted. The rose diagram of the total automatic remotely sensed lineaments, of PC1, MNF1 and DEM showed the main sets of WNW-ESE, ENE-WSW and NE-SW as the predominant fracture systems (Figure 7c).

4.2.1.2. Lineament density map

The line density map of the study area is shown in Figure 7b. High-density zones are highlighted by red, orange and yellow colours which mainly appear in the southern, western side, western south and eastern north corners of the study area (Figure 7b). High-density contours characterise the strongly sheared ophiolites and listvenites as Haimur gold mine area and the scattered, fractured ophiolites as Hariari gold mine in north and Nagib gold mine in the middle east of study area. Likewise, the strongly fractured metavolcanic rocks within Umm Garayat, Nile Valley Blocks, Atshani and Marahaib gold mining area in the west middle of Allaqi-Heiani Suture are differentiated by highly line-density contours which also characterise the deformed/altered granitic rocks within the south at Filat gold mine, and northern part of west AHS (Figure 7b).

4.2.1.3. Percentage and orientation of the hydrothermal alteration system

To detect the surface orientation systems associated with the mineralised occurrences, the remotely sensing extracted lineaments were limited to alteration-bearing zones only (after masking unaltered areas, and the Saharan Cretaceous deposits in the west and the recent alluvial deposits of Wadi Allaqi). The lineaments of hydroxyl-bearing altered zones were separated as a single layer (Figure 8a). Also, lineaments of iron-oxides altered zones separated as an individual shapefile (Figure 8b). The number of lineaments of hydroxyl-bearing alterations equals 12,924 lineaments (Figure 8a) (represent 12.13% of, 106,542, the total remotely sensing extracted lineaments; Figure 7a). While lineaments of the iron oxide alterations are 24,804 (Figure 8b) (represent 23.29% of all extracted lineaments; Figure 7a). The total percentage of features and fractures of altered rocks (hydroxyl and iron oxide-bearing alterations) represents 35% of the total fracturing system over the west AHS. This high percentage also means the mineralisation and alteration systems are positively correlated with the fracturing system (structurally controlled mineralisation) in Allaqi-Heiani Suture which represents a suitable environment for gold, copper and other mineralisation occurrences.

Figure 8. Automatically, remotely sensing, extracted lineaments for alteration zones: (a) Lineaments of hydroxyl-bearing zones. (b) Lineaments of iron-oxide altered zones. (c) Lineaments orientation rose diagram of hydroxyl-bearing zones. (d) Lineaments length rose of hydroxyl altered zones. (e) Lineaments trend rose diagram of iron-oxides. (f) Lineaments length rose diagram of iron-oxide altered zones.

Trends of hydroxyl-bearing alterations were examined by plotting a bearing rose diagram which shows the WNW-ESE, ENE-WSW and NE-SW, especially N45E as the main trend systems (Figure 8c). Likewise, the orientation system of the iron oxide-bearing alteration zones shows the same main sets of hydroxyl alterations in the WNW-ESE, NE-SW, especially N45E, and ENE-WSW, respectively (Figure 8e).

In general, the alterations trends are following the general, observed, structure system of west AHS with predominating trends in WNW-ESE, ENE-WSW and NE-SW, especially N45E, fracture systems, respectively (Figure 7c). This interconnection between the orientation systems of altered zones and whole fracture system indicates the structurally controlled mineralisation occurrences in the west Allaqi-Heiani belt.

