Evaluation of gold mineralisation potential using AHP systems and weighted overlay analysis

The demand for sustainable development goals and the absence of systematic development and organised exploration for gold has prompted this study to integrate magnetic and radiometric datasets with lithology to evaluate the gold mineralisation potential in the Ilesha schist belt. This study considers 3168.72 km2 of the Ilesha schist belt in southwestern Nigeria, a frontier belt for gold deposits. The high-resolution airborne magnetic and radiometric datasets were processed using enhancement techniques, including the analytical signal, lineament density, and K/Th ratio. CET grid analysis, Euler deconvolution, and analytical signal depth estimation methods were used to aid the interpretation. The spatial integration and interpolation were performed using the Analytical Hierarchy Process (AHP) and weighted overlay analytical tools within the ArcGIS environment. The dominant structural controls for potential mineralisation are ENE–WSW and ESE–WNW trends. The depth of the magnetic sources revealed by the analytical signal ranged from 63.17 to 629.47 m, while depths ranging from 47.32 to 457.22 m were obtained from Euler deconvolution. The delineated highly magnetic edge sources, dense lineaments, radiometrically highlighted alteration zones, and lithological hosts for gold mineralisation were integrated to establish the gold mineralisation potential map. The AHP deductions reveal that 10.52% of the study site is within the high mineralisation potential class, a remarkable 60.39% falls within the moderate class, a significant portion (28.86%) falls within the poor class, and 0.23% is considered unfavourable. The result was optimised by validation using known mines, with 94% (i.e., 15 out of 16 mining sites) plotting within the high mineralisation potential class. This assessment provides invaluable insight for stakeholders and policymakers to embark on gold exploration and exploitation and promote sustainable mineral development.


Geologic settings and mineralisation
The study site is situated in southwestern Nigeria, bounded by Longitude 4° 30ʹ and 5° 00ʹ E and Latitude 7° 30ʹ and 8° 00ʹ N. Its entire area is 3168.72 km 2 .It is dominated by the Ilesha schist belt (Fig. 1).The Ilesha schist belt is well known for its lode gold and alluvial mineralisation 54 .The belt stretches nearly 200 km from north to south; it widens to its most significant point in the south at 60 km.The structural unit is dominated by the NNE-SSW trending fault, which is notable for the Ife fault.The Ife fault divides the lithology into distinct units 10 .The site is dominated by amphibolite, schist, gneiss, migmatite, and granitic rocks.There are two basic gold deposits within the belt: alluvial and auriferous.Within the amphibolite, there are notable alluvial gold occurrences at Igun and Itagunmodi.The auriferous quartz veins in the granite gneiss are where the lode gold occurrences at Iperindo are located 9 .Iperindo gold is the most notable deposit within the Ilesha schist belt due to its prominence 54 .Other alluvial occurrences (artisanally mined) are distributed sporadically around these notable deposits (Fig. 2).
The gold within the belt, especially that of Iperindo, occurs with pyrite, pyrrhotite, and a lesser amount of magnetite, galena, chalcopyrite, sphalerite, and ilmenite.Its host gneiss has undergone hydrothermal alteration, and it is adjacent to the veins that contain gold 9 .Based on structural correlations, the secondary faults are primarily responsible for controlling mineralisation.The gold emplacement at Iperindo was thought to be late Pan-African 9,58,59 .
The terrain comprises lowlands and hilly areas ranging in elevation from 233 to 757 m above sea level (Fig. 2).Data from the Shuttle Radar Topography Mission (STRM) elevation, one arc-second global (30 m resolution) obtained from the USGS website (www.usgs.gov) on November 24, 2023, was used to generate the elevation map.Amphibolite and gneiss are typically found along riverbeds and exist as low-lying hills and outcrops.The gold mines are situated at shallow elevations where occurrences are found.

Materials and methodology
The airborne magnetic and radiometric datasets were obtained from the Nigerian Geological Survey Agency, Abuja, Nigeria (https:// ngsa.gov.ng/).These datasets were acquired by Xcalibur Airborne Geophysical.Both the magnetic and radiometric data were collected concurrently with the following acquisition parameters: 150 m inter-profile spacing, 50 m flight height, and 1500 m tie line interval.Figure 3 shows the methodological flowchart adopted in the data analyses.

