DRASTIC Method Research Articles

Assessing Groundwater Vulnerability to Pollution in a Rapidly Urbanizing River Basin Using a Modified DRASTIC Land Use–Lineament Density Method

ABSTRACTGroundwater quality assessment is crucial for ensuring safe drinking water, protecting public health, and maintaining sustainable water resources for agricultural and industrial uses. The Awash River basin faces significant groundwater quality challenges due to rapid population growth, high urbanization, large‐scale irrigation, and industrial pollution. The main objective of this study is to evaluate the intrinsic vulnerability of aquifers to pollution in the Awash Basin and identify hotspots requiring urgent intervention using a modified DRASTIC overlay analysis method. In addition to the seven parameters considered in the generic DRASTIC overlay analysis (depth to the water table, recharge, aquifer media, slope, soil media, vadose zone, and hydraulic conductivity), we incorporated land use/land‐cover (LULC) and lineament density (LD) distributions (DRASTIC‐LU‐LD). This modification allowed us to produce more realistic groundwater vulnerability maps for the basin. To identify the most influential parameters in the overlay, sensitivity analysis was conducted using Map Removal Sensitivity Analysis (MRSA) and Single Parameter Sensitivity Analysis (SPSA). Initially, the generic DRASTIC index in the area ranges from 69 to 181, categorizing the area into four vulnerability zones: very low (21%), low (51%), medium (27%), and high (1%). After incorporating LU and LD, the index values ranged from 90 to 240. Based on the percentage of the total area studied, the inclusion of LU decreased the very low and low vulnerability zones from 72% to 44% and the inclusion of LD increased the high‐ and very‐high‐vulnerability zones from 14% to 27%. The areas most vulnerable to groundwater pollution are in the western (upper Awash), middle Awash, and northwestern regions, particularly in city centers where groundwater abstraction is significant. These high‐vulnerability zones coincide with municipal, industrial, and agricultural pollution sources. The vadose zone parameter has the highest impact in both MRSA and SPSA, with a variation index value of 3.08% and a mean effective weight of 27.93%, respectively. By identifying areas vulnerable to groundwater pollution, this study provides a valuable basis for informed decision‐making and the development of effective strategies for protecting groundwater from urban, industrial, and agricultural pollution sources.

Protection of groundwater resources from multi-hazards: An integration of risk assessment and watershed management

The improving groundwater resource protection depends on aligning contamination risk mapping with aquifer protection measures. However, existing methods often fail to align with aquifer protection goals, hindering practical implementation. The study demonstrated integrating vulnerability assessment, hazard evaluation, and groundwater protection initiatives using an L-Matrix approach. Testing this model in a well-documented watershed with multiple hazards to illustrates its applicability, providing a reference for similar scenarios. The risk was assessed through evaluation of aquifer vulnerability, using the DRASTIC method, coupled with the evaluation of a hazard potential considering industrial, livestock/agriculture and urban infrastructure sources in separate. The L-Matrix diagrams consisted of cartesian representations of V versus H, ranging from 0 to 1, segmented into four quadrants bounded by the cut-off lines V = 0.5 and H = 0.5. The quadrants were then linked to aquifer protection measures, with site inspection (high V and H), groundwater monitoring (low V and high H), land use planning (high V and low H), and tolerable expansion of hazardous activities (low V and H). The Paraopeba River basin was used as test site. Mining areas are abundant in this basin and were allocated high values of H and low to high values of V, requesting implementation of groundwater monitoring and periodic inspection of mining sites, even more because groundwater contamination with iron could be clearly associated with the presence of mine-tailings. In the results, it was observed that the DRASTIC vulnerability map of the aquifers has a mean vulnerability around the central value (V = 0.48 ± 0.08). Regarding the distribution of hazards, the average rounds to the value for to (Hindustry = 0.41 ± 0.37, Hagriculture = 0.33 ± 0.09 and Hinfrastructure = 0.27 ± 0.14). Thus, it is noted that for industrial activities, the risk of contamination could be moderate to high, while for agricultural and livestock activities, the risk lies between low and moderate contamination impact.

