Multivariate Statistical Analysis and Structural Sovereignty for Geochemical Assessment and Groundwater Prevalence in Bahariya Oasis, Western Desert, Egypt

Abstract

The Bahariya Oasis is an example of an extremely hyperarid environment and it is characterized by an extensive nonrenewable Nubian Sandstone Aquifer System (NSAS), which is deemed the crucial provenance for agrarian and national development ventures. The present work aimed to assess the groundwater occurrences in the NSAS, and to document the main factors that control the geochemistry of the groundwater in the Bahariya Oasis. Groundwater samples were collected from 52 locations in April 2019 and were analyzed for a total of 13 water-quality physicochemical parameters. A diverse geological and structural setup has greatly impacted the groundwater flow pattern and has diverted it towards the NE by the great Bahariya anticline structure, the ENE-oriented Bahariya mid dextral strike-slip fault, and NE-striking normal faults, while NW-oriented normal faults cause the groundwater to diverge perpendicular to the groundwater flow lines. The groundwater is highly contaminated by trace metals (Fe²⁺ and Mn²⁺), which exceed the permissible limit for different purposes. Conventional graphical plots and geochemical modeling integrated with multivariate factor analysis (FA) revealed that the chemical composition of the groundwater is strongly affected by its interaction with the lithologies of the NSAS. The dissolution of aquifer host rocks (carbonates and iron oxides) and chloride salts through the infiltration of groundwater, and the incorporation of cations by the ionic exchange of Na⁺ by Ca²⁺ in clay minerals, emerged as worthy mechanisms for the groundwater development. Furthermore, the region’s rapidly increasing population, agricultural expansion, and the associated anthropogenic practices have generated a need for groundwater-quality assurance as a prime source of the water supply. Consequently, reducing the effects of the NSAS’s unsustainable extraction requires long-term monitoring and the ongoing evaluation of the groundwater.

1. Introduction

In the Egyptian deserts, such as the areas of the Western Desert, there has recently been a lot of attention paid to agricultural developments and sustainable growth, and their need to be consistent with the most important goals of both the National Water Resources Plan 2017 [1] and the Government of Egypt’s Strategy for Sustainable Agricultural Development 2030 [2]. The Sustainable Agricultural Development Plan for 2030, which was launched by the Ministry of Agriculture and Land Reclamation in the late 2000s, takes into account all facets of agricultural production. In terms of water conservation, the aim was to increase water-use production from 50% to 80% by 2013, which would “save volumes of water measured at 5.3 and 12.4 billion m³ by 2017 and 2030, respectively and also allow for a 5 million feddan expansion” [2]. Such fanciful numbers fail to understand the distinction between the water-use potency on the plot and in the NSAS system, which is a problem that the MWRI described and discussed as early as the mid-1990s.

Any debate as to the ways to improve production and how to fulfill it are inevitably linked to the question of the overall water availability. Since the only provider source of water in these dry areas is the groundwater abstracted from the Nubian Sandstone Aquifer System (NSAS), which is designated as a nonrenewable groundwater resource [3], the groundwater system requires appropriate assessment and management procedures for permissible usage and long-term production. The overuse of groundwater from the NSAS, which lacked a comprehensive management plan, resulted in significant water deterioration across the area [4]. Groundwater hydrochemical characteristics are used to track groundwater flow processes, and to identify the mechanisms that govern groundwater chemistry and the geochemical processes that impact the groundwater quality. Groundwater mineralization is influenced by natural elements, including the aquifer lithology, the overriding geological structure, geochemical operations within the aquifer, and the quantity of the groundwater flow from the recharge to discharge areas across its passage, as well as anthropogenic activities, groundwater overpumping, unregulated well drilling, and unrestricted agricultural and industrial activities [5,6,7]. Understanding aquifer geochemistry and identifying the causes of groundwater depletion are vital for regulating and improving groundwater resource management [8].

The Bahariya Oasis in the Western Desert is undergoing a continuous lowering of the groundwater level as a result of unplanned water control, which causes severe soil salinization, resulting in a reduction in the quality and productivity of related resources [9,10,11]. Such environmental concerns have arisen as a result of the rapid expansion of agricultural practices, which has resulted in a significant increase in the use of groundwater. As a result, it is critical to continuously consider the aquifer characterization in the Bahariya Oasis in order to define the area’s areal distribution of aquifer layers and to assess the groundwater potentialities.

Groundwater chemistry studies certainly benefit from multivariate-statistical-analysis approaches [12,13]. These approaches aid in the classification of groundwater and in the identification of the main factors that influence the groundwater chemistry [14]. To interpret the hydrochemical parameters, these approaches might be combined with graphical representations. They are often regarded as the optimum and, in many cases, the only viable option for assessing massive amounts of hydrochemical data. Geochemical modeling is often employed for the study of the water–rock interaction. Its goal is to identify the chemical processes that take place along a flow channel. The combination of hydrochemical-interpretation visualizations, multivariate statistical analysis of the groundwater chemistry, and inverse geochemical models is an effective approach for assisting the geochemical process and identifying the main factors that control the groundwater quality in their respective study areas [8,15,16]. The main objectives of this study were to emphasize the roles of geologic structures in monitoring groundwater flow regimes, to identify the hydrochemical characteristics and the processes that control the groundwater quality, and to identify the mineralization of the groundwater in the Bahariya Oasis.

2. Study Area

The Bahariya Oasis is a natural oval depression in the midst of Egypt’s Western Desert, spanning approximately 2200 km² and positioned NE–SW (Figure 1a). The area is dominated by hyperarid climatic conditions [17]. The oasis’s population of 34,000 people is mostly focused on agriculture, with farming settlements scattered over the landscape. Scientists are interested in it because of its economic value, industrial activity (mostly iron ore resources), tourism, and vast reclaimable fields.

A highly precise Phased Array type L-band Synthetic Aperture Radar (PALSAR) dataset (12.5 m resolution) was employed to analyze the geomorphic properties of the Bahariya, and it was processed using ArcGIS 10.6. The topography of Bahariya is marked as a flat rough, rocky surface owing to the weathering of hard rocks, and it ranges in height from 71 to 370 m above mean sea level (AMSL) (Figure 1b). The Bahariya Oasis is divided into three main parts: northern, central, and southern. The northern and southern parts are flat, while the central sector is rocky and higher in altitude. The plateau surface, bordering escarpments, oasis floor, wadis, residual hills (questas), alluvial fans and terraces, volcanic cones and sheets, sabkhas, sand sheets, and sand dunes are the most notable geomorphic features [18].

According to Egypt’s Strategy for Sustainable Agricultural Development 2030, the Bahariya Oasis is subject to about 80 km² of the 1.5-million-acre project, and the nonrenewable groundwater withdrawn from the Nubian Sandstone Aquifer is the only source of water for agriculture and residential use. Thus, decisionmakers must work to sustain, enhance, and plan groundwater-quality investigations, as well as identify effective ways to reduce both natural and human pollutants.