Airborne magnetic data
The airborne magnetic dataset was acquired in the form of a total magnetic intensity (TMI) grid (with the regional field of 33,000 nT removed) (Fig. 4).The geomagnetic field is expected to show variations in inclination and declination from the equator to the pole due to Nigeria's low magnetic latitude.Consequently, a reduction to the equator (RTE) modification is done to the TMI to put the maxima of magnetic strengths above their sources (Fig. 5) 60 .A magnetic inclination and declination of − 8.248 and − 0.620, respectively, were obtained from the IGRF calculator.The Geosoft program (Oasis Montaj) version 8.4 was used to process the data.Several magnetic enhancement techniques were performed to yield favourable results.www.nature.com/scientificreports/

Analytical signal and analytical signal depth analysis
The analytical signal (AS) is an enhancement technique and complex function suitable for locating subsurface magnetic structures and measuring magnetic potentials 24 .Using the evaluation of 61 , we obtain the following as the analytical signal's amplitude function: where MI is the intensity of the magnetic field.∂MI ∂x , ∂MI ∂y , ∂MI ∂z is the gradient of the magnetic field intensity (in the first-order) in the x, y and z directions, respectively.
An analytical signal can be employed to estimate the depth to the top of magnetic causative bodies, with minimal assumptions made about the nature of the anomalies 17 .Usually, the anomalies are identified as twodimensional magnetic source bodies 62 .The expression for the analytical signal depth (ASD) sources 63 is given as: where AS is the analytical signal expressed in Eq. (1), AS 1VG is the analytical signal of the first-order vertical gradient, while SI is the structural index geometry of the magnetic sources, where SI can be 1, 2, 3, or 4 represent- ing contacts, dike, pipe, and sphere, respectively 64 .Other depth estimation methods, such as radially averaged power spectrum (RAPS) and Euler deconvolution, were also used. (

Euler deconvolution
The apparent depth of a magnetic source can be estimated using Euler's homogeneity equation (Euler deconvolution, Eq. ( 3)) 65 .With the use of a structural index (SI) to represent the degree of homogeneity, the equation links the magnetic field and its derivative components to the locations of the causal sources 66 .The SI varies from 0.0 to 3.0 in general potential fields.In the magnetic field, 0, 1, 2, and 3 represent contacts, dyke or sill, pipe, and sphere, respectively.The equation below is the mathematical expression for Euler's homogeneity 66,67 .
where x 0 , y 0 , z 0 stands for the magnetic source's position, SI is the structural index, and B is the regional field of magnetism.The minimum curvature griding algorithm was used to interpolate the depth solutions after Euler's homogeneity equation was applied to the magnetic field.This was done to produce a map of the depth source distribution.The RAPS was employed to checkmate the depth estimates from both ASD and the Euler deconvolution.The theoretical details of the RAPS model can be found in 68 .

Centre for Exploration Target grid analysis and lineament density
The CET grid analysis (CETGA) is utilised for textural analysis and lineament delineation.It reveals all the structural controls connected to the magnetic data 69 .The details of the theoretical background of this analytical technique can be found in Refs. 69,70.The lineament features extracted from the data were used to produce a lineament density (LD) distribution in the ArcGIS environment (ver 10.5).The LD is suitable for the exploration of minerals 39,71 .Additionally, to ascertain the dominant distribution of structural trends, the extracted lineaments were plotted on a rose diagram.