4.2.2. Subsurface structures and depth to buried magnetic sources

To ensure that the surficial, remotely, traced lineaments are related to deep magnetic bodies and a buried structure system, analysing of aeromagnetic data took place. The RTP map reflects the mineralogical composition and the magnetic response of the buried rock units. The high anomalies in the RTP map are related to the altered ultramafic rocks (serpentinites), mafic metavolcanics and granitic intrusions (Figure 9a). Serpentinites represent the elongated (and bent) high magnetic anomalies/belts, west the study area, taking the NW-SE and E-W directions, respectively, as the main trends. However, the ultramafic rocks of Gabel Haimur anticline (mainly embedded within metasediments and metavolcanics; middle west of the study area) and some patches within the elongated serpentinite belts didn’t show high magnetic anomalies over the airborne RTP map, why? Although serpentinisation of ultramafic rocks increases the magnetic susceptibility (mainly due to the formation of magnetite mineral during serpentinisation), however carbonation of serpentinites/ultramafics decreases the magnetic susceptibility significantly (mainly due to the destruction of magnetite during carbonation and increasing the amount of magnesite minerals forming talc-carbonates and quartz) (Tominaga et al. 2017; Mitchinson et al. 2020; Cutts et al. 2021; Godard et al. 2021). Therefore, the high degree of hydrothermal carbonation of ultramafic/serpentinite rocks within the Haimur area resulted in decreasing the recorded magnetic readings. Such an explanation is reinforced by the previous field studies which described serpentinites within Haimur area as highly deformed/altered serpentinites and silicified and carbonatized serpentinites (talc-carbonates and listvenites) (Emam and Zoheir 2013; Zoheir and Emam 2014; Abd El Monsef and Abdelnasser 2020; Araffa et al. 2022). While metavolcanics within the Umm Garayat mines area show a positive magnetic anomaly expected to extend deeper due to deeply, mineralised-rich, Volcanogenic Massive Sulphide (VMS) deposits settled beneath the area (Ramadan and Sultan 2003; ElGalladi et al. 2022). Meanwhile, the highest magnetic readings are related to the older and younger granitic intrusions (like granodiorites, and monzogranites and microgranites) with distinctive oval shapes in the north, middle and south of the study area (Figure 9a). The geological correspondence between granitic intrusions with high magnetic readings may indicate iron/magnetite rich content of these granites (typically biotite-rich [K₂(Mg,Fe)₆(Si₆Al₂)O₂₀(OH)₄] granites). The biotite-rich granites type was confirmed by Saleh and Makroum (2003) at Gabel Filat, south of the study area, and by Emam and Radwan (2021) at Nasb-Zurar area, northwest of the study area (north Um Ashira).

Figure 9. (a) RTP map. (b) Tilt derivative (TDR) map. (c) Zero contours of TDR map to outline the edges of magnetic anomalies, with their rose diagram. (d) Euler solutions estimated the depths to magnetic contacts at shallow depths (≤−150 m), with their rose diagram (ElGalladi et al. 2022).

The tilt derivative (TDR) map sharpens the magnetic anomalies (Figure 9b). Whereas zero contours of TDR (Figure 9c) highlight the edges of these anomalies and thence detect the structures/boundaries. The heavy numbers/distributions of the TDR zero-contours reflect the highly fractured/structured system buried beneath the AHS.

Likewise, ElGalladi et al. (2022) traced the subsurface contacts using the structural index (SI = 0) of the Euler deconvolution technique beneath the study area. The Euler solutions were restricted to the shallow depths (≤–150 m) to define the contacts/edges of residual magnetic (possible mineralised) sources (Figure 9d). These tracked contacts also revealed an intensity crowded, fractured zone, and showed high similarity with TDR zero-contours. The main structural trends of these subsurface/magnetic lineaments referred to the N-S (NNE-SSW and NNW-SSE), N45W and E-W directions as the predominant trend systems. These magnetic trends exhibit a good correlation with the surface remotely sensed lineaments.

The analytic signal maps of the magnetic anomalies (AS; Figure 10a, applied to the RTP map) and of the first vertical derivative (AS1; Figure 10b) maxima over the magnetic contacts and outlined the edges of magnetisation sources (Roest et al. 1992). AS and AS1 highlighted the granitic intrusions like area in the north (microgranites and monzogranites around Hariari mine), in the south (monzogranites around Filat mine) and above the granodioritic rocks in the middle (near Abu Swayel mine). Also, the AS and AS1 anomalies peaked over the magnetic transitions like areas of contact between the NW-SE belts of serpentinites and mafic metavolcanic rocks with the adjacent metasediments and felsic metavolcanics (like the zone around Um Ashira mine).

Figure 10. (a) Analytic signal map of the total magnetic field (AS). (b) Analytic signal of the first vertical derivative (AS1). (c) Depth estimations to the magnetic sources using the analytic signal technique. (d) Depth solutions to the magnetic sources/contacts, beneath AHS, using the SPI technique.

Depth estimations to the magnetic sources using the analytic signal technique were calculated by dividing the AS by AS1 (i.e. N = 0 in Equation 3)(Equation. 3) $D=\frac{AS}{AS1}\ast \left(N+1\right)\left(m\right)$(Equation. 3) to highlight the magnetic contacts/faults beneath the AHS. Depths range from shallower rocks (around −23.3 m) to the deeper sources (at −620.5 m and deeper) (Figure 10c). Likewise, the Source Parameter Imaging (SPI) technique outlined a general depth image of the magnetic sources under the study area (Figure 10d). SPI results approximately show high similarity with the contacts and depths deduced from the analytic signal method. The SPI shows a depth variation from −41.8 to −592.5 m with an average depth equal to −192.7 m, which indicates the near-surface/shallow nature of the magnetic/mineralised sources and the buried structures beneath the Allaqi-Heiani belt.

