Comparative Assessment of Analytical Hierarchy Process (AHP) and Fuzzy Overlay Analysis (FOA) Models in Groundwater Potential Zone Mapping Using Sensitive Analysis: A GIS-RS Integrated Approach

doi:10.21203/rs.3.rs-4368423/v1

Download PDF

Research Article

Comparative Assessment of Analytical Hierarchy Process (AHP) and Fuzzy Overlay Analysis (FOA) Models in Groundwater Potential Zone Mapping Using Sensitive Analysis: A GIS-RS Integrated Approach

https://doi.org/10.21203/rs.3.rs-4368423/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

This study addresses the pressing need for assessing groundwater potential in diverse regions worldwide, emphasizing the challenges posed by industrialization and urban expansion. Leveraging remote sensing (RS) data within an ArcGIS framework, a comparison was conducted on the effectiveness of the Analytical Hierarchy Process (AHP) and Fuzzy Overlay Analysis (FOA) models in delineating groundwater potential zones. Ten influencing factors underwent rigorous multicollinearity checks to ensure independent analysis. Both methodologies categorized the output into five classification zones, revealing variations between methods. The sensitivity analysis identified specific influential layers for each method, showcasing heightened sensitivity in assessing groundwater potential zones. Notably, AHP emphasized factors such as NDWI, Lineament density, and Land Use Land Cover (LULC), while FOA prioritized Soil Texture, Lineament Density, and NDWI. Evaluating the reliability of groundwater potential assessments, AHP demonstrated fair agreement (59.18%), while FOA exhibited substantial agreement (77.5%). FOA emerged as superior, offering a more nuanced and precise representation of spatial relationships and uncertainties. The promising performance of FOA in our study suggests its potential applicability in accurate groundwater potential assessment beyond the study area (India) to other countries regions with heterogeneous hydrogeological conditions.

Groundwater potential

Fuzzy Overlay Analysis (FOA)

Analytical Hierarchy Process (AHP)

Sensitivity analysis

Hydrogeological assessment

Groundwater, also known as subterranean water, encompasses all water sources located beneath the Earth's surface (Humphreys, 2008). In addition to serving as a natural reservoir, it offers rural and urban towns a dependable and reasonably priced source of water. Beyond its importance in agriculture, groundwater plays a crucial role in industrial and domestic sectors (Mukate et al., 2020; Nepal et al., 2021; Ullah et al., 2022). Nearly half of all irrigation water comes from subterranean reservoirs, serving as the principal supply of water for approximately 2 billion people worldwide (Khilchevskyi and Karamushka, 2021). Its importance is impossible to overestimate in India, where groundwater supplies more than 85% of the country’s rural residential water needs, 50% of its urban water needs, and more than 50% of its irrigation needs (Paria et al., 2021).

Nevertheless, unsustainable behaviours, insufficient governance, swift urbanization, industrial growth, and a variety of human actions have worsened the worldwide water crisis, with the depletion of groundwater emerging as a notable issue (UN Water 2007; Mukherjee et al., 2020). In India, the decline in groundwater levels has become critically severe in recent times. Therefore, it is crucial to pinpoint potential groundwater areas to guarantee its prudent use while adhering to sustainable development objectives.

A complex interaction between geological, hydrological, atmospheric, and biological elements takes place in a variety of locales to generate groundwater. However, because of topographical, geological, and climatic differences, groundwater occurrence is exceedingly erratic in India (Moharir et al., 2023). Because groundwater is hidden and difficult to identify, it might be difficult to locate and explore in a given area. Although stratigraphic studies and drilling experiments are frequently employed techniques to ascertain the sites of borewells and the thickness of aquifers, their high prices and extended durations provide substantial challenges (Bertoni et al., 2020). Because the Geographical Information System (GIS) can handle large amounts of data accurately, integrating it with remotely sensed (RS) data has shown promise in tackling many global concerns, including groundwater issues (Benjmel et al., 2020). This technology has drawn interest due to its effectiveness, dependability, and time-saving features. Delineating groundwater potential zones has become easier and faster by integrating geological and hydrogeological factors with GIS technology (Saravanan et al., 2021).