2.1. Geological Setting

With regard to the subsurface sequence, from bottom to top (Table 1 and Figure 2), a detailed stratigraphic conception of the Bahariya Oasis is as follows:

• The Precambrian basement is made up mostly of igneous and metamorphic rocks (1391–1718 m) [19];
• Intercalated siltstone, sandstone, and clay constitute 458 m-thick Cambrian rocks;
• Cretaceous rocks: 660 m from the bottom to the top, and there are four formations:

  (a)

  The Bahariya Formation (Lower Cenomanian). This formation is made up primarily of flat-lying fluviatile sandstone and it serves as the oasis’s floor (total: 705 m thick, with a maximum exposed thickness of 173.5 m). The Bahariya Formation is marked by abundant intercalations of ferruginous layers [20,21].

  The thickness of the Bahariya Formation ranges from 90 m in the central section of the depression, to more than 170 m in the northern parts (Gabal El Dist) (Figure 3d). It displays a portion of the Bahariya Formation (Cenomanian, Upper Cretaceous) and is overlain by the Naqb Formation (Middle Eocene). In the Sandstone Hills region and at Gabal Miteilaa Radwan, the lower sandstone member is well exposed. In the northeastern scarps, the Bahariya Formation is surmounted by Middle Miocene basaltic rocks at Gabal El-Hefhuf.

  The Black Desert is located 50 km south of El-Bawiti, is composed of conical-shaped mountainous hills, and is fundamentally related to the Bahariya Formation (Cenomanian ferruginous sandstone and chalky clays). The majority of these hills are covered with ironstones and dolerite crystals (Figure 3e);

  (b)

  The El-Heiz Formation (Upper Cenomanian). This formation, which is unconformably topped by the El-Hefhuf Formation, is composed primarily of fossiliferous limestone, shale, calcareous sandstone, and marl (Figure 3f) [19,22,23]. The El-Heiz Formation reaches a maximum thickness of 40 m in the southern part of the depression at Naqb El Sellem, and then drops to around 12 m in the northern scarp;

  (c)

  The El-Hefhuf Formation (Campanian–Turonian). This formation is a 130 m-thick succession of light-brown siliceous dolomite, dolomitic limestone, and sandstone, and it unconformably superimposes the El-Heiz Formation (Figure 3g) near Gabal El-Hefhuf [24]. At the Bahariya north scarp, the El-Hefhuf Formation is absent, and the Bahariya Formation is unconformably overlain by Lower Eocene limestones [25]);

  (d)

  The Khoman Formation (Maastrichtian). This formation is from 30 to 45 meters thick and consists of huge white chalky limestone that covers most of the Bahariya southern scarps [19,26,27]);

• Paleocene rocks are clearly identified by: (a) The Tarawan Chalk (20–30 m thick), which is composed essentially of chalky limestone and limestone and which is naked to the south of the depression; and (b) The Esna Shale, which is featured in the plateau’s southern escarpment (Figure 3h), separating the Bahariya and El-Farafra depressions [19,23]);
• Eocene rocks. They constitute the Farafra and Nqab formations and represent the top of the plateau that surrounds the Bahariya depression:

  (a)

  The Farafra Formation (Lower Eocene), which is featured in the eastern and southern escarpments of the depression. Only the upper member of the Farafra Formation is visible;

  (b)

  The Naqb Formation, which mainly comprises pink to violet Nummulitic dolostone and limestone accumulations (Figure 3i). Iron ore deposits have emerged in the northeastern Bahariya Depression (the El-Gedida, Gabal Ghorabi, Nasser, and El-Harra districts), where iron-bearing solutions have substituted the Naqb’s carbonate rock [28]);

  (c)

  The Qazzun Formation, which is constituted of chalky Nummulitic limestones with firm dolomitic-limestone ball-like concretions (7–10 km-wide band);

  (d)

  The El-Hamra Formation (Middle-Upper Eocene), which is on the eastern plateau of the depression (63 m of fossiliferous limestone deposits) [29];

• Tertiary rocks comprise: (a) Volcanic rocks. Numerous isolated basalt outcrops in the northern and central parts of the depression (Measera, Mandisha, El-Hefhuf, and Basalt Hills), sheets of Oligocene-aged volcanic rocks, and mainly extrusive basalt and dolerite cap the sedimentary deposits (Figure 3j,k). A WNW-oriented dolerite dike and a thin laccolith may be seen at Gabal El-Hefhuf; (b) Ferruginous grits and layers of sandstone and clayey sandstone constitute the Radwan Formation [19,25,29]); (c) The Qatrani Formation (Oligocene) is found in the northeastern plateau and represents the weathering products of various Cretaceous and Eocene rock groups;
• Quaternary rocks include: (a) aeolian sands, forming dispersed seif dunes; (b) sabkhas and salt deposits, established owing to the infiltration of water from naturally flowing springs and wells on the depression surface; (c) playa deposits, which are made up mostly of fine sand, silt, clay, gypsum, and halite, and are found in freshwater depressions.

Table 1. Composite stratigraphic section of El-Bahariya Oasis (compiled from [23,30]).

Age	Rock Unit	Thickness (m)	Lithology
Quaternary			Gravel terraces, talus, sabkha and playa deposits, wind-blown sand, and alluvium
Early Miocene (16–20 Ma)	Basalt flows and sills	20	Olivine basalt
Oligocene	Qatrani Formation		Weathering products of the different rock units of Cretaceous and Eocene ages and occurs in the northeastern plateau
Oligocene	Radwan Formation	40	Ferruginous grits and beds of pale-brown to yellow sandstone and clayey sandstone
Unconformity
Middle and Late Eocene	El-Hamra Formation	63	Fossiliferous limestone beds with a few clastic intercalations with inward dips of 10–40°
Middle Eocene	Qazzun Formation	32	Chalky Nummulitic limestones with ball-like concretions of hard dolomitic limestone
Unconformity
Middle Eocene	Naqb Formation(Ghorabi iron ore member)	68	Pink to violet Nummulitic dolostone and limestone forming isolated hills
Lower Eocene	Farafra Formation	34	Limestone
Upper Paleocene–Lower Eocene	Esna Shale	20	Greenish-gray to gray shale
Lower Paleocene	Tarawan Chalk	20–30	Chalky limestone and limestone
Unconformity
Early Maastrichtian	Khoman Chalk	30–45	Massive white chalky limestones
Campanian	El-Hefhuf Formation	120	Dolostone and sandstone with sandy clay intercalations
Unconformity
Late Cenomanian	El-Heiz Formation	30	Clastics with carbonate interbeds and a dolostone member at the top
Cenomanian	El-Bahariya Formation	170+	Variegated, cross-bedded sandstone
Cambrian	Cambrian rocks	458	Intercalated siltstone, sandstone, and clay
Pre-Cambrian			Metamorphosed igneous basement

Table 1. Composite stratigraphic section of El-Bahariya Oasis (compiled from [23,30]).