Moreover, if you know that, the total solutions/depths calculated by the SPI technique reached 557,729 points (magnetic sources) beneath the study area. Meanwhile, the number of sources with a depth shallower than >−500 m equals 533,778 sources with a percentage of 95.7%, whilst sources deeper than <−500 m equal 23,950 sources with a percentage of 4.29% (Figure 11). These numbers indicate the shallow occurrences of the magnetic sources while the large variations in the percentages between the numbers of shallow and deeper sources (95.7%: 4.29%, respectively) could indicate a ductile–brittle transition zone (DBZ) where fractures at depth are fewer and increase upward at/near the surface (Figure 11, and Figure S1 attached as an online resource; available to download in a supplementary file). Such brittle–ductile shear zone was recorded at surface structures and deformed minerals/veins in the Allaqi district and is mainly associated with the development of auriferous quartz veins (El-Kazzaz 1995; El-Kazzaz and Taylor 2001; Zoheir 2007; Zoheir and Klemm 2007; Zoheir et al. 2017). This DBZ could interpret why the residual magnetic and remote sensing maps traced large numbers of lineaments/fractures and explain why the Allaqi district is rich with auriferous quartz veins and hosts altered, mineralisation-rich, zones.

Figure 11. (a) 3D gridding, (b) 3D symbol simulation, of the depths to magnetic sources/contacts estimated using the SPI technique which suggests a ductile–brittle transitional/deformational zone (DBZ).

The anomalies/solutions of analytic signal and SPI maps (in Figure 10c,d) show high similarity to the magnetic bodies of TDR map (Figure 9b) and their edges (zero contours of TDR map, Figure 9c) and further offer the depth estimations for these contacts/sources. These contacts/sources mainly take the NW-SE and NNE-SSW as the main trend systems. These anomalies could represent the lithological structures or magnetite-rich magma that ran upward to infill fractures carrying possible bearing mineralised sources. While the known mines within the area show good examples of the mineralised bearing anomalies in these maps.

The intensity and trend system of the traced magnetic/subsurface contacts revealed a highly fractured area and reported a strong tectonic, deformation, history along the AHS life. While the very shallow estimated depths (−23.3 m: −620.5 m) and the small scale of these depth variations (~0.5 km of range variation) revealed that the magnetic (probably mineralised bearing) sources/structures occurred at the near-surface to represent a very encouraging zone for the mineral exploration and indicate the efficiency of surficial tracing of these structures/lineaments using the geological surveys and remote sensing data.

4.3. Geological discrimination and mapping of different rock units

The true colour image (TCC) was obtained by combining OLI- B4, B3 and B2 as RGB, respectively. The natural colour image was pan-sharpened using the high spatial resolution B8 to enhance the ground resolution of the image (Figure 12a). False colour images (FCC) were created to differentiate between various rock units and match the mutual units that have the same spectral response for the bands used in the various false colour images.

Figure 12. Landsat 8 band composition images over the west AHS. (a) Pan-sharpened natural (true) colour of OLI band composition as Red: B4, Green: B3 and Blue: B2. (b): (g) False colour composite (FCC) images using OLI and TIRS bands, band ratios and PCA images. (Bands included in the FCC composition, in the RGB channels, are labelled over each image).

Figure 12. (Continued).

Initially, band ratio 6/7 was executed for mapping clay and other hydroxyl alteration minerals in addition to carbonates. Also, band ratio 5/7 was useful in identifying clay minerals. Band ratios 6/5, 6/4 and 5/4 were useful in recognising ferrous, mafic iron oxide rocks. Ferric-iron oxide minerals were sharpened by using band ratio 4/2. Also, band ratios resulting from the dividing of bands in visible wavelength region such as BR 4/3 and BR 3/2 were useful for distinguishing ferric iron rocks. Band ratio 6/2 was purposive in recognition of monzogranite and microgranite (granites) over the study area.

PC4 and MNF3 were fruitful in differentiating hydroxyl-bearing alteration, clay and carbonate minerals (similar to the OLI band ratio 6/7). PC5 and MNF4 helped characterise iron oxide minerals. PC3 and MNF5 were also useful in identifying altered rocks.

Thermal infrared (TIR) bands (B10 and B11) of Landsat 8 helped identify silicate, and quartz units and differentiate the mafic/felsic rocks which have various spectral responses for the thermal infrared (TIR) bands.

