Groundwater Vulnerability Mapping Using the Susceptibility Index (SI) Method and Tritium Isotopes: A Case Study of the Gharb Aquifer in Northwestern Morocco

.


Introduction
Today, there has been a notable rise in the need for fresh water in Morocco's semi-arid areas [1,2].This increased demand stems from multiple reasons, such as reduced rainfall, population growth, the expansion of industrial and agricultural pursuits, and deteriorating surface water quality [3,4,5,6].The efficient use of groundwater resources is crucial, and measures need to be taken to prevent contamination.The most efficient method for managing and tackling groundwater pollution is to introduce early-stage mitigation strategies [7,8].The protection and management of groundwater require specific strategies and guidelines to target particular land areas and prevent contamination.Effectively utilizing and developing groundwater in a safe manner requires a comprehensive grasp of the aquifer system and the hydrological conditions within the area, enabling the creation of a map outlining vulnerability to groundwater pollution [9,10].
Groundwater vulnerability is a crucial concept, acknowledging the idea that the physical environment can play a substantial role in protecting groundwater from human-made impacts or influences originating from natural sources, especially pollutants that infiltrate the subsurface environment [11,12].Groundwater vulnerability assessments are now integral to land use planning and sustainable socio-economic development, aligning with the goal of preserving groundwater resources.These assessments play a crucial role in preventing potentially harmful activities that could impact groundwater quality.They involve delineating groundwater zones, each exhibiting varying degrees of sensitivity.Given that aquifers are susceptible to contamination from anthropogenic and land-use influences, addressing groundwater pollution is a significant challenge.Groundwater, by its nature, possesses inherent sensitivity, making certain aquifers vulnerable to pollution based on geological and hydrogeological factors [13,14].Groundwater vulnerability corresponds to the inclination or likelihood of common contaminants mixing with groundwater as it enters the subsurface environment.It's also noteworthy that susceptibility to contamination can vary across different land segments [15].Nitrates, frequently utilized as fertilizers in agriculture, stand as a prevalent origin of groundwater pollution [16].They can also be considered potential indicators of contaminant movement because they do not occur naturally in groundwater.
Assessing groundwater threats is critical to ensuring adequate water supplies and mitigating adverse impacts.Therefore, assessing the vulnerability of groundwater to contamination is currently considered one of the most important studies.Various models and methods have been developed and tested globally to assess and map groundwater threats [17,18,19,20,21,22].These methodologies combine diverse hydrogeological parameters to generate maps illustrating the susceptibility of groundwater to contamination [23].Models such as DRASTIC, SEEPAGE, AVI, SINTACS (for pinpointing high vulnerability zones via surface water-aquifer interactions), GOD (applicable for vulnerability mapping on a small to medium scale in karst carbonate aquifers), and EPIK (for broader scales) stand as examples.Additionally, the German Geological Agency GLA method and GOD sensitivity index (SI) method initially proposed by Ribeiro [24] and further developed by Anane et al. [25] have evolved into an integrated approach for assessing groundwater threats.The SI method assesses the impact of a variety of factors, including land use, fertilizers, human activity and natural impacts, as potential sources of groundwater contaminants.Each parameter is assigned a value between 1 and 10, reflecting its importance in the vulnerability map.Implementing these methods involves several key steps, including analyzing the raw data, rendering feature rankings on a map, integrating the maps, and then classifying the integrated maps based on index values.
The purpose of this study is to assess groundwater vulnerability and develop a map delineating pollution-sensitive zones in the Garb aquifer in northwest Morocco.The study area includes a large portion of agricultural land (79.73% of the total study area), which has a significant impact on groundwater quality.The DRASTIC method does not consider the impact of human activities on groundwater quality.Therefore, this study adopts the SI method, where "LU" represents a new parameter called land use.Various factors considered in developing SI models include groundwater depth, net recharge, aquifer media, topography, and land use.This study demonstrates the importance of SI methods in providing an efficient and cost-effective method for groundwater vulnerability and risk mapping studies that can also be applied in other fields.The study also provides scientific guidance for the protection and management of groundwater resources in the region.

Description of the Study Area
The Gharb Plain is situated in the northwestern part of Morocco, spanning the low valley of the Sebou River and covering an area of over 3000 km 2 .It encompasses a substantial aquifer system and is surrounded by the Mamora Plain to the south, the Dradère-Souière Basin to the north, the Atlantic Ocean to the west, and conglomerate outcrops marking the basin's eastern boundaries.The study area comprises a thick sedimentary series with highly diverse deposits, ranging from Miocene marls to recent Quaternary silts [26].The Gharb Plain experiences a climate that varies from sub-humid in the coastal zone to temperate in winter and semi-arid with hot winters further inland.The average annual precipitation stands at 460 mm.Surface drainage is primarily managed by the Sebou River, contributing approximately 6 billion m 3 /year.The region is predominantly agricultural and has undergone significant hydraulic developments and agricultural improvements since 1933.Gravity irrigation is the most commonly employed method, with commonly irrigated crops including forage, sugarcane, and sugar beets [26].