Age	Rock Unit	Thickness (m)	Lithology
Quaternary			Gravel terraces, talus, sabkha and playa deposits, wind-blown sand, and alluvium
Early Miocene (16–20 Ma)	Basalt flows and sills	20	Olivine basalt
Oligocene	Qatrani Formation		Weathering products of the different rock units of Cretaceous and Eocene ages and occurs in the northeastern plateau
Oligocene	Radwan Formation	40	Ferruginous grits and beds of pale-brown to yellow sandstone and clayey sandstone
Unconformity
Middle and Late Eocene	El-Hamra Formation	63	Fossiliferous limestone beds with a few clastic intercalations with inward dips of 10–40°
Middle Eocene	Qazzun Formation	32	Chalky Nummulitic limestones with ball-like concretions of hard dolomitic limestone
Unconformity
Middle Eocene	Naqb Formation(Ghorabi iron ore member)	68	Pink to violet Nummulitic dolostone and limestone forming isolated hills
Lower Eocene	Farafra Formation	34	Limestone
Upper Paleocene–Lower Eocene	Esna Shale	20	Greenish-gray to gray shale
Lower Paleocene	Tarawan Chalk	20–30	Chalky limestone and limestone
Unconformity
Early Maastrichtian	Khoman Chalk	30–45	Massive white chalky limestones
Campanian	El-Hefhuf Formation	120	Dolostone and sandstone with sandy clay intercalations
Unconformity
Late Cenomanian	El-Heiz Formation	30	Clastics with carbonate interbeds and a dolostone member at the top
Cenomanian	El-Bahariya Formation	170+	Variegated, cross-bedded sandstone
Cambrian	Cambrian rocks	458	Intercalated siltstone, sandstone, and clay
Pre-Cambrian			Metamorphosed igneous basement

Finally, the iron ore localities in the Bahariya are El-Gedida (Figure 3l), Nasser, El-Harra, and Ghorabi. Ref. [31] identified three genetic forms of El-Gedida iron ore: (1) massive hydrothermal metasomatic; (2) porous or massive, resulting from mobilized iron and manganese redeposits in freshwater lakes, probably from biogenic activity; and (3) a result of the weathering of glauconitic clays and sands, typically with an oolitic or pisolitic form.

2.2. Structural Setting

The Bahariya depression is deformed by an NE-trending right-lateral wrench-fault structure (Figure 2), which is associated with several doubly plunging folds and extensional faults [19,32] of the large Syrian Arc belt [19,32,33,34,35,36,37,38,39]. The faults trending E-W, ENE, and WNW are the masters and are characterized by a dextral sense of movement, while those of the NS, NNW, and NNE directions are subordinate, with a sinistral sense of dislocation [32].

The Bahariya Oasis depression is essentially a huge anticline that runs NE–SW, from Gabal Ghorabi in the north to El-Heiz in the south, passing through the depression’s middle hills (Figure 2). Folding occurred after the deposition of the thick Cenomanian Bahariya clastics [30].

The Bahariya great anticline consists of a number of major anticlines (e.g., El-Gedida, Naqb El-Harra, Ghorabi, El-Tebaniya, Heiz, and Ris), major synclines (e.g., El-Hefhuf, Miteilaa Radwan, and Tobog), and one monocline (Sandstone Hills), as well as dozens of small-scale folds. The dips of bedding are as high as 55° in the north (Figure 2).

One of the prominent wrench faults, cutting through the Bahariya depression and associated with severe folding on both sides of its trace, is a major dextral strike-slip fault that was termed the Bahariya mid fault by [32]. The Bahariya great anticline is markedly displaced by the Bahariya mid fault.

In the northern segment of the Bahariya Oasis occurs the El-Gedida fold, and the Gabal Topog and Gabal Ghorabi structural segments. The El-Gedida anticline (Figure 2) is a major SW-plunging anticline that occurs in the central part of the mine area. The iron ore beds show a diagnostic fold in profile view, with axes parallel to the major fold. The El-Gedida anticline, interpreted by [36] as a wrench-related folding, indicates that the wrenching ceased during the Late Eocene time.

The Gabal Tobog fault is an NE-striking dextral strike-slip fault (Figure 4). Two phases of movement were recognized on the fault surface by [19]. The first displacement occurred prior to the deposition of the Naqab Formation and resulted in the complete erosion of the El-Hefhuf Formation, which was located on the southern side of the fault at a higher structural elevation (hanging wall). The second movement took place after the deposition of the Naqab Formation, and this is because the fault split the Naqab Formation northward of Naqb El-Harra on the eastern plateau (Figure 4a).

In Naqab El-Harra, and to the south of the Gabal Tobog fault, there is an NE-striking reverse fault that is associated with a hanging-wall anticline in the Bahariya Formation (Figure 4a). In the eastern plateau, part of this anticline is exposed along the new road cut into Naqab El-Harra (Figure 4b), and the folded beds are displaced by a WNW-striking normal fault (Figure 4c).

To the south of Gabal Topog, there are three hills along one line, and mainly of the Bahariya Formation overlain by the El Heiz Formation. These hills occur as three major doubly plunging NE-trending synclines (Figure 4a,d), with the largest syncline displaced by an NNE-striking normal fault (Figure 4d), whereas the smallest syncline is displaced by an E–W striking normal fault. To the north of these synclines, there is a major doubly plunging anticline along the same line as the synclines. These major doubly plunging asymmetrical synclines and anticline, together with the El-Gedida and Naqab El-Harra anticlines, are characterized by right-stepped en echelon sequencing (Figure 4a), perhaps because of a Late Cretaceous right-lateral shear couple.

A small manmade exposure in the Upper Bahariya Formation to the west of Gabal Topog shows an overturned fold (first documented in this study). Variegated clay, shale, and sandstone constitute the majority of these rocks. The fold axis plunges towards the NE, and its overturned limb is displaced by an NE-striking dextral strike-slip fault (Figure 4e,f).

The Gabal Ghorabi structural segment is affected by the ENE-striking Ghorabi fault (Figure 5a). This dips 65° toward the south, cuts the central part of Gabal Ghorbi, and displaces its southern part downward. Along the Ghorabi fault, [19] identified two phases of movement. The first is a right-lateral strike-slip displacement that occurred post-Cenomanian and before the Middle Eocene, leading to a southward steep dip of the Bahariya Formation on the southern section of the fault, as well as the creation of a reverse fault and an NE-trending anticline and NE-striking normal faults in the Bahariya Formation (Figure 5b,c). These normal faults dip moderately (45–50°) toward the NW and SE (Figure 5b–d). The SE-dipping faults displace the NW-dipping ones. During the second movement (post-Middle Eocene), the Middle Eocene Naqb Formation (e.g., iron-rich beds) was southerly displaced along the Ghorabi right-lateral strike-slip. This movement led to the development of several doubly plunging anticlines and synclines, as well as monoclines trending NE–SW, and they have a right-stepped en echelon arrangement (Figure 5e). The Gabal Ghorabi structural segment is made up of multiple normal-slip faults that strike WNW to NW and are arranged in an echelon. Displacement along these normal faults during the second movement (post-Middle Eocene) generated various horst-and-graben formations, as well as many WNW-plunging anticlines and synclines (Figure 5f–h).