Airborne radiometric data
The radiometric method of data acquisition was performed to quantify the naturally occurring radioactivity of gamma rays.These rays come from uranium, thorium, and potassium radioactive isotope-containing earth materials 16 .The high-resolution airborne radiometric dataset can be used to detect hydrothermal alterations, delineate radioactive geologic deposits, and map petrologic units 26,72 .Radioelemental ratios (e.g., K/Th, U/Th, U/K) are essential analytical tools suitable for identifying hydrothermal alterations 73 .
The radiometric dataset was acquired in the form of three grids: potassium (K%), uranium (eU in ppm), and thorium (eTh in ppm).This study employs K/Th to delineate the hydrothermal alteration zones.www.nature.com/scientificreports/

Gold mineralisation potential assessment
The assessment of gold mineralisation potential was carried out based on the factors that favour its delineation within the geologic settings under consideration.The gold mineralisation of the area is structurally controlled, occurs with metallic deposits, and its host is hydrothermally altered 9,54 .As a result, four parameters were adopted to determine the potential of gold mineralisation.These are lithology, analytical signal, lineament density, and K/Th ratio.The degree of limits to the specific requirements listed in Table 1 was taken into consideration when conducting the assessment of the gold mineralisation potential.By the standard deviation of the distribution of the individual parameters, the areas were categorised into highly favourable (HF), moderately favourable (MF), poorly favourable (PF), and not favourable (NF).
The raster layers of the aforementioned parameters were then transferred to the ArcGIS environment for further analysis.Units deemed to be targets for gold occurrences (i.e., using mine sites as control and expert opinions) of gold mineralisation potential were assigned to differentiate the features representing the raster data in order to perform reclassification.The standard deviation of targets was used as the basis for the classification.With the aid of the reclassification tool in ArcGIS 10.5, the features were transformed into reclassified raster layers (Fig. 13).

AHP
The analytical hierarchy process (AHP) is a well-known and often-used method for classifying variables arranged hierarchically.It is a multicriteria analysis method that enables users to choose the parameters' weights while addressing multicriteria problems.In this study, weight values were assigned, criteria were compared using expert judgements, and the AHP model was then utilised to process the weight values [74][75][76] .Reference 75 created the AHP model, which enables comparison of the relative weights of each criterion using a scale ranging from one to nine, which reports the relative ranking of each variable (Table 2).The weights of the pairwise comparison matrix priorities were then established.
The consistency ratio (CR) was computed to assess the consistency of the pairwise comparisons 77 .Equation (5) was used to calculate the CR, with values ranging from 0 to 1 following 75,76 .Desirable consistency is indicated by a CR score of less than 0.1.The consistency index (CI) is essential to assess gold mineralisation potential, and it is computed using Eq. ( 4).
where max is the pairwise comparison matrix's largest eigenvalue, and d is the dimension of the matrix.CR is given as: where RI denotes the mean random consistency index.

Results and discussion
Airborne magnetic data

AS and ASD
The AS map shows the distribution of the magnetic zones; as a result, the areas were classified into high, intermediate, and low magnetic zones, with a corresponding range of 0.2-0.6 nT/m, 0.04-0.2nT/m, and 0.01-0.04nT/m, respectively (Fig. 6).The zones of higher maxima are mostly trending NE-SW.These zones could be representations of structures, contacts, and intrusives that could be promising targets for mineralisation 23,27 .
The high and intermediate zones of magnetism can be associated with granitic rocks and very high-grade metamorphic rocks (such as migmatite, amphibolite, and granite gneiss) that contain ferromagnesian minerals 49,78 .Meanwhile, the low magnetic zones are rocks made up of more felsic minerals 16 .In addition, the structures of magnetic discontinuities that could potentially be targeted for mineral prospecting are depicted on the map.Furthermore, the method of 63,79 was used by applying Eq. ( 2) to the RTE data in order to determine the depth distribution of magnetic sources (Fig. 7).As a result of the discordant magneto-structural features observed in Fig. 5, an N value of 2 for the dyke-like feature was chosen.The analytical signal's depth estimates varied from 63.17 to 629.47 m.Upon using SI of 1 (i.e., dyke-like features), Euler's deconvolution depth estimates of magnetic sources range from 47.32 to 457.22 m (Fig. 8).This shows some level of correlation with ASD.Fig. 6.AS map of the study site.