Materials and methods
The assessment of groundwater vulnerability in the Gharb phreatic aquifer relies on the SI method within the GIS environment.To apply this method, various spatial data inputs are required.Initially, the GIS software ArcGIS 10.4 is utilized to create different raster layers, forming the basis for generating the ultimate vulnerability map [27].For the generation of a percentage slope map in the study area, ASTER DEM satellite data with a spatial resolution of 10 meters are utilized.A satellite image derived from a Land Cover Land Use (LC/LU) classification, conducted in 2022 using Sentinel-2 satellite data from the Esri Microsoft Impact Observatory source, is analyzed to identify different land use categories within the region.Groundwater levels from 20 observation wells were collected, and the recorded water level data were imported and interpolated using the Inverse Distance Weighting (IDW) technique within a GIS environment to create the depth-to-water level map.

The SI method
The Susceptibility Index (SI) is a parametric approach utilized to evaluate the precise vertical susceptibility of groundwater to pollution, especially stemming from agricultural activities.It was specifically developed for the assessment of aquifer vulnerability [28].The SI method is based on the DRASTIC model, keeping only four original parameters (water table depth, recharge, aquifer media, and topography), and adding a new parameter, land use, as suggested by Ribeiro [24].The SI is computed using the following formula: In the formula, D represents the water table Depth, R stands for aquifer Recharge, A denotes Aquifer media, T represents Topography, and LU corresponds to Land use.The variables "p" and "c" represent the weight and the rating value assigned to each parameter, respectively [9].While the classes and ratings for the first four parameters remain consistent with those employed in the DRASTIC model, the weights associated with them, as proposed by Aller et al. [29], are multiplied by 10.The resulting SI values, indicating aquifer vulnerability, are categorized into four classes representing distinct vulnerability degrees, as outlined by Ribeiro [24].

Results
To delineate groundwater hazard and risk zones in the study area, five influencing factors were identified, including groundwater depth, net recharge, aquifer medium, topography, and land use (LU).These topic-level characteristics are explained below:

Depth to Water Level (D)
In general, an aquifer's ability to protect groundwater is directly related to its depth.Therefore, a deeper aquifer has a higher protective potential [30].The depth of the aquifer is categorized into five classes, ranging from 6.7 to over 70 meters (Figure 2A), illustrating spatial variability.The depth of the water table is one of the most critical parameters for assessing pollution vulnerability because it indicates the distance contaminants must travel before reaching the aquifer.Consequently, the risk of contamination is higher when the water level in aquifers is high [31].In this study, groundwater levels from 20 observation wells were collected.The recorded water level data were imported and interpolated using the IDW technique within a GIS environment to create the depth to water level map (Figure 2A).It's noteworthy that the high vulnerability zone, ranging from 6.7 to 25 meters in depth, accounts for 11.81% of the total area of the Gharb aquifer.Conversely, the zone of moderate vulnerability, between 25 and 35 meters in depth, covers 85% of the total area.On the other hand, the low vulnerability zone, corresponding to a depth greater than 35 meters, only covers 2.12% of the total area.

Net Recharge (R)
Recharge or infiltration is the movement of water per unit area of soil, penetrating the surface and reaching the water table.Precipitation is a major source of groundwater recharge and provides a pathway for surface contaminants to be transported vertically to the water table and horizontally to the aquifer.Areas with high recharge rates have the highest vulnerability to groundwater contamination, while areas with low recharge rates have the least vulnerability [27].The net recharge of the study area was determined by collecting rainfall data from 2011 to 2021 at more than 30 stations within the study area, using the average annual precipitation data obtained from NASA Power Data.The block-wise

Topography (T)
Topography in a given area reflects the variations in steepness or slope on the Earth's surface.Regions with lower slope angles tend to retain water for extended periods, facilitating more substantial water seepage into the ground.As a result, these areas are highly vulnerable to contamination [11].Conversely, areas characterized by steep slopes, leading to excessive runoff and reduced infiltration, are less susceptible to groundwater contamination [31].In this study, ASTER DEM satellite data, with a spatial resolution of 10 meters, are employed to generate a percentage slope map of the study area.The slope varies from 0% to 40% and is further divided into five categories: 0-1, 2-2, 3-4, 5-7, 8-11 and 12-40.The majority of the study area exhibits slopes ranging from 0% to 10%, making it vulnerable to groundwater pollution.In contrast, the minority of the study area, characterized by slopes higher than 15%, is less vulnerable.The resulting slope map is shown in the Figure 2C.