The FCC images resulting from the combination of different greyscale images according to the previously mentioned aspects are shown in (Figure 12, and Figure S2 attached as an online resource; available to download in a supplementary file). The extension of different geological units was revealed by the band composite images which correlated and labelled according to the previously published geological maps over the study area, such as the geological maps in Table 4.

Table 4. Geological maps that were used for correlation and labelling the Landsat 8, TCC and FCC, images over the west Allaqi-Heiani Suture (AHS).

Geological Map	Reference
West Allaqi-Heiani Suture area	CONOCO (1987); EGSMA (1996); Kusky and Ramadan (2002); Abdelsalam et al. (2003); Zoheir et al. (2018); Eldosouky et al. (2017).
Wadi Shalman area (east central study area)	Abdeen and Abdelghaffar (2008); Azer (2013).
Wadi Haimur area (west central study area)	Khalil et al. (2014).
Haimur gold mine area	Emam and Zoheir (2013); Zoheir and Emam (2014); Ramazanov and Ali (2012).
Um Garayat (Umm Qurayyat) gold mine area	El-Makky and Sediek (2012); Zoheir et al. (2018).
Marahib (Marahiq) gold mine area	Emam et al. (2014).
Um Ashira gold mine area	El-Desoky et al. (2021); Emam and Radwan (2021).

4.4. Order ranking of the prospective mineralisation zones over AHS

Previous magnetic results confirmed that the remote sensing detected lineaments/features are related to the buried structures. Thence, the surface, remotely sensing, traced lineaments may be used for investigating new probable mineralised zones along the Allaqi-Heiani Belt.

Moreover, the results concluded by applying remote sensing techniques show that

• 1) By integration between the remotely sensing extracted lineaments, lineament density map (Figure 7a, b respectively) and the known gold mines, overlying the maps, it is noticed that there is a positive correlation between the high lineament density areas and the known gold mineralisation sites.

• 2) Also, by superimposing the extracted alteration zones (hydroxyls and iron-oxides extracted from the threshold of band ratios and Crosta images) on the lineament density map, it is noticed that there is a positive correlation between the high lineament density areas and the extracted alteration zones (Figure 13a).

  Figure 13. (a) Superimposing image combines hydroxyl-bearing, iron-oxide alterations and fracture system, the highest earlier three lineament density contours, all over a real colour image of Google Earth. (b) Remotely sensed ranking of the prospective mineralisation zones in west AHS, Diagram, c represents a legend of the probable mineralisation categories. (c) The pyramid scheme describes the ranking order of prospective mineralisations according to the existence of alteration zones and the structural control system. The top of the diagram (red colour) represents the highest ranking (first order) of the probable mineralisation occurrences, while the bottom is the expected lowest potential ranking.

  

• 3) The heavy lineaments, (106,542 features) with a total length of 106,458.11 km, extracted from remote sensing data (Figure 7a), reinforced by magnetic results, that are oriented in different directions (Figure 7c), indicate a highly fractured system dominated the AHS area and reflect the intense density of the subsurface structures.

• 4) Moreover, Araffa et al. (2022) detected listvenite alterations within the Haimur gold mine area, west middle of the study area, using remote sensing techniques as the highest fractured and the most heterogeneous zones including admixtures of various alteration minerals as carbonate-, hydroxyl- and iron-oxide minerals within serpentinite rocks. In Wadi Allaqi, gold is enriched in listvenite alterations up 400–1000 times in comparison to its occurrence in the original serpentinites (Azer 2013).

These concluded results prove the big role of tectonic and structural events in the occurrence and distribution of mineralisations over the west Allaqi-Heiani belt (structurally controlled mineralisation zone). This may be a key for discovering new mineralisation potential zones in the west AHS.

From the previous results, the probable mineralised zones in AHS were ranked, using remote sensing data, according to three factors: (1) first, the existence of hydroxyl-bearing minerals and/or carbonates; (2) second, the presence of iron-oxide-bearing minerals; and (3) the third factor is the intensity of fracturing system controlling the mineralised zone (obtained from lineament density map) as the probability of mineralisation occurrence increases at intense fracturing zones.

4.4.1. The prospective mineralisation GIS-ranking strategy

1. The highest order of this tri-factor ranking pyramid represents areas in which hydroxyl-bearing/carbonates minerals (H) intersect iron-oxide bearing (Fe) with the highest fracture system (highest lineament density).

2. The second order is areas with intersection between hydroxyls and iron oxides with the second-highest fracture system (second-highest lineament density).

3. The third order is the altered zones in which hydroxyls and iron oxides cross the third-highest fracture system (third-highest lineament density).