The Gabal El-Hefhuf structural segment occurs in the central part of the Bahariya depression, along the ENE-oriented Bahariya mid fault (N50°E/75°NW) and the related folds (Figure 6a). Gabal El-Hefhuf represents the downthrown side of the Bahariya mid fault. Gabal El-Hefhuf occurs as a major doubly plunging hanging-wall syncline dissected by several NW-striking normal faults. The most conspicuous NW-striking fault is that between Gabal El-Hefhuf and Gabal El-Basalt (Figure 6b). Another NW-striking normal fault displaces Oligocene basalt in Gabal El Engleez (Figure 6c) and forms a thick bed of breccia parallel to the fault plane.

The Gabal El-Hefhuf hanging-wall syncline (Figure 6d), and the small right-stepped en echelon anticlines and synclines to the west of Gabal El-Hefhuf, are directed east–northeast, subparallel to the Bahariya mid fault, but they shift directions in places and create an acute angle with it. The Upper Cretaceous rocks of Gabal El-Hefhuf are pierced by an NW-oriented right-lateral strike-slip fault that displaces the axis of the hanging-wall syncline by up to 1 km (Figure 2).

The Gabal Miteilaa Radwan structural segment (Figure 2) occurs along the Bahariya mid fault and consists of several ENE-oriented right-stepped en echelon plunging anticlines and synclines (Figure 6e) that form acute angles with the main plane of the Bahariya mid fault.

The Sandstone Hills form an ENE-oriented monocline and have steep flanks, with a dip-angle range between 35 and 55° (Figure 2). According to [19], the existence of a deep-seated ENE-oriented fault beneath the Sandstone Hills contributed to the creation of the monocline.

To the west of the Sandstone Hills, on the southern part of the Bahariya depression, there are two large NE-oriented doubly plunging anticlines (Ris and Heiz) in the Upper Cretaceous rocks. These folds generate an acute angle with the Sandstone Hills monocline’s continuation and the Bahariya mid fault.

2.3. Hydrogeological Setting

The Nubian sandstone, which is largely composed of continental clastic layers, and predominantly sandstone with shale and clays intermingled, is the biggest water-bearing source in the investigated region. The groundwater system is hydraulically attached to adjacent and inherent aquifers via feasible conduits that allow upward outflow. Therefore, the NSAS is defined as a multilayered “leaky” artesian aquifer system that operates as a coherent hydrogeologic framework [40]. According to [23], the NSAS hydraulic conductivity averaged 5.67 m/day, with values ranging from 0.46 m/day in the El-Bawiti area, to 10.88 m/day in the El-Heiz region (southern part). Furthermore, the transmissivity reached its lowest point (236 m²/day) in the north (El-Bawiti), and its highest point (3045 m²/day) in the south (El-Heiz). This is likely owing to the restricted development of shale deposition in the relevant aquifer in that direction, demonstrating the aquifer’s high potentiality and efficiency in the south.

Although the isotopic findings provide strong arguments that the fossil water of the basin was infiltrated during humid periods in the Late Pleistocene and Holocene times, it is likely that the leakage of water from the Nile into the NSAS can occur, but on very local scales [41]. The groundwater of the Nubian Sandstone aquifer was extracted at a rate of 56.6 million m³ per year, of which 50.7 million m³ per year was used to irrigate an area of 12,500 feddan, while the remaining 5.9 mcm/year was allocated for domestic purposes [42]. The exploitation of the groundwater is increasing each year. In the past 40 years, over 40 billion m³ of water has been extracted from the system in Libya and Egypt. This has produced a maximum drawdown of about 60 m and has led to the disappearance of almost all of the naturally flowing wells and springs [43].

3. Methodology

3.1. Sampling and Analysis

A random sampling scheme was initiated in April 2019 to explore the spatial distribution of the groundwater-quality parameters, and measurements of the water level from different wells were also achieved under static conditions to examine the head pattern and flowing directions. Concurrently, a total of 40 groundwater samples were taken from the confined deep aquifer, with documented depths ranging from 760 to 1170 m. Meanwhile, a total of 12 groundwater samples were gathered from the confined shallow aquifer, and their well depths ranged from 250 to 495 m (Figure 1b). The wells were chosen on the basis of the aquifers’ availability, geographical distribution, accessibility, and representation. These wells were operated for 10 min before gathering fresh samples of the relevant aquifer to eliminate any sluggish or polluted water. Samples were filtered through 0.45 μm-pore-diameter filter paper and collected in cleaned polythene bottles of a 1 L capacity. Then, samples were transported to the laboratory and maintained at 4 °C for subsequent chemical analysis.

In the preliminary assessments, the key physicochemical parameters comprising temperature (T °C), electric conductivity (EC), and total dissolved solids (TDS) were measured in situ using portable meters. Using the Direct Reading/2000 Spectrophotometer, the total quantity of iron as a trace element was also analyzed in situ. Major cations and anions, as well as Mn²⁺ as a trace element and NO₃⁻, were detected shortly after collection following Hach’s normal analytical techniques [44]. A handheld GARMIN GPS was used to acquire the geographic coordinates and elevation of the sampling locations. The data were checked for accuracy, and the computed ion-balance discrepancy was deemed to be within the acceptable range of 5%. Table 1 shows the general hydrochemical statistics for all the water samples collected from the different aquifers in the Bahariya Oasis, which can be used to understand their status and to determine their trend evolution.

3.2. Geostatistical and Geochemical Modeling

Semivariography and ordinary kriging (OK) were used to distinguish the spatial distributions of the hydrochemical variable concentrations in the groundwater system [45,46,47]. OK has emerged as one of the most popular linear unbiased estimators for constructing an appropriate spatial model in random-information areas while allowing for indiscriminate model uncertainty [47]. The spatial-variation maps were generated with ArcGIS 10.6’s Geostatistical mapping tool.

Geochemical modeling has been commonly used to assess the geochemical mechanism that causes groundwater to evolve [48,49,50,51]. The hydrochemical software PHREEQC is used to perform the mass-balance simulation, which shows chemical reactions and modifications in the chemical nature, such as the dissolution/precipitation of minerals and gases in the groundwater flow route [52]. Saturation-index (SI) results illustrate the propensity of minerals to dissolve or precipitate in the groundwater system. Geochemical modeling was performed using the following equation [53]:

SI = log (IAP/Kt);

where IAP is the ion-activity outcome in a solution of the disintegrated chemical species, and Kt is the mineral’s equilibrium-solubility component. The SI for any mineral phase with a negative value implies that the water is undersaturated with regard to that mineral and that it will dissolve until equilibrium is achieved, while a positive value denotes oversaturation, and that the mineral will precipitate.