CETGA and LD
To identify areas of structural complexity and deposit occurrence favourability, CETGA was used to assess the texture and locate laterally consistent fissures, such as lineaments along the edges and ridges of magnetic bodies.The CETGA was employed on the RTE grid.A step-by-step process was used to estimate magnetic variation using textural analysis; detect any laterally consistent and continuous structures of discontinuity using phase symmetry; amplitude thresholding to suppress background signals and noise to improve lineaments; generate lineaments using skeletonisation (Fig. 9); convert features to smooth the generated lineaments; and, lastly, create a lineament density map that distinguishes regions based on their lineament density (Fig. 10).The lineament density map has been used in several works as a primary indicator for mineralisation 20,27,28,39 .Consequently, a rose diagram was plotted, and it revealed the dominant structural trend to be ENE-WSW followed by ESE-WNW (Fig. 11).The structural signature shows some level of similarity with the analytical signal and the first vertical gradient.Within the study area, 50 obtained NE-SW and W-E as the dominant trends.The NE-SW fault has been thought to be a result of the Pan-African orogeny 10 .The NE-SW and NNE-SSW trends were reported to be favourable structural controls for gold mineralisation 9 .

Airborne radiometric data K/Th
The K/Th ratio detected the regions of potassium enrichment, which are unique zones of hydrothermal alterations 14,49 .High K/Th ratios are reliable markers of hydrothermal alteration 48 .The K/Th ratio map (Fig. 12) shows that the alteration zones are characterised by 0.09-0.23%/ppm.However, this shows some level of similarity with the range (0.013-0.362%/ppm) obtained by Ref. 48.The large K/Th ratio zone at the northern part of the site appears to be a sizeable felsic rock rich in potassium or a metamorphosed felsic rock 16,49 , while the southern part of the study area shows some high K/Th ratio that appears to be structurally controlled.
The relatively high K/Th ratio can be considered for hydrothermally altered zones.This is because there could be depletion in the radioactive elements due to intense weathering 49 , and this is the case of the known mineralisation host rocks 58 .The results obtained for the K/Th are very similar to those obtained by Ref. 48.Not much variation was observed using higher-resolution data (as employed in this study) (Fig. 13).

Weight of AHP
The specialists who created the AHP scale and ranked the criteria according to their professional judgement have a big say in how the gold mineralisation potential is judged in the area.After carrying out a background survey and a review of pertinent literature, five criteria were selected for the study.A pairwise matrix was utilised to acquire the expert's judgement on these criteria, hence providing support for the AHP.The weights obtained from the AHP, as derived from professional judgement, were then used in the weighted overlay algorithm to give each criterion layer a priority.
According to the result of the AHP weight, the parameter "AS, LD, K/Th, and Geology" has an equal cumulative weight index (CWI) value of 0.250 (Table 3).This implies that parameters have equivalent levels of contribution to the consistency outcome.The AHP analysis yielded a cumulative index and random index of 0 and 0.9, respectively.Consequently, a consistency ratio of 0% was obtained, which is most desired 75,76 .

Gold mineralisation potential assessment
The AS, LD, K/Th, and lithology were evaluated as rasters.Overlaying four raster layers, each of which corresponded to a criterion mentioned in Table 3, allowed for the construction of a gold mineralisation potential map over the Ilesha schist belt (Fig. 14).Table 4 presents the split of the favourability classes of various areas, expressed in km 2 and percentages (%).The favourability classes are "Highly favourable," "Moderately favourable," "Poorly favourable," and "Not favourable".
The highly favourable class covers 329.11 km 2 , which is about 10.52% of the total area of the study site.The moderately favourable class encompasses a much larger area than the highly favourable class, covering a total of 1889.28 km 2 , constituting approximately 60.39% of the entire area.The poorly favourable class appears to cover 28.86% of the area, which is equivalent to 902.87 km 2 .Meanwhile, the "Not favourable" class covers a smaller area of 7.19 km 2 , accounting for 0.23% of the entire area.The AHP methodologies have been employed in recent studies on appropriateness evaluation to determine the weights of critical criteria 50,80,81 .These investigations have identified important parameters as essential components for modelling.Based on previous research and expert knowledge, the AHP weights assigned a higher priority to each criterion inside the weighted overlay.These weights were from a matrix that had been standardised.The results of the AHP weights showed that the lithology, analytical signal, lineament density, and K/Th ratio all have a significant effect on the gold mineralisation because they are directly linked to the nature of the occurrence of the gold mineralisation under study.The work of Ref. 50 within the southern portion of the study area utilised lineament density, resistivity data, and geological distribution.However, this study covers a regional part of the Ilesha schist belt, and the focus of this study is tilted toward the style of the gold mineralisation within the study site.Additionally, higher-resolution airborne magnetic and radiometric datasets were employed to unveil the gold mineralisation potential.The metallic association, structural control, and hydrothermal alterations of the gold mineralisation style have shown the relevance of the aforementioned enhancement techniques.