Many techniques, such as multi-influencing factor (MIF), multicriteria decision making (MCDM), logistic regression, artificial neural network, fuzzy logic, analytical hierarchy process (AHP) (Chen et al., 2020; Doke et al., 2021; Gandhi and Patel, 2022; Mandal et al., 2021; Melese and Belay, 2022; Nguyen et al., 2020) have been used by researchers on integrated RS-GIS platforms to identify potential groundwater zones. Fuzzy-based geospatial techniques have also been helpful in tackling a number of global issues, including defining flood-prone areas, mapping the quality of groundwater, researching landslides, and locating artificial recharge locations (Choudhury et al., 2023; Dandapat et al., 2024; Dhaoui et al., 2022; Gopinath et al., 2023; Mansour et al., 2024; Mao et al., 2023; Melese and Belay, 2022; Nguyen, 2023; Qin et al., 2021; Şener, 2022; Shekar and Mathew, 2023).

The Surat district of Gujarat, India stands as a testament to significant population growth, urbanization, and industrial expansion, solidifying its status as the nation’s economic capital and a global centre for diamond, textile, and agricultural industries. With a current population estimated at 7,489,742, the district has witnessed rapid conversion of agricultural lands and water bodies into urban areas due to an influx of migrants from other Indian states seeking livelihoods, intensifying the demand for water resources. Despite agriculture covering over 70% of the district, exacerbating water demands, the primary water source, the Ukai dam on the Tapi River, grapples with challenges in meeting these needs. Moreover, the district faces a water crisis exacerbated by the depletion of groundwater resources, attributed to inadequate planning and management, as highlighted by the Central Groundwater Board of India’s report in 2013. For the purpose of mapping groundwater potential zones, the watershed of the Kakrapar Right Bank Canal (KRBC) has been identified along the Tapi River near Kakrapar village in the Mandvi taluka of Surat district which is and undergoing rapid development and industrializations. Constructed in 1954 primarily to facilitate irrigation, the Kakrapar dam covers a catchment area of 59904 km² and divides the irrigation supply into the KRBC and the Kakrapar Left Bank Main Canal (KLBC), each extending 64 km. Notably, the KRBC exhibits a discharge capacity of 70.23 m³/sec, encompassing a gross command area of 110220 ha and a cultivated command area of 58745 ha, achieving an annual irrigation rate of 162.16%, the highest among all main canals. This strategic location presents an opportunity to assess and delineate groundwater potential zones in conjunction with existing surface water infrastructure, aiding in sustainable water resource management and agricultural practices in the region. It should be noted that in recharging the groundwater in the dam’s command area, the KRBC does not play a direct role, as it is entirely concrete-lined.

The current research has undertaken an assessment of groundwater potential zones within the study area utilizing ten thematic layers to achieve higher accuracy. Notably, unlike many previous studies, particular attention has been given to conducting multicollinearity checks across various thematic layers, thereby enhancing the overall quality of the analysis. This process involves scrutinizing each thematic layer to identify and eliminate redundancy, thus refining the accuracy of the results. Despite the existing body of research, few have utilized GIS-based FOA techniques to delineate groundwater potential zones, particularly within the Agricultural KRBC region of Surat district, where such efforts have been notably absent. To better comprehend the impact of thematic layers, a map removal sensitivity study has been conducted. Furthermore, this study endeavours to assess the groundwater potential of the region through the combined application of GIS Fuzzy methodology, specifically Fuzzy Overlay Analysis (FOA) (Al-Rashidi et al., 2023) and the AHP, offering a comprehensive understanding of the area’s groundwater dynamics. In hydrological and hydrogeological research, validating findings is crucial. Typically, this involves comparing outputs with mathematical models or ground truth data, especially in groundwater studies. While previous literature has focused on methods like regional curvature or well yield for validation, this study introduces the use of a confusion matrix and kappa analysis to correlate existing groundwater data with fieldwork data and derive groundwater potential zone (GPZ) maps. By accurately identifying GPZs and assessing their correlation with sensitive criteria, this study has significant potential for groundwater management in the area, presenting a unique and innovative approach. The study aims to (1) map groundwater potential using the AHP and FOA techniques, (2) compare the outcomes of both methods and (3) validate the identified groundwater potential zones using kappa statistics and a confusion matrix. This research holds promise not only for local groundwater resource management, policy development, and mitigating groundwater scarcity in the region but also for similar areas facing analogous challenges.