Land Use (LU)
Anthropogenic activities, including industrial waste, wastewater, mining, and agricultural practices, are the primary contributors to groundwater quality degradation [28].A satellite image derived from an LCLU (Land Cover Land Use) classification using Sentinel-2 satellite data (classification conducted in 2022) from the Esri Microsoft Impact Observatory source was analyzed to identify various land use categories within the region.These categories include agricultural lands, built-up areas, forests, tree-covered areas, bare soils, and water bodies.The results indicate that the predominant land use class in the study area is agricultural land, covering 3055,03 km 2 (79,73%), followed by rangeland at 296,98 km 2 (7,75%) and built-up areas at 284.01 km 2 (7,41%).The resulting land use map is displayed in the Figure 2B.

Groundwater aquifer vulnerability
The vulnerability map obtained (Figure 2F) reveals a Susceptibility Index (SI) ranging from 31 to 160, indicating varying degrees of aquifer vulnerability from low to high.Regions characterized by low vulnerability cover 3% of the total aquifer and are primarily found in forest and rangeland areas with considerable depths to the water table.The medium vulnerability zone encompasses 70% of the aquifer and includes irrigated perimeters, urban areas, and industrial regions.Factors contributing to this moderate vulnerability include low slope, agricultural land, high recharge rates, and clay-sand soil.The high vulnerability area, constituting 26% of the aquifer, is predominantly located in a significant portion of the irrigated perimeter.These high vulnerability zones are chiefly influenced by factors such as low slope, sandy soils, low water table depth, and a very high recharge rate.Therefore, the vulnerability map analysis indicates that the Gharb aquifer exhibits moderate to high vulnerability, primarily associated with factors such as a high recharge rate, the presence of clay-sand and sandy levels in the aquifer lithology, low slope, and the prevalence of extensive agricultural activity in land use.

Validation of Groundwater Pollution Map
Validation is an essential procedure to ensure the attainment of accurate and reliable results.Without validation, models lack scientific significance.The delineated Susceptibility Index (SI) map was validated using nitrate (NO 3 ) and tritium ( 3 H) concentrations, two groundwater pollutants.Nitrate does not naturally occur in groundwater but infiltrates the aquifer system due to the excessive use of fertilizers in agriculture [26].To investigate the correlation between the SI index and NO3 and 3H concentrations, 63 groundwater samples were collected for NO3 analysis.On one hand, chemical analysis results for all water samples showed that NO3 concentrations ranged from 0.03 to 257.72 mg/l.Among these, 23 water samples fell within the moderate to high-risk pollution zones, ranging from 50 to 257.72 mg/l, which exceeded the permissible limit of 50 mg/l (Figure 2G).The NO3 concentration map was overlaid onto the SI index map using the 'Spatial Join' tool in ArcGIS, allowing data to be transferred from one layer to another.Through this operation, the index values of both maps were extracted and spatially joined in the attribute file of NO3 samples.The concordance rate between nitrate concentrations in groundwater and moderate to high vulnerability degrees was 83.12%, as shown in (Figure 3A), which illustrates the distribution of NO3 vs. the Vulnerability index.2H).Furthermore, the distribution map of tritium in groundwater reveals that high 3 H concentrations are found in the northern and southwestern zones, gradually decreasing towards the central and northeastern parts of the plain.Elevated values are also observed in the eastern portion of the aquifer.The low tritium levels observed in the central part of the aquifer suggest that ancient recharge is responsible for these waters, as the average tritium content in the waters is approximately 0.54 TU.This can be explained by the slow groundwater movement due to the low permeability of the geological layer (clay layer).However, this observation does not apply universally to all samples, as there are areas where tritium levels exceed 2 TU, indicating a mixture of both ancient and recent waters, signifying active recharge.The heterogeneous distribution of tritium activities within the aquifer highlights the presence of two recharge zones occurring during different periods: firstly, ancient waters originating from infiltration in the central and northeastern parts of the plain, and secondly, recent waters recharged by infiltration in the northern and southwestern areas of the plain.Given the observed young age of the groundwater, it can be predicted that the aquifer is continually replenished and potentially sustainable for exploitation.However, this also