Optimisation by validation
Validation was made by plotting the coordinates of existing mining sites for correlation purposes to ascertain the effectiveness of the highlighted target zones for gold mineralisation potential.Some mining sites were observed.Additionally, the mining sites observed by Ref. 50 are remarkable sites within the study area.However, they are all artisanal workings.They are also favoured by the topography of the locations (Fig. 2), where they occur as alluvial deposits.Sixteen of the gold mines were plotted.94% (i.e., 15 out of 16) of the mines fall within the highly favourable class, while 6% fall within the moderately favourable class.None of the known mine sites fall into the poorly favourable or not favourable class.In summary, all the mine sites fall within the highly and moderately favourable class (i.e., 100% agreement).This shows a high level of accuracy in the prediction.In comparison to the deductions of Ref. 50 , 79% are in agreement, plotting within the high to moderate potential zones.This high percentage of agreement may be a result of contributions from the analytical signal and the radiometric ratios, which have not been integrated in previous works.This indicates that there are potential prospects within the study area that are yet to be unveiled.The northwestern part and other regions delineated to be favourable for gold mineralisation across the study area are relevant targets (Fig. 14).

Conclusion
The demand for sustainable development goals and the absence of systematic development and organised exploration for gold has prompted this study to integrate magnetic and radiometric datasets with lithology to evaluate the gold mineralisation potential in the Ilesha schist belt.As a result, several signal enhancement techniques were employed to aid data interpretation, including analytic signal, lineament density, and K/Th.With the use of expert opinions, the Analytical Hierarchy Process (AHP) model was used to compare criteria and assign weight values in order to determine the potential of gold mineralisation.
The AS map categorised the area's magnetic potential into low, intermediate, and high.It reveals structures, contacts, and intrusives that could be promising targets for mineralisation.The CETGA delineates the lineament distribution of the area, revealing a dominant trend of ENE-WSW, followed by the ESE-WNW.The K/Th was used to reveal the hydrothermal alteration zones.The analytical signal depth estimate of magnetic sources reveals 63.17-629.47 m of shallow to deeper sources, while the Euler deconvolution delineated a depth range of 47.32-457.22 m.The delineated hydrothermal alteration zones, high lineament density, and highly magnetic zones are significantly associated with active gold mines.The gold mineralisation potential map reveals 10.52% of the area to be highly favourable, 60.39% of the area to be moderately favourable, while the poorly favourable class covers the larger part of the study site with an aerial coverage of 28.86%.The unfavourable class covers the least portion (0.23%).The result was optimised by using known mines as a validation tool.94% (i.e., 15 out of 16 mining sites) plotted within the high mineralisation potential class.The higher level of accuracy may be attributed to the higher resolution of datasets used and/or contributions from the analytical signal and the K/Th ratio.
Policymakers and stakeholders should prioritise the highly favourable class for further exploration and development.Putting these strategies into effect will aid in the optimisation of gold production in the study area and, of course, promote sustainable mineral development in Nigeria.

Fig. 1 .
Fig. 1.(a) Geology of Nigeria showing the major schist belts and the study area location (after 10 ).(b) Geologic map of the study site (after Ref. 57 ).

Fig. 2 .
Fig. 2. Elevation map of the study site.

Table 1 .
Criteria adopted for the utilised parameters in the Ilesha schist belt.

Table 2 .
Fundamental scale used in AHP pairwise comparisons.

Table 4 .
Distribution of gold mineralisation potential in the Ilesha schist belt, southwestern Nigeria.