3.3. Multivariate Statistical Analysis

Statistical investigation offers more attractive options in environmental science, though the results may deviate more from real situations. Using the XLSTAT statistical software, the 13 hydrochemical parameters (TDS, EC, pH, Ca²⁺, Mg²⁺, Na⁺, K⁺, HCO₃⁻, SO₄²⁻, Cl⁻, NO₃⁻, Fe²⁺, and Mn²⁺) in both shallow and deep aquifers were correlated, and factor analysis (FA) was also performed. The Pearson’s coefficients (r) between the different ion pairs and their relation to TDS could be used as a first pointer for the types of salts that dominate the groundwater composition and the major salinization processes [54,55,56,57]. FA’s feasibility has been shown in prior groundwater-quality investigations [15,46,58].

Factor analysis (FA) is a multivariate statistical approach that is used to investigate the fundamental connections between multiple noted variables in terms of a potentially lower proportion of unobserved variables, which are referred to as factors, with minimal native-information loss. A correlation matrix is produced for the hydrochemical dataset, which designates the connectivity between variables. The generated parameters will represent the entire dataset’s properties. The overall variation (percent) was addressed using Kaiser’s criterion (eigenvalues > 1), and significant principal components (PCs) that constituted a percentage of variance > 10% were extracted from the original parameters. Varimax-rotation-generating factors (FA) were applied to these PCs to limit the impact of variables of little relevance. The FA provided the criteria of the most important water-quality measures and highlighted the relationship between the variables and sample locations in their geographical distributions, which were grouped together on the basis of comparable environmental conditions [15,46,51,59]. The eigenvalues and factor loadings of the correlation matrix were computed, and a scree plot was created. The variances and covariances of the variables were used to create the extraction factors. Prominent variables are defined as those with eigenvalues larger than one. Finally, the rotating mechanism enhances each variable’s loading on one of the extracted components while reducing the loadings on the other factors.

4. Results and Discussion

4.1. Groundwater Flow Pattern

The NSAS piezometric levels in the Bahariya Oasis were generated by using the available water-level observations over 2001, 2009, and 2019 (Figure 7). This level reached its peak during 2001 (149 m) in the El-Heiz region, while it dropped to 90 m and was mostly found in the northern part of the Bahariya Oasis (Figure 7a), suggesting that the pressure head of the NSAS in the southern part is higher than in the northern part. However, this level declined during 2009, and it ranged between 74 m and 140 m in the northern part and southern parts, respectively (Figure 7b). According to Egypt’s 2030 Strategy for Sustainable Agricultural Development, the Bahariya Oasis is home to about 20,000 acres of the 1.5-million-acre project, with the nonrenewable groundwater extracted from the NSAS serving as the region’s only source of water for irrigation and domestic use. This produced a noticeable decline in the piezometric level during 2019, and minimum values of 59 m in the northern part and 137 m in the southern part were recorded (Figure 7c).

The overall flow of the groundwater in the Bahariya aquifer is directed SW to NE, following the slope directions. The groundwater in the area flows in the same direction (SW–NE) as the groundwater in the Great NSAS in Egypt’s Western Desert [11,23,40,60,61,62]. Furthermore, due to significant groundwater pumping, the pressure head rises to the east, west, and south of the oasis, while the flow shifts to the east and west, approaching El-Bawiti town and its outskirts in the northern area [11,23].

The existing structural elements were greatly impacted by the groundwater conditions, as Bahariya is situated on the boundary between Egypt’s tectonically stable and unstable shelves [30]. Consequently, by comparing piezometric maps to the geological maps of the area (Figure 2), it was found that the anticlinal axis, with its associated strike-slip, reverse, and normal faults, as well as joint system, permits the movement of the groundwater to roughly follow the orientation of the anticline axis, plunging NNE.

Ultimately, it is reported that the piezometric head rises along structurally high areas, while the piezometric head decreases across structurally low areas. The wells drilled at El-Harra in 2001, 2009, and 2019 had maximum heads of 126, 115, and 111 m, respectively, whereas lower values of the head were matched with the wells that were drilled in the adjoining structurally low areas (El-Bawiti and its surroundings).

4.2. Hydrochemical Characteristics

Table 2. Descriptive statistics of the hydrochemical parameters of the shallow and deep aquifers. Units of the ion concentrations are in mg/L; Min: minimum; Max: maximum; St. dev.: standard deviation.

Variable	Shallow-Aquifer Samples n = 12 Samples				Deep-Aquifer Samples n = 40 Samples				Acceptable Limits
	Min.	Max.	Mean	St. dev.	Min.	Max.	Mean	St. dev.	(WHO 2011)
TDS (mg/L)	147.9	1740.0	455.3	434.0	133.0	354.0	189.4	50.6	1500
EC (μ.S/cm)	296.0	3530.0	916.2	881.3	267.0	708.0	380.0	98.9	-
pH	5.7	6.8	6.2	0.3	5.6	7.0	6.2	0.3	6.5–8.5
Ca2+	7.7	128.7	34.6	33.9	6.1	22.2	10.6	3.5	200
Mg2+	9.8	196.1	39.1	51.6	6.7	26.2	12.2	4.0	150
Na+	20.1	334.7	71.6	84.8	11.6	57.2	31.8	9.7	200
K+	11.5	158.3	39.6	40.6	6.9	26.2	15.4	5.3	30
HCO3−	28.0	56.0	37.0	8.3	6.0	98.0	46.9	21.7	300
SO42−	29.0	825.0	182.4	236.6	9.0	81.0	30.2	18.1	250
Cl−	49.0	728.0	146.8	188.1	27.0	126.0	61.3	21.0	250
NO3−	0.1	5.7	2.1	2.2	0.1	9.2	4.4	2.2	50
Fe2+	0.2	21.5	5.6	6.8	0.1	7.2	2.7	1.7	0.2–0.3
Mn2+	0.4	49.0	7.3	14.3	0.1	2.9	1.1	0.6	0.05–0.1

highlights the descriptive statistics of the hydrochemical parameters (lowest, maximum, average), together with the standard World Health Organization (WHO) values [63]. The following are the most commonly varying parameters for shallow aquifers: TDS (147.9–1740 mg/L); pH (5. 7–6.8); Ca²⁺ (7.7–128.7 mg/L); Mg²⁺ (9.8–196.1 mg/L); Na⁺ (20.1–334.7 mg/L); K⁺ (11.5–158.3 mg/L); HCO₃⁻ (28–56 mg/L); Cl⁻ (49–728 mg/L); and SO₄²⁻ (1229–825 mg/L), which exemplify the hydrochemical evolution’s intricacy. Furthermore, the ranges of Fe2+ (0.2–21.5 mg/L) and Mn2+ (0.4–49 mg/L) were unusually broad, implying that these elements emerged from water-bearing sediments. Meanwhile, these parameters for deep aquifers were not widely ranged: TDS (133–354 mg/L) and pH (5.6–7.0). Furthermore, the calcium and magnesium concentrations were nearly close and ranged between (6.1–22.2 mg/L) and (6.7–26.2 mg/L), respectively. The mean sulfate and chloride concentrations were 30.2 and 61.3 mg/L, respectively, and were located within the WHO standards.