Only the early highest three fracture systems (highest three density areas) were included in this classification.

• (4) In the fourth order, the areas with the intersection between one alteration type (hydroxyl or iron oxide) with the earliest fracture system (highest lineament density), for example, altered zones with hydroxyl-bearing minerals that cross the highest fracture system (highest lineament density) represent this class, which is equal to iron oxide-bearing areas with the highest fracture system (highest lineament density) order in this category.

• (5) In the fifth grade, altered areas with hydroxyls or irons bearing intersect the second-highest fracture system.

• (6) In the sixth class, hydroxyls or irons alterations cross the third-highest fracture system (third-highest lineament density).

• (7) In the seventh group, areas with the intersection between the two extracted alteration types, hydroxyls with iron oxide-bearing areas, with medium fracture system (medium contoured areas in the lineament density map) gradually to the lowest fracture areas.

• (8) In the eighth ranking, areas with medium to the lowest, gradually, fracture system that crosses only one type of the extracted alteration type, hydroxyls or iron oxide altered zones.

The remotely sensed ranking of the prospective mineralisation potential areas in the west Allaqi-Heiani Suture according to these three factors: hydroxyl-bearing minerals, iron-oxide altered zones and density of fracture system is shown in (Figure 13b) and diagram in (Figure 13c). Also, the ranking process was zoomed in for the known gold mines in west AHS and it matches very well (Figure 14). Therefore, the highly ordered areas could represent encouraging provinces for any future drilling programmes and mining exploitations.

Figure 14. Comparison between the remotely sensed ordering with the known gold mines in west AHS, overlie Google Earth images: (a) Haimur gold mine. (b) Nile Valley Blocks and Um Garayat gold mines. (c) Marahib gold mine. (d) Atshani gold mine. (e) Nagib gold mine and area of Abu Swayel copper mine. (f) Filat gold mine. (g) Um Ashira gold mine. (h) Hariari gold mine. Diagram, c, in Figure 13, represents a legend of the colours in this figure (Figure 14) and describes the ranking order of the prospective mineralisation occurrences.

Particularly, hydroxyl (H) alterations and iron oxides (Fe), extracted from BR and Crosta eigenimages of PCA and MNF (Figures 4–6, respectively) were merged into a single layer for each alteration type (i.e. separated H-shapefile and Fe-shapefile). The lineament density map (Figure 7b) was converted to dissected shapefiles, every class with its shapefile (i.e. the highest lineament density, the second, the third and so on, areas become individual/separated shapefiles). By using the Intersect algorithm of ArcGIS software, the shapefiles of hydroxyl (H), iron oxide (Fe) and lineament densities were used as input for the algorithm to order the prospective mineralisation zones mentioned above.

5. Conclusion

Wadi Allaqi district is a historical gold source; however, it contains undiscovered encouraging zones for futurity mineral exploration plans. Remote sensing techniques recorded diverse types of alterations over the district. Hydroxyl-bearing and carbonate alterations, ferrous-bearing and ferric-bearing oxides were delineated to represent probable mineralisation zones intruding ophiolitic, metavolcanics, metasediments and monzogranite rocks. A highly fractured, tectonized, system demonstrating the west AHS was observed using remote sensing techniques. Such an intense fractured system was indicated using magnetic maps to prove that the mineralisation occurrences in AHS are structurally controlled. Moreover, magnetic data indicated a shallow nature of these structures to represent a very encouraging area for future mineral exploration. Also, magnetic maps indicated the efficiency of the surficial tracing of structures/lineaments and alteration/mineralised sources using geological surveys and remote sensing techniques over AHS and similar environments. According to the distribution/depths of magnetic sources, a ductile–brittle deformation zone (DBZ) was suggested that interpret why the Allaqi district is rich with auriferous-quartz veins and hosts altered, mineralised-rich, zones. Therefore, a GIS strategy depending on the intensity of the fracturing system and the existence of more than one type of alteration occurrence was used for ranking the prospective mineralisation areas. The high orders ranked zones matched very well with the historical/known mines within the district. Therefore, the highly potential-ranked zones could represent encouraging provinces for any future drilling programmes and mining exploitations. Furthermore, different rock units over the west AHS were differentiated.

Acknowledgements

The USGS, NASA and METI are thanked for the free Landsat 8 and ASTER-GDEM images. The authors extend their sincere appreciation to the anonymous reviewers for their valuable comments and helpful suggestions which greatly improved the early version of this manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data collected for this work are available.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/20909977.2022.2156068