Multi-contamination groundwater risk assessment based on integrated OSPRC framework considering receptor and consequence components

This study evaluates the risks associated with geogenic and anthropogenic contaminants in the Azarshahr aquifer situated in the Lake Urmia watershed in NW Iran, an area critically affected by both natural mineral deposits and intensive agricultural activities. This region, characterized by its significant geological diversity and extensive use of pesticides and fertilizers, presents a unique opportunity to study the interaction between natural and human-induced groundwater contamination. Employing the Origin-Source-Pathway-Receptors-Consequence (OSPRC) framework, the research focuses on the linkage between these contaminants and waterborne diseases, particularly examining the less-studied “Receptor” and “Consequence” components using the Groundwater Quality Index (GQI) and health risk indices. Our approach integrates vulnerability assessments through the DRASTIC and SPECTR methods, enhanced by Sugeno fuzzy logic, to produce a detailed risk map highlighting two critical zones: Risk Cell 1, impacted predominantly by agricultural contaminants including nitrate and chlorpyrifos, and Risk Cell 2, affected by geogenic contaminants such as arsenic, lead, nickel, and chromium. This comprehensive analysis not only maps out the source and migration pathways of these contaminants but also evaluates their impact on human health. The findings underscore a strong correlation between identified risks and health impacts, emphasizing the pressing need for targeted health interventions and improved management of water resources in the region. By advancing the application of the OSPRC framework, this research fills a vital gap in our understanding of aquifer contamination dynamics and sets a new standard for future groundwater risk assessments.

Assessment of groundwater vulnerability in a semi-arid basin: a comparative study of DRASTIC and SI methods (case study of Boulefreis Wadi watershed—northeast Algeria)

Assessment of groundwater vulnerability in a semi-arid basin: a comparative study of DRASTIC and SI methods (case study of Boulefreis Wadi watershed—northeast Algeria)

Assessment of risk intensity in the triffa plain aquifer: Integration of hazard quantification, land use analysis, dynamic vulnerability GCITF, and DRASTIC method

This study aims to assess the level of groundwater pollution risk in the Triffa Plain, stemming from various pollution sources, with a specific focus on agricultural intensification and excessive pesticide use. Data were collected from ABHM (the Moulouya Hydraulic Basin Agency) and ORMVAM (Regional Office of Moulouya Agricultural Development), and a field visit was conducted in July 2023. The study utilized two main methodologies for mapping risk intensity: the COST 620 action and overlaying vulnerability maps with the land use (LU) factor. Vulnerability maps were generated using the original DRASTIC method and the new dynamic vulnerability assessment method, GCITF.The research approach involved comparing the COST 620 risk index (Ri) and the land use risk index (LU) to identify the most suitable method for the study area. Due to challenges in inventorying various point sources of pollution, the study focused on mapping the primary sources of pollution (diffuse, point, and linear). The key findings indicate that the predominant pollution intensity in the region is high, covering 60.66% of the total surface area. These results closely align with those obtained using the European COST 620 approach and the spatial distribution of nitrates. The study demonstrates the reliability of integrating the GCITF method in assessing groundwater contamination risk, highlighting the significant negative impacts of agricultural activities on the Triffa Plain aquifer, primarily due to uncontrolled pesticide and fertilizer use and point sources of pollution.This study suggests that implementing targeted technical solutions for specific pollution sources is crucial for reducing contamination in the Triffa Plain. These pollution risk intensity maps can assist planners and decision-makers in understanding the origin of pollution and identifying necessary actions to minimize contamination risk, contributing to better groundwater management and protection in the region.

Assessment of the Vulnerability of Aquifers in Basement Areas to Pollution from Agriculture: The Case of the Boulbi Rice Plain in Burkina Faso

Groundwater is a major source of water, meeting the domestic water needs of more than 70% of Africa's population. Although prized for its relatively good quality compared with surface water, groundwater is increasingly subjected to multiple sources of pollution. Long thought to be the solution to increasing agricultural production and achieving food self-sufficiency, agricultural inputs are now being pointed out in Burkina Faso as a major source of water pollution. However, few studies exist showing the contribution of agricultural inputs to groundwater pollution. The aim of this study is to show the impact of the use of agricultural inputs on groundwater quality: the case of the Boulbi valley rice-growing area in Burkina Faso, West Africa. Soil properties were measured using a double-ring infiltrometer and Harmonized World Soil Database. Groundwater recharge was assessed by Thornthwaite’s equation. The DRASTIC, GOD and SI methods were applied to map the valley’s vulnerability. Fertilizers and phytochemicals were recorded by surveys. A sampling of surface and groundwater was done in 32 locations and the chemical characteristics (pH, EC, NO<sup>3-</sup>, SO<sub>4</sub><sup>2-</sup>, PO<sub>4</sub><sup>2-</sup> and K<sup>+</sup>) confronted with the vulnerability indices. Results show that the soils were predominantly clay (41%), silt (37%) and silty sand (22%). Twenty types of phytochemicals were used, among which 35% were composed of the controversial glyphosate (denounced as carcinogenic) and 30% made with paraquat chloride also accused of being responsible for several self-poisoning. All the three methods pointed to a low vulnerability risk, partly because of the purification role of clay. The average pH is 8.2 ± 0.4, explaining the low-rice yield (<4.0 tons/ha), in spite of fertilizer use. Although the risk assessment rendered non-alarming situation, preventive measures about health and environment need to be taken.