Surat district (latitude 21° 30′ N to 21° 35′ N, longitude 72° 35′ E to 74° 20′ E), located in the southern part of Gujarat state, India and along the western coast of the Deccan Peninsula, encompasses diverse geography with hilly regions, piedmont slopes, alluvial plains, and coastal plains (Fig. 1). The district relies heavily on the perennial river Tapi as a primary water source, flowing westward towards the Arabian Sea. With a tropical savanna climate, Surat experiences distinct seasons, notably April and May as the hottest months with temperatures averaging 37 degrees Celsius, and a monsoon season witnessing an average rainfall of approximately 1143 mm. Agriculture covers over 70% of the area, predominantly cultivating sugarcane and paddy. The northern KRBC region benefits from the Kakrapar Right Bank Canal, constructed in 1954, boasting a gross command area of 2470 km² and a culturable command area of 1453 km² (Batliwala et al., 2014). This vital canal, spanning 64 km, irrigates crops such as rice, wheat, cotton, and sugarcane while also providing drinking water to villages and supporting industrial activities, thus playing a crucial role in the socio-economic development of the region. Additionally, crops like jowar, wheat, groundnut, banana, and vegetables contribute to the agricultural diversity of the district. The KRBC watershed area (study area) is around 2317 km².

[Fig. 1 near here]

3.1. Data acquisition

This study employs a variety of comprehensive remotely sensed datasets for analysis. The dataset compilation is carried out using ESRI ArcGIS 10.8 and Geomatica 2016 image processing software. Key parameters such as soil texture, geomorphology, and land use/land cover (LULC) are derived from IRS P6 LISS III imagery at a scale of 1:50000. Additionally, the Normalized Difference Wetness Index (NDWI) is computed using Level-1-Geo Tiff Data from Landsat 8 OLI/TIRS (Band 4 and 5), sourced from http://earthexplorer.usgs.gov. The ASTER Digital Elevation Model (DEM), obtained from the United State Geological Survey (USGS) portal (http://earthexplorer.usgs.gov), is utilized with a resolution of 30 meters. Further analysis involves the preparation of drainage networks and region slope using ArcGIS 10.6, leveraging the DEM data. Lineament mapping is conducted in Geomatica 2016 software, also utilizing DEM data. Rainfall data spanning 22 years (from 2001 to 2022) across 21 stations in the Surat district is gathered from the State Water Data Center (SWDC), Gandhinagar, and processed into thematic layers using ArcGIS 10.8. Additionally, the comprehensive methodology is illustrated in (Fig. 2.).

[Fig. 2 near here]

3.2. Parameter Optimization using Multicollinearity Assessment

In the context of a multiple regression model, multicollinearity arises when there is a significant correlation between at least one input parameter and a combination of other input parameters. This indicates that one input parameter, such as LULC or Soil, can be accurately predicted from another input parameter, leading to potential inaccuracies in the model's output. Therefore, it is crucial to assess multicollinearity among input parameters before conducting regression analysis. This assessment involves treating one input parameter as the dependent variable and the rest as independent variables, calculating correlation (R²) values for each parameter. These R² values are then used to determine tolerance Eq. (1) and variance inflation factor (VIF) values Eq. (2), aiding in the validation of multicollinearity (Asadi Shamsabadi et al., 2023).