Discussion
The Gharb region is mainly agricultural, with extensive use of chemical fertilizers and pesticides.Combined with household emissions, this poses an ongoing risk to groundwater quality [32,33].The mapping of vulnerability to pollution of the shallow aquifer in the study area identified three vulnerability classes: low vulnerability, moderate vulnerability, and high vulnerability, which deserve special attention in terms of protection.The moderate vulnerability class represents the largest area, accounting for 71.47% of the total, highlighting a predominance of moderate intrinsic vulnerability in the study region.Similar conclusions, showing three vulnerability classes, were observed by Hamza et al. [34] in their study on diffuse agricultural pollution in the semi-arid groundwater regions of northeastern Tunisia, as well as by Batchi et al. [35] in their comparative study of two models (DRASTIC and SI) assessing the sensitivity to agricultural pollution in the Mnasra groundwater aquifer (northwest Morocco).The values of the vulnerability index SI calculated in this study range from 31 to 84.These results are consistent with those obtained by Batchi et al. [35] in the Mnasra groundwater aquifer (northwest Morocco) (40 to 82).Trends in groundwater vulnerability to pollution in the Gharb aquifer was confirmed by nitrate concentrations in the wells within the study area.The concurrence rate of nitrate concentrations in groundwater with the designated vulnerability classes is 83.12%.Similar findings have been reported in various studies.Specifically, research conducted by Hamza et al. [34], AKE et al. [28], and Batchi et al. [35] has underscored that the susceptibility to nitrates is more accurately depicted by the Specific Vulnerability Index (SI) method.

Conclusion
The study delves into the Gharb aquifer in Morocco, a vital source for both drinking and irrigation water.However, the escalating demands of a rapidly growing economy, coupled with the widespread use of chemical fertilizers, have resulted in groundwater contamination and land degradation.To tackle this pressing issue, we conducted a vulnerability assessment utilizing the Susceptibility Index (SI).The assessment considered five crucial parameters: groundwater depth, recharge, slope, soil type, and land use.The outcomes categorized the region into low, medium, and high-risk zones, determined by index values ranging from 31 to 160.Spatial analysis unveiled notable diversity and moderate sensitivity, particularly in areas with shallow groundwater and sandstone dominance in the northwest and southwest.These vulnerable regions comprise 3%, 70%, and 26% of the total area, equivalent to 101 km 2 , 2725 km 2 , and 986 km 2 , respectively.Furthermore, tritium isotope analysis and nitrate content verification affirmed the accuracy of the hazard map, exhibiting a robust correlation with the hazard (R 2 values of 0.86 for tritium and 0.55 for nitrate concentration).This comprehensive assessment yields valuable insights into crafting effective strategies for safeguarding the Gharb aquifer and preserving Morocco's precious water resources.

Fig 1 :
Fig 1: General map of Morocco and location of the study area.The study area comprises a thick sedimentary series with highly diverse deposits, ranging from Miocene marls to recent Quaternary silts[26].The Gharb Plain experiences a climate that varies from sub-humid in the coastal zone to temperate in winter and semi-arid with hot winters further inland.The average annual precipitation stands at 460 mm.Surface drainage is primarily managed by the Sebou River, contributing approximately 6 billion m 3 /year.The region is predominantly agricultural and has undergone significant hydraulic developments and agricultural improvements since 1933.Gravity irrigation is the most commonly employed method, with commonly irrigated crops including forage, sugarcane, and sugar beets[26].

Fig 2 :
Fig 2 : Mapping input and output for SI: (A) Water table depth; (B) Aquifer recharge; (C) Land slope; (D) Land use map (E) Soil map (F) Vulnerability map based on SI method.(G) Distribution of Nitrate in the generic SI maps and (H) Distribution of Tritium in the generic SI maps

Fig 3 :
Fig 3: The scatter plot between A) the vulnerability indices and nitrate B) the vulnerability indices and TritiumOn the other hand, Figure2Hdisplays the distribution of 3 H vs. the vulnerability index.It indicates that tritium levels in groundwater exhibit heterogeneous temporal variability, ranging from 0.01 TU to 3.6 TU within the study area (Figure2H).Furthermore, the distribution map of tritium in groundwater reveals that high 3 H concentrations are found in the northern and southwestern zones, gradually decreasing towards the central and northeastern parts of the plain.Elevated values are also observed in the eastern portion of the aquifer.The low tritium levels observed in the central part of the aquifer suggest that ancient recharge is responsible for these waters, as the average tritium content in the waters is approximately 0.54 TU.This can be explained by the slow groundwater movement due to the low permeability of the geological layer (clay layer).However, this observation does not apply universally to all samples, as there are areas where tritium levels exceed 2 TU, indicating a mixture of both ancient and recent waters, signifying active recharge.The heterogeneous distribution of tritium activities within the aquifer highlights the presence of two recharge zones occurring during different periods: firstly, ancient waters originating from infiltration in the central and northeastern parts of the plain, and secondly, recent waters recharged by infiltration in the northern and southwestern areas of the plain.Given the observed young age of the groundwater, it can be predicted that the aquifer is continually replenished and potentially sustainable for exploitation.However, this also increases its susceptibility to contamination.Furthermore, this confirms the validation of the used method, where young waters indicate areas with a high vulnerability, and vice versa.