Integrated assessment of groundwater pollution vulnerability in Goulmima: a comparative analysis of DRASTIC and GOD methods

Integrated assessment of groundwater pollution vulnerability in Goulmima: a comparative analysis of DRASTIC and GOD methods

Groundwater vulnerability and risk assessment coupled with hydrogeochemical analysis in the Taza aquifer (Morocco)

This study assesses the vulnerability of aquifers to increased contamination, influenced by rapid peri-urbanization and deficient sanitary infrastructure. The research focuses on the Taza aquifer in northeastern Morocco, where we applied the DRASTIC and GOD models. We further enhanced the DRASTIC method by the addition of a Fracture Index (F) ‘’DRASTICF‘’ to incorporate the effects of tectonics and geological discontinuities on groundwater behavior. The vulnerability mapping was performed for two distinct dry and wet periods of 2016 and 2021 respectively. Moreover, the use of hydrogeochemical analysis of 22 groundwater samples helped to identify potential sources of pollution and to understand the hydrogeological process. The results indicate a dynamic in vulnerability between these two periods due to changes in the recharge parameter and groundwater depth. The comparison between obtained vulnerability maps and nitrate concentrations reveals vulnerable zones that closely resemble regions affected by nitrate pollution. This area is predominantly classified as of high vulnerability at 27 %, associated with an estimated pollution level of 36%. The findings underscore also the importance of integrating the fracture index into the assessment of aquifer vulnerability alongside the different models. This integration offers an enriched perspective for effectively managing risks associated with groundwater contamination. The risk maps of groundwater contamination were also established by overlapping the vulnerability map with the land use layer. The main hydrogeochemical facies determined in the Piper Diagram allowed classifying the waters as Ca–Mg–HCO3 type. The significant correlations indicated that groundwater mineralization in the study area originates from the dissolution of limestone and evaporated formations. The water quality within the study area shows substantial variability, ranging from excellent to unsuitable, primarily due to the leaching of fertilizers and wastewater.The different results can be very helpful to the farmers, and local authorities for better groundwater sustainable management.

Sensitivity Analysis and GIS Tools for Groundwater Vulnerability Assessment. (Application in the Middle Chellif Plain, Algeria)

Contamination due to rapid urban development, industrialization, and agricultural sources is increasingly threatening the groundwater resources of the plioquaternary phreatic aquifer of the Middle Western Cheliff. An intrinsic vulnerability assessment was carried out based on the DRASTIC method on parameters involved in the vertical transfer of pollution from the soil surface. The vulnerability maps obtained showed a high extension of areas of medium vulnerability (54%). On the other hand, areas of low vulnerability occupied about 36%. The high and very high vulnerability was mainly related to low water depth values (8% and 2%, respectively). The results of a sensitivity analysis using the two tests—the single parameter sensitivity analysis and the map removal sensitivity analysis—show that every parameter is, without fail, required for the computation of the vulnerability index. The validation of the vulnerability map produced by the DRASTIC method confirmed the evolution of this sensitivity which decreases towards the S.W. of the plain with nitrate concentrations between 30 and 120 mg/L. Planners can use the produced risk maps as tools to make a preliminary choice of priority locations for various forms of environmental sustainability.

Groundwater potential assessment in parts of Nnewi Industrial Zone: Implications for sustainable development and conservation