Tolerance of p^th variable $\left({T}_{p}\right)=1-{R}_{p}^{2}$ (1)

VIF of p^th variable $\left(VI{F}_{p}\right)=1/{T}_{p}$ (2)

If a ${T}_{p}$ < 0.10 or VIF ≥ 10 indicates a multicollinearity problem (Venkatesh and Parimalarenganayaki, 2023), the same layers should be eliminated from the analysis.

To assess the presence of multicollinearity across different thematic layers, a sampling method was employed wherein 500 points were randomly chosen from the study area using the 'create random points' tool within ArcGIS 10.8 for each thematic layer. Subsequently, the values for each point were extracted utilizing the 'extract values to points' tool, assigning values based on the reclassified layer. This process yielded 500 values for each thematic layer, which were then analyzed using SPSS software.

3.3. AHP Method

AHP analysis in ArcGIS facilitates decision-making by systematically evaluating alternatives based on multiple criteria. Initially, relevant criteria are identified, encompassing factors crucial to the decision at hand (Cammerino et al., 2023). Through pairwise comparisons, criteria are assessed in relation to each other to ascertain their relative importance. The consistency of these comparisons is checked to ensure reliability (Tan et al., 2023). Once consistency is established, weights are calculated for each criterion, representing their significance in the decision-making process. These weights are determined by averaging the values of each criterion’s column in the pairwise comparison matrix. With criteria weights established, alternatives can be evaluated and ranked accordingly, using methods such as weighted overlays. Sensitivity analysis allows for an examination of how changes in criteria weights impact the ranking of alternatives, enhancing the understanding of decision robustness. AHP analysis in ArcGIS offers a transparent and structured approach, aiding in informed decision-making across diverse spatial applications (Kang et al., 2024).

3.4. Fuzzy Overlay Analysis (FOA)

FOA method in ArcGIS combines multiple layers of geographic data with uncertain boundaries using fuzzy logic. Fuzzy logic assigns membership values between 0 and 1 to features based on their degree of belongingness to specific classes (Nourani et al., 2023). Operations like intersection, union, or difference are performed on fuzzy sets, considering these membership values (Atagün and Kamacı, 2023).

The basic Eq. (3) for fuzzy overlay operations can be represented as:

$$\text{Output Membership }=f\left({\text{ Input Membership }}_{1},\text{ }{\text{ Input Membership }}_{2},\dots \right)$$

Here, $f$ represents the fuzzy overlay operation, and ${\text{ Input Membership }}_{1},\text{ }{\text{ Input Membership }}_{2}$ are the membership values from input layers.

FOA allows for the representation of uncertain or ambiguous spatial relationships, crucial in applications such as land suitability modelling, potential zone mapping, environmental assessment, and risk analysis. It enables analysts to account for uncertainties in spatial data, leading to more robust and realistic decision-making processes.

3.5. Groundwater Potential Zone Evaluation and Comparison

In this study, in the exploration of groundwater potential zones, two distinct methodologies have been employed: the AHP and FOA method. The delineation of groundwater potential zones often relies on the AHP method, a structured decision-making technique (Farhat et al., 2023). Beginning with the identification of key parameters, ten parameters were utilized and compared within this framework, the AHP method organizes these criteria hierarchically, assigning weights based on their perceived importance. Through pairwise comparisons and expert judgment, the relative priorities of these factors are determined, facilitating a comprehensive evaluation process (Grošelj et al., 2023). Integrated with GIS, weighted criteria are then utilized to generate groundwater potential maps, offering insights crucial for effective land use planning, water resource management, and environmental conservation efforts (Moharir et al., 2023). By systematically assessing influencing factors, the AHP method enhances our understanding of groundwater dynamics, enabling informed decision-making in diverse landscapes (Sulaiman and Mustafa, 2023). In recent years, the Fuzzy Integrated Geospatial Technique has emerged as a valuable tool in addressing various global challenges, including the identification of flood-prone areas, assessment of drinking water quality, analysis of sedimentation and soil erosion, prediction of air pollution, landslide hazard studies, and more (Aloui et al., 2024). To comprehend groundwater potentiality, researchers have employed diverse approaches. In this particular study, remote sensing data integration coupled with a fuzzy-logic-based approach was utilized to discern the impact of various hydrologic and hydrogeologic parameters on groundwater potentiality. Considering factors such as geomorphology, NDVI, rainfall distribution, existing groundwater levels, lineament density, drainage density, soil texture, LULC, relief, and slope, fuzzy membership values were assigned to each thematic layer. A crucial aspect of the fuzzy logic technique involves establishing criteria for delineating groundwater prospective zones. Groundwater experts, including professors, scientists, research supervisors, and field engineers of the study area, assessed the relative significance of each factor and its sub-classes, assigning appropriate ratings on a scale of 1 to 10. Subsequently, reclassed maps were assigned fuzzy logic membership values ranging from 0 to 1, representing varying degrees of groundwater occurrence likelihood. These fuzzy function-assigned layers were overlaid using the ArcGIS 10.8 overlay tool to elucidate groundwater prospective zones in the study area. Furthermore, a fuzzy GAMMA operator was employed in this study to integrate different layers of influencing parameters. The outcome of this integration was categorized into five classes: Very good, good, moderate, poor, and very poor, representing the groundwater prospective zones.