Low-salinity regions for shallow aquifers were frequently identified on the depression’s ground along the fault plane, implying the probability of a hydraulic connectivity to the deep aquifers beneath (Figure 8a). The decreased sand content and widespread shale intercalations, with silt northward (El-Bawiti), consequently results in a reduction in transmissivity, which inhibits the groundwater flow rate and promotes salt dissolution. The TDS spatial distribution in shallow and deep aquifers, which indicates the northward rising trend in salinity towards the El-Bawiti village, supports this hypothesis (Figure 8a).

Furthermore, the naturally coherent presence of Fe²⁺ and Mn²⁺ trace metals was found to be contaminating most of the groundwater, which suggests that the development of these elements from water–rock interactions is the source of these minerals, rather than land-use activities [64]. Their abundances exceeded the permissible limits of the WHO. The Fe²⁺ concentration ranged from 0.2 to 21.5 mg/L (av. 5.6 mg/L) and 0.1–7.2 mg/L (av. 2.7 mg/L) for shallow and deep aquifers, respectively, while the Mn²⁺ content varied from 0.4 to 49 mg/L (av. 7.3 mg/L) and 0.1–2.9 mg/L (av. 1.1 mg/L) for shallow and deep aquifers, respectively (Table 2). The broader quantities of these metals (Figure 8b,c) were distributed spatially, in agreement with the TDS pattern (Figure 8a), with the groundwater flow focused primarily in the central and southern sections of the area (Figure 7c). The nitrate concentration varied between 0.1 and 5.7 mg/L and 0.1 and 9.2 mg/L for the shallow and deep aquifers, respectively. NO₃⁻ contaminated groundwater in certain regions, which is hazardous to human health and the environment, was most likely caused by agricultural activity [65]. Conversely, diverse fertilizers transmit not only nitrate to the groundwater, but also calcium, magnesium, chloride, and potassium [66], which potentially change the geochemistry of the groundwater.

4.3. Hydrogeochemical Facies

When taking into account the salinity, the water samples were plotted as the total ionic salinity (TIS) (Figure 9a) [67]. For shallow water, the majority presented a TIS between 0 and 20 meq/L, and only one sample exhibited a higher TIS, above 50 meq/L. For deep groundwater, a TIS between 10 and 20 meq/L was noticed, whereas the bulk of the other samples had a lower TIS (<10 meq/L).

The hydrochemical constituents of groundwater were displayed on a Piper layout to highlight the primary hydrochemical facies and to illustrate the numerous operations that govern groundwater [68,69]. The existence of Cl⁻ enhances the anionic triangle, while the prevalence of Na⁺ characterizes the cationic triangle (Na⁺/Cl⁻ water type, or saline water) (Figure 9b). In accordance with the Na⁺-Ca²⁺/SO₄²⁻ type, the evaporation, cation exchange, as well as the leaching and dissolution of marine deposits, have an influence on the aquifer system (a few samples from both aquifers). The groundwater type changed to Ca²⁺-Mg²⁺/Cl⁻HCO₃⁻ and Na⁺/Cl⁻-HCO₃ when this water was combined with the underlying freshwater aquifers in the northern section. It is worth noting that the change in the chemical composition gives rise to a deviation from the conservative mixing line on the Piper’s diagram, which shifts the samples upward into the upper right portion of the diamond throughout the leaching with aquifer material and the cation-exchange process, while moving westward into the diamond’s central portion during mixing [70].

For a better understanding of the hydrochemistry and discrimination of water types, Chadha’s diagram [71] was used. It is evident from the results that about 25% and 56% of the samples of the shallow and deep aquifers, respectively, fall into the Na–Cl water type, reflecting evaporation processes, whereas the remaining 75% and 44% of the samples of the shallow and deep aquifers, respectively, fall into the Ca–Mg–Cl water-type group, reflecting the advanced reverse-ion-exchange reactions (Figure 9c).

4.4. Geochemical Processes and Groundwater Evolution

4.4.1. Dissolution and Weathering

Gibbs diagrams have been used to differentiate between the rock–water interaction and the sources of dissolved ions, such as precipitation, rock weathering, and evaporation, all of which have an influence on the water chemistry [69,72]. This linkage has been frequently utilized to evaluate how the water chemistry developed [17]. The majority of the data were aggregated around the rock-dominance vicinity, signifying that rock weathering and dissolution were the primary processes impacting the groundwater hydrochemistry (Figure 10a,b). Evaporation effectively raised the ion concentrations in some samples in the shallow aquifer in the northern portion, resulting in excessive salinity.

Several scatter plots for major ions serve to highlight the different potential processes that contribute to the evolution of the Bahariya groundwater, such as leaching, mixing, and ion exchange [73]. The Na⁺/Cl⁻ scatter plot (Figure 11a) demonstrates that the majority of the sample (85% of the total samples for both aquifers) plots are below the 1:1 line, suggesting the supremacy of chloride over sodium. This is mostly due to the excessive addendum of chloride salts from the marine source to the water throughout leaching operations. Furthermore, this ratio is larger than unity in the remaining samples (15%), which might be due to the transfer of Na⁺ to Ca²⁺ or Mg²⁺ in relations with clays [74].

The dissolution of calcite and dolomite in the weathered zone is depicted by the Ca²⁺/Mg²⁺ ratio of the groundwater samples [75]. Dolomite dissolution should occur if the Ca²⁺/Mg²⁺ ratio is equal to unity, while calcite dissolution could be indicated if the ratio is greater than one [76]. All samples from both aquifers had ratios less than 1, which may imply dolomite-dissolution activity (Figure 11b).

Furthermore, the sample points on the Ca²⁺+Mg²⁺ against the HCO₃⁻+SO₄²⁻ scatter plot appear on the equiline, suggesting that the weathering of carbonates and silicates is the principal source for these ions [75,77,78,79]. Most of the sample plots are above the 1:1 line and are on the Ca²⁺+Mg²⁺ side versus the HCO₃⁻+SO₄²⁻ side (63% and 75% of the total samples for shallow and deep aquifers, respectively) (Figure 11c), referring to the prevalent carbonate weathering in the hydrogeochemical process. Nevertheless, this ratio was dispersed around the 1:1 line and was relatively distant from the 1:1 line for the remainder of the samples, revealing that the preponderance of carbonate and silicate weathering is consistent with the earlier finding.