Integrated field geological and geophysical (Vertical Electrical Sounding VES and 2D Electrical Resistivity Tomography ERT) investigations were carried out in the Nnewi Industrial Zone, SE Nigeria to identify and characterize productive aquifer, aquifer hydraulic properties, and aquifer vulnerability. Data acquired from 17 VES points and across a profile line of about 800 m for the 2D ERT were processed, interpreted, and modeled. Results from the geologic mapping showed that the outcropping units are mainly the sandstone, shaly-sandstone, and less commonly shale of the Nanka and Ogwashi-Asaba Formations which are relatively permeable and offer poor protective cover to the underlying aquifer. Models from the resistivity data showed that the depth-to and thickness of the aquifer vary from 38.60 to 98.80 m and 30.10–177 m, respectively. Aquifer properties estimated from the geophysical data gave values ranging from 0.611127 m2/day to 246.6576 m2/day and 0.1609 m/day to 5.6325 m/day for transmissivity and hydraulic conductivity, respectively, and suggestive of low – moderate aquifer potential. Aquifer Protective Capacity APC distribution modeled from the longitudinal conductivity values, and the aquifer vulnerability modeled using the DRASTIC method indicate that the study area is characterized by poor – moderate APC and low – moderate – high aquifer vulnerability, respectively. Analysis shows that the modeled aquifer parameters, APC, and aquifer vulnerability have similar trend which tends to improve towards the southern and more specifically southeastern parts of the study area, suggesting that even though aquifer units were identified all through the study area, the southeastern parts are best suited for the development of groundwater exploitation schemes. Also, aquifer vulnerability model results recommend that proper and efficient waste disposal schemes are put in place to conserve groundwater quality from pollution from industrial waste since the aquifer in the area is relatively vulnerable.

Groundwater Vulnerability Comparison Using DRASTIC and GOD Methods in Surakarta City

Groundwater Vulnerability Comparison Using DRASTIC and GOD Methods in Surakarta City

GROUNDWATER VULNERABILITY MAP OF BASARA BASIN, SULAIMANI GOVERNORATE, IRAQI KURDISTAN REGION

The Basara basin which is one of the most promising hydrogeological basin in Iraqi Kurdistan Region, located in north east of Iraq, 25 km west of Sulaimani city, between the 496652 - 537752 East and 3911038 - 3951906 North in Universal Transverse Mercator (UTM) and lie in Zone 38N. The basin has a rectangular shape and covering an area of 571 km². The present hydrogeological investigations have revealed three inhomogeneous and anisotropic water bearing formations: Eocene Karstic Fissured Aquifer (EKFA), Intergranular Aquifer represented by Alluvium and Pliocene (AIA) & (PIA), as well as Miocene Complex Aquifer (MCA).For the first time, not only in Kurdistan but also in Iraq, Groundwater Vulnerability Map has been constructed in this study, using DRASTIC method with the assistance of Geographic Information System (GIS) to show zonation area of high and low groundwater susceptibility to pollution. Accordingly, vulnerability classes of the study area were classified into four classes. Most of the basin shows the highest extension of the zones with very low and low vulnerability zones, in contrast the zones with high vulnerability are distributed mainly in the mountain areas, solely in the eastern Uloblagh and Kuwaik mountains, in addition to that small zones in the farthest northern corner and south western corner of the area have less or no human activity.

Validation of the intrinsic vulnerability to pollution of fractured siliciclastic aquifers using natural background levels

Aquifers are important water reserves in water-scarce areas. However, uncontrolled human activities and lack of knowledge of aquifer characteristics increase their vulnerability to pollution. Assessing intrinsic vulnerability is essential for improving groundwater management. However, traditional tools, such as DRASTIC, ignore relevant parameters that, in particular contexts, lead to underestimating vulnerability. This work assesses the intrinsic vulnerability to pollution of a siliciclastic aquifer system (Shallow and Upper) by comparing the DRASTIC and modified DRASTIC methods. Modified DRASTIC included parameters such as Land Use and fracturing patterns relevant in populated areas where hydrogeological configurations involve highly cemented and fractured siliciclastic rocks with secondary porosity that define the main pathways for transporting contaminants from the surface. The intrinsic vulnerability was validated using Natural Background Levels (NBLs) of contaminant indicator ions (NO3−, PO43− and SO₄2-) to identify areas where the concentration of these indicators surpassed the NBLs. These areas were compared with the vulnerability classes from DRASTIC methods. NBLs were: 2.12 mg/l, 0.05 mg/L and 26.77 mg/L (Shallow aquifer), and 3.17 mg/L, 0.17 mg/L and 5.06 ml/L (Upper aquifer) for the ions of NO3−, PO43− and SO₄2-, respectively. The ion concentrations surpassing NBLs were spatially distributed, being more notable in the dry season and showing a greater coincidence with the high vulnerability class of the Modified DRASTIC. Thus, the modified DRASTIC better represented the high vulnerability of the aquifer system to pollution and the negative effect of anthropic activities. Additionally, Pearson's coefficients were calculated between the concentrations of contaminant indicator ions and intrinsic vulnerability indices of DRASTIC methods. However, the coefficients lack statistical significance. These results support the relevance of adding land use and fracturing patterns to assess the intrinsic vulnerability of fractured, siliciclastic aquifers in populated regions and show NBLs as a better validation method than Pearson's coefficients in data-scarce contexts.