For a better understanding of the mineral reactivity in groundwater, the saturation indices (SIs) were determined. The SI readings of the essential minerals in the groundwater are shown in Figure 12a. The negative values indicate that carbonate minerals (calcite and dolomite), sulfates (anhydrite and gypsum), and halite are miserly soluble in groundwater, while iron minerals (goethite and hematite) are supersaturated, causing good screen encrustation and obstruction, which results in significant hydraulic head losses due to overpumping.

4.4.2. Ion Exchange

The concentration of ions in groundwater is controlled by ion-exchange interactions with clay particles [75]. Equations (1) and (2) can be employed to define the direct and reverse ion exchanges, respectively [80]:

2Na⁺ + CaX₂ = 2NaX + Ca²⁺

(1)

Ca²⁺ + 2NaX = CaX₂ + 2Na⁺

(2)

where X indicates the soil exchanger [81]. The substitution of Na⁺ in the water with Ca²⁺ or Mg²⁺ in clay material can generate an oversupply of Ca²⁺ or Mg²⁺ in groundwater, whereas the substitution of Ca²⁺ or Mg²⁺ in the water with Na+ in clay components might cause an imbalance of Na⁺ [82].

The projection of (Ca²⁺ + Mg²⁺) − (HCO₃⁻ + SO₄²⁻) versus (Na⁺ + K⁺) − Cl⁻ in meq/L has been used to explore the potential role of ion exchange in groundwater development. This pattern implies that the Ca²⁺ and Mg²⁺ were controlled by excluding the impacts of other controls (such as carbonate or silicate weathering), while the (Na⁺ + K⁺) Cl⁻ formula shows that the cations K and Na were obtained from sources other than their respective chlorides [83]. The samples should exhibit a line with a slope of 1 if ion exchange is the key mechanism in the system [84,85]. Figure 12b reveals that the slope of the shallow and deep samples is −0.54, suggesting that cation exchange had an impact on the groundwater chemistry of these samples.

Equations (3) and (4) were employed to develop chloro alkaline indices (CAIs) [86], which were then used to figure out the specific ion exchange between the groundwater and its surroundings [80]):

CAI−1 = Cl⁻ − (Na⁺ + K⁺)/Cl⁻

(3)

CAI−2 = Cl⁻ − (Na⁺ + K⁺)/SO₄²⁻ + HCO₃⁻ + CO₃²⁻ + NO₃⁻

(4)

All measurements are in meq/L. Direct ion exchange occurs when Na⁺ or K⁺ in water interacts with Mg²⁺ or Ca²⁺ in rock. Both the aforementioned indices will be positive as Na⁺ or K⁺ decreases in water. The reverse ion exchange, on the other hand, will result in negative indices [86]. Figure 12c demonstrates that the majority of samples (60%) exhibit negative values for both indices, implying reverse ion exchange in the groundwater system. Meanwhile, the remaining samples have positive values for both indices, confirming direct ion exchange.

4.5. Groundwater Parameter Interrelationships

4.5.1. Pearson Coefficient

Interelement relationships generally provide interesting information regarding element sources and pathways [87]. Some statistics were performed for the purpose of evaluating the results.

Table 3. Pearson’s correlation matrix for all parameter pairs of shallow-groundwater-sample data. Bold letters indicate coefficients greater than 0.5.

Variables	TDS	pH	Ca	Mg	Na	K	HOC3	SO4	Cl	NO3	Fe	Mn
TDS	1
pH	−0.49	1
Ca	0.98	−0.61	1
Mg	0.99	−0.44	0.97	1
Na	0.98	−0.37	0.92	0.97	1
K	0.99	−0.55	0.98	0.98	0.96	1
HOC3	0.03	−0.15	0.05	−0.05	0.00	−0.06	1
SO4	0.94	−0.61	0.97	0.95	0.89	0.98	−0.15	1
Cl	0.97	−0.32	0.91	0.97	0.99	0.94	0.03	0.86	1
NO3	−0.27	0.19	−0.20	−0.23	−0.24	−0.20	−0.27	−0.14	−0.29	1
Fe	0.07	−0.65	0.15	−0.04	−0.04	0.11	0.29	0.11	−0.07	0.05	1
Mn	0.19	−0.62	0.33	0.16	0.02	0.27	−0.15	0.44	−0.02	0.04	0.38	1

and

Table 4. Pearson’s correlation matrix for all parameter pairs of deep-groundwater-sample data. Bold letters indicate coefficients greater than 0.5.

Variables	TDS	pH	Ca	Mg	Na	K	HOC3	SO4	Cl	NO3	Fe	Mn
TDS	1
pH	0.50	1
Ca	0.67	0.32	1
Mg	0.83	0.49	0.89	1
Na	0.67	0.37	0.31	0.54	1
K	0.74	0.40	0.57	0.67	0.56	1
HOC3	0.14	0.09	0.25	0.26	0.41	0.21	1
SO4	0.45	0.19	0.75	0.67	0.13	0.30	−0.08	1
Cl	0.71	0.36	0.41	0.64	0.85	0.68	0.12	0.15	1
NO3	0.14	0.07	0.19	0.12	−0.06	0.04	0.04	0.05	−0.13	1
Fe	−0.14	−0.41	0.24	−0.06	−0.38	−0.15	−0.04	0.34	−0.39	−0.08	1
Mn	0.05	−0.22	0.09	0.00	−0.07	−0.07	−0.23	0.28	−0.09	−0.29	0.61	1

provide the correlation inspection of the physicochemical groundwater-quality parameters, using Pearson’s coefficient (r) for the current study. A correlation coefficient is a bivariate technique for measuring the strength of a connection between two hydrochemical variables. A strong correlation coefficient (around 1 or −1) indicates a significant link between two variables, whereas a value of zero indicates no such relationship at a significance level of 0.05. Parameters with r > 0.7 are regarded as strongly associated, whereas those with an (r) between 0.5 and 0.7 are considered moderately correlated. This clarifies that the matrix highlighted a positively strong correlation (r > 0.7) for TDS, with Mg²⁺, Ca²⁺, Na⁺, and Cl⁻ for both aquifers. This demonstrates that the NSAS’s groundwater chemistry is largely influenced by activities that involve these ions. The same is true for Na⁺, Cl⁻, Ca²⁺, SO₄²⁻, and Mg²⁺, which marked a strong linkage. These findings imply that the geochemical characteristics of the water from many samples are similar. The significant relationship between Na⁺ and Cl⁻ (r = 0.85 − 0.99) might indicate that halite decomposition is a key generator of these ions. Due to the substantial links between SO₄²⁻ with Ca²⁺ (r = 0.75–0.97) and Mg²⁺ (r = 0.67–0.95), gypsum and anhydrite, as well as epsomite, are possible sulfate sources. The positively moderate coefficients for SO₄²⁻ with K⁺ (r² = 0.51) and Cl⁻ (r² = 0.56) indicate an anthropogenic influence [47]. Additionally, the occurrence of iron in groundwater is usually linked to the dissolution of iron-bearing geological formations: the lower the pH, the more soluble the iron is, as evidenced by the negative association between iron and the groundwater pH levels (−0.65) [46]. Concurrently, the coexistence of Fe²⁺ with Mn²⁺ is also shown (r = 0.61).