Groundwater vulnerability assessment using a GIS-based DRASTIC method in the Erbil Dumpsite area (Kani Qirzhala), Central Erbil Basin, North Iraq

Groundwater vulnerability assessment using a GIS-based DRASTIC method in the Erbil Dumpsite area (Kani Qirzhala), Central Erbil Basin, North Iraq

Comparative study of the vulnerability to groundwater pollution by the DRASTIC and SINTACS methods on the Amizour plain (Béjaia, North Algeria)

The present study deals with the vulnerability and the risks of pollution on the Amizour plain aquifer (north Algeria), already threatened by several sources of pollution (e.g., industries, agriculture, illegal dumping, etc.) that collapsed this region, without any planned environmental protection measures. In the aim to study the sensitivity of the Amizour plain against pollution two methods (DRASTIC and SINTACS models) were used and the results were compared and evaluated. The maps showed a similarity in the degrees of vulnerability ranged from high to medium. Most of the plain is affected by a high degree of vulnerability of 76% and 67% as estimated by DRASTIC and SINTACS, respectively. In fact, such remarkable degree is justified by the shallow position of the groundwater and the type of aquifer. The degree of groundwater vulnerability was 18% and 33% in the northern and southern part of the studied zone as estimated by DRASTIC and SINTACS methods, respectively. According to the DRASTIC approach, an area with a very high vulnerability (6%) was observed amidst the Amizour plain. Superimposing both maps of vulnerability and the pollution source sites allowed us through the risk map to frame the areas at high risk of groundwater contamination. This investigation will facilitate to make decisions in implementing of an accurate and urgent management project for safeguarding the studied zone. The applicability of these findings has been discussed and suggestions for attenuating the risk of contamination have been given.

Enhancing groundwater vulnerability assessment for improved environmental management: addressing a critical environmental concern

Groundwater serves as a primary water source for various purposes. Therefore, aquifer pollution poses a critical threat to human health and the environment. Identifying the aquifer’s highly vulnerable areas to pollution is necessary to implement appropriate remedial measures, thus ensuring groundwater sustainability. This paper aims to enhance groundwater vulnerability assessment (GWVA) to manage aquifer quality effectively. The study focuses on the El Orjane Aquifer in the Moulouya basin, Morocco, which is facing significant degradation due to olive mill wastewater. Groundwater vulnerability maps (GVMs) were generated using the DRASTIC, Pesticide DRASTIC, SINTACS, and SI methods. To assess the effectiveness of the proposed improvements, 24 piezometers were installed to measure nitrate concentrations, a common indicator of groundwater contamination. This study aimed to enhance GWVA by incorporating new layers, such as land use, and adjusting parameter rates based on a comprehensive sensitivity analysis. The results demonstrate a significant increase in Pearson correlation values (PCV) between the produced GVMs and measured nitrate concentrations. For instance, the PCV for the DRASTIC method improved from 0.42 to 0.75 after adding the land use layer and adjusting parameter rates using the Wilcoxon method. These findings offer valuable insights for accurately assessing groundwater vulnerability in areas with similar hazards and hydrological conditions, particularly in semi-arid and arid regions. They contribute to improving groundwater and environmental management practices, ensuring the long-term sustainability of aquifers.

Application of DRASTIC method to identify the groundwater vulnerability to pollution in the sub-district of West Limboto and Limboto, Gorontalo Regency, Indonesia

The increment in population and agricultural sectors in West Limboto and Limboto, Gorontalo, affects the increased production of pollutants. The community uses shallow groundwater as the main water source. Since shallow groundwater is vulnerable to pollution due to community activity, this study aims to identify its vulnerability in West Limboto and Limboto. It was identified using the DRASTIC Method that will be integrated with Geographic Information System (GIS) Technique. The used parameters are Depth to Water Table, Recharge, Aquifer Media, Soil Media, Topography, Impact of Vadoze Zone, and Hydraulic Conductivity of the Aquifer. Depth to Water Table data is obtained from direct field measurement. Recharge amount data is obtained from the calculation of groundwater recharge. Aquifer Media, the Impact of Vadoze Zone, and Soil Media data are obtained from surface and subsurface geological data. Topography data is obtained from Digital Elevation Model Map. Hydraulic Conductivity is obtained from Pumping Test Data. The analysis result shows that the groundwater vulnerability of the study area was classified into 2 classes, the high index was in the southern part, and the low index was in the middle to the northern part of the study area.

A comparison of methods for assessing groundwater vulnerability in karst aquifers: the case study of Terminio Mt. aquifer (Southern Italy)

The assessment of groundwater vulnerability to pollution is becoming even more important all over the world due to the increase of impacts of human activities on groundwater resources and the related risks to the human health, economics, and the environment. Owing to the variability of methods known for estimating groundwater vulnerability, basically depending on hydrogeological parameters considered and the scale of analysis, the comparison of results of different methods appears straightforward for identifying the best approach in a given hydrogeological condition and reference scale. In such a view, this work attempts to assess the groundwater vulnerability of the Terminio Mt. karst aquifer, by applying four different groundwater vulnerability methods, index-based, and comparing results in order to identify the best performing one in karst environments. The study aquifer, located in the Picentini Mts Regional Park (Campania region, southern Italy) represents a strategic drinking water resource since Roman times and hosts massive groundwater resources which outflow mainly from tapped basal and subordinately perched springs.The peculiar characters of the study karst aquifer, which favour direct infiltration and groundwater recharge processes, as well as the occurrence of industrial, agricultural and grazing activities, make it very vulnerable to groundwater pollution, thus requiring a proper and careful territorial management.Beside the most frequently and generally used methods for assessing groundwater vulnerability, such as the DRASTIC and SINTACS, also DAC and COP methods specifically designed for karst aquifers were applied and mutually compared. Results of SINTACS, DRASTIC and DAC methods show groundwater vulnerability maps of the Terminio Mt. karst aquifer as chiefly characterized by two classes of intrinsic groundwater vulnerability, varying between the medium and high degrees. Furthermore, high and extremely high values of groundwater vulnerability were found in areas controlled by the shallow depth of the water-table. Instead, the COP method resulted as the most effective in identifying the endorheic areas and the related karst morphologies as very high groundwater vulnerability zones, therefore the most suitable in capturing specific hydrogeological features of karst areas that control groundwater pollution and vulnerability.Results obtained will support decision tools aimed at the land use planning and protection of karst aquifers from pollution in karst areas.

Assessment of groundwater vulnerable zones using conventional and Fuzzy-AHP DRASTIC for Visakhapatnam district, India

Given the increasing impact of extreme climate change and the unpredictability of monsoons, the importance of preserving groundwater resources has never been more critical. Conducting a thorough groundwater vulnerability assessment enables one to pinpoint the zones most susceptible to contamination and take proactive measures to safeguard them. This paper is an attempt to enhance the conventional DRASTIC method by integrating Fuzzy Logic and Analytical Hierarchy Processes (FAHP) and applying this approach to the Visakhapatnam district. Spatial data for the seven parameters used in this study were created and analyzed using ArcGIS. The study determines and compares the groundwater vulnerability index using both the conventional DRASTIC method and the FAHP-DRASTIC method. Based on the index values, the groundwater zones in the study area have been classified into five categories: Very Low, Low, Medium, High, and Very High Vulnerability. Model validations were conducted using the ROC-AUC technique against Nitrate concentration. The results reveal that the FAHP-DRASTIC model, with an AUC of 0.92, is the most suitable model for the present study.

Enhancing groundwater vulnerability assessment: Comparative study of three machine learning models and five classification schemes for Cuddalore district

Most of the groundwater vulnerability assessment methods using machine learning are binary classification. This study attempts multi-class classification models to map the groundwater vulnerability against Nitrate contamination. Further, the significance of the number of classes used in the multi-class classification is studied by considering three and five classes. Three machine learning models, namely Random Forest, Extreme Gradient Boosting and CART, with two classification schemes, were developed for the present study. The parameters used in the conventional DRASTIC method and with an additional parameter, Landuse, have been employed for the study. Evaluation metrics such as Accuracy, Kappa, Positive Predictive Value, Negative Predictive Value, and Area Under the Curve of the Receiver Operating Characteristic (AUC-ROC) were compared among all six models to select the optimal one. Based on the model evaluation metrics and consistent distribution of area among the classes Random Forest model with a three-class classification with an AUC of 0.95 is considered optimum for the selected objective. This study highlights the importance of the data classification process and the selection of the number of classes for ML model prediction in assessing groundwater vulnerability. Leveraging the effectiveness of the Geographic Information system and advanced machine learning techniques, the proposed approach offers valuable insights for enhanced groundwater management and contamination mitigation strategies.