4.5.2. Factor Analyses

FA discovered three leading factors that accounted for 83% of the variations in the shallow-groundwater data, whereas four key factors accounted for 74% of the total variances in the deep-groundwater data (

Table 5. Loads of top three factors (F1, F2, and F3), and communality for each quality parameter obtained by FA of shallow-groundwater samples. Bold letters indicate the factor of which squared cosine takes the maximum.

Variable	F1	F2	F3	Communality
TDS	0.99	0.10	−0.07	1.00
pH	−0.58	0.77	0.07	0.94
Ca2+	0.99	−0.07	−0.02	0.98
Mg2+	0.98	0.18	0.03	1.00
Na+	0.95	0.26	−0.08	0.98
K+	1.00	0.02	0.06	1.00
HCO3−	0.00	−0.17	−0.77	0.63
SO42−	0.97	−0.09	0.24	1.00
Cl−	0.94	0.32	−0.14	1.00
NO3−	−0.23	−0.05	0.32	0.16
Fe2+	0.11	−0.70	−0.19	0.54
Mn2+	0.27	−0.64	0.35	0.61
Eigenvalue	7.13	1.76	0.95
Variability (%)	59.5	14.6	7.9
Cumulative %	59.5	74.1	82.1

and

Table 6. Loads of top three factors (F1, F2, and F3), and communality for each quality parameter obtained by FA of deep-groundwater samples. Bold letters indicate the factor of which squared cosine takes the maximum.

Variable	F1	F2	F3	F4	Communality
TDS	0.89	0.05	−0.13	0.01	0.81
pH	0.51	−0.21	−0.06	0.23	0.37
Ca2+	0.79	0.48	0.23	0.21	0.95
Mg2+	0.94	0.21	0.07	0.16	0.96
Na+	0.75	−0.35	−0.07	−0.38	0.83
K+	0.76	−0.05	−0.06	−0.04	0.58
HCO3−	0.31	−0.24	0.85	−0.36	1.00
SO42−	0.51	0.61	−0.01	0.19	0.66
Cl−	0.80	−0.30	−0.32	−0.27	0.91
NO3−	0.08	−0.03	0.20	0.38	0.19
Fe2+	−0.20	0.84	0.16	−0.25	0.83
Mn2+	−0.06	0.63	−0.27	−0.44	0.67
Eigenvalue	4.75	2.07	1.04	0.91
Variability (%)	39.56	17.28	8.69	7.54
Cumulative %	39.56	56.84	65.53	73.08

). The variables’ spatial distributions in the spaces are specified by the first (F1) and second (F2) factors (Figure 13).

F1 exhibits from strong to moderate positive loadings with TDS, EC, Na⁺, Mg²⁺, Ca²⁺, SO₄²⁻, and Cl⁻, and contributes 59.5% and 39.5% of the total variance in the shallow- and deep-groundwater samples, respectively (more than 0.75). Accordingly, F1 denotes that salinity governs the provincial hydrochemical variability, implying that mineral disintegration has progressed this process [88].

The second factor (F2) of the shallow-groundwater samples, accounting for 14.6% of the total variance, is associated with high negative loadings of Fe²⁺ (−0.7) and Mn²⁺ (−0.64), while there is a strong positive loading value for the pH (0.77). Accordingly, F2 is assumed to be indicative of the natural processes and the water–rock interaction (geogenic sources) because the presence of Mn²⁺ and Fe²⁺ in the groundwater is a direct result of the reductive dissolution of the mineral that typically exists [11]. On the other hand, F2 of the deep-groundwater samples explains 17.28% of the total variance and is associated with high positive loadings of Fe²⁺ (0.84), Mn²⁺ (0.63), and SO₄²⁻ (0.61), which is consistent with the aforementioned conclusion.

Factor 3 contributes 7.9% of the total variance and 0.95 of the eigenvalues, and it is marked by significant negative loadings with HCO₃⁻ (−0.77), which reflects the mixing of freshwater from the underlying deep aquifers with the water of the shallow aquifer. Furthermore, the presence of a weak positive loading value for NO₃⁻ (0.32) suggests the emergence of anthropogenic pollutants due to the implementation of chemical fertilizers on soil, which causes pollution, and thus can be referred to as the agriculture contamination factor [89]. Likewise, F3 for the deep datasets, which contributes 8.7% of the total variance and 1.04 of the eigenvalues, is associated with substantial positive HCO₃⁻ loadings (0.85) and depicts rainwater and groundwater mixing in aquifers [46]. Similar to the shallow-groundwater results, F4 accounts for 7% of the total variance and is positively loaded with NO₃⁻ (0.38), which is relevant to the prior assertion.

5. Conclusions

The current research highlights the significance of multivariate statistical analysis and geochemical modeling to inspect groundwater geochemistry and the factors that control its evolution and spatial variability. The surface of the Bahariya Oasis is dissected by several NE–SW and NW–SE normal and reverse faults. These faults intersect with each other and with the Bahariya mid fault, and they are present in large numbers in the north and northwest segments of the Bahariya depression. The hydrochemical results demonstrate that the majority of the samples were freshwater and were mainly identified in the northernmost section of the area, along fault planes.

Groundwater mineralization is controlled by the leaching and disintegration of marine deposits in freshwater, whether from irrigation, rain, or upward flows from the underlying aquifers. The wide variation in the Fe²⁺ and Mn²⁺ concentrations suggests that their contents were controlled by several intermixed processes. The significant positive relation between Fe and Mn indicates that the Fe source may be attributed to rock weathering. The water samples were largely divided into two categories, according to Chadha’s (1999) suggested hydrochemical processes: Na–Cl and Ca–Mg–Cl. The dissolution/precipitation of aquifer minerals (including calcite, dolomite, gypsum, halite, and iron minerals) was responsible for the geochemical processes in the Bahariya groundwater.

A factorial analysis revealed that three main factors were found to account for 83% and 67% of the dataset for shallow and deep aquifers, respectively, and revealed that ion-exchange, water-mineralization, and carbonate-weathering mechanisms, as well as the imprint of anthropogenic activities, such as fertilizers use, were the major determinants of the region’s overall groundwater geochemistry. The findings of this research can be employed to build a more beneficial agricultural management strategy and for the drilling of groundwater wells into the Nubian sandstone aquifer, and particularly in the central part, to provide significant sustainable groundwater supplies for the study region. In order to establish a comprehensive picture of the state of such a deep aquifer, subsurface lithology and groundwater research should be carried out using geophysical investigation and pore-hole testing. For instance, the overexploitation of groundwater, and the creation of groundwater management and simulation models based on the current outcomes to alter the abuse of groundwater, can be controlled by supplanting outdated irrigation-system frameworks with efficient ones, and by providing a more appropriate scenario for choosing the optimum pumping rate for the NSAS in the region.