3.6. Investigation of Sensitivity and Influence of Different Factors on Potential Zones

In the current study, a sensitivity analysis utilizing the map exclusion method is employed to assess the significance of various parameters in determining groundwater potentiality as per their assigned ranking (Mathewos et al., 2024). It suggests which map or maps exert greater or lesser influence on the delineation of groundwater potential zones (Danso and Ma, 2023). This technique is applied for both methods used in the study. In this approach, one thematic layer is excluded at a time, followed by recalculating the groundwater potential zones. Eq. (4) is then employed to compute the highest, lowest, average, and standard deviation values for each map exclusion scenario.

$${S}_{i}=\left(\left|P/N-P{x}_{i}/n\right|\right)$$

Where, ${ S}_{i}$= sensitivity of the removed parameter (i),

$P$ = the groundwater potential index

$P{x}_{i}$ = the index value of groundwater potentiality estimated after removal of one factor (i), n= N-1 factors

3.7. Validation of a model

Validation serves as a crucial checkpoint in hydrological and hydrogeological inquiries, ensuring the integrity and reliability of study findings (Abd-Elmaboud et al., 2024). Although RS and GIS have ushered in a paradigm shift in understanding global issues, their ongoing development underscores the importance of robust validation procedures. In the realm of groundwater studies, validation typically entails comparing outputs against various mathematical models or ground truth data sources (Abhilash Singh et al., 2024). Numerous methodologies have been proposed in the literature for this purpose, including assessments based on region curvature, groundwater well yield, or existing groundwater level data. In the context of the study area, accuracy assessment involves scrutinizing the correlation between groundwater depth data and identified potential zones. Moreover, techniques such as confusion matrices or error assessments may be deployed to provide a comprehensive evaluation of accuracy levels. The determination of overall accuracy often involves applying specific equations, with (Ray, 2024) formulation being a notable example Eq. (5). By rigorously validating study outcomes, researchers can enhance confidence in the reliability and applicability of their findings, thereby contributing to informed decision-making processes in water resource management and environmental conservation efforts.

$$\text{Accuracy }=\frac{\text{ No. of correct well location}}{\text{ Total No. of well location}}$$

Additionally, the consistency between raters is often evaluated through Cohen's kappa statistic (k). The importance of rater reliability lies in its ability to assess the accuracy of the data collected in the study concerning the variables under investigation (Li et al., 2023). Kappa value (k) is defined by the following formula, which quantifies the agreement between raters beyond what would be expected by chance alone. This statistical measure is particularly valuable in research settings where multiple observers or coders are involved, as it provides insight into the reliability of their judgments (Ali et al., 2023). By assessing inter-rater dependability, researchers can gauge the consistency of their data collection process, thereby enhancing the overall validity and robustness of their findings. Kappa value (k) is defined by the following formula Eq. (6), which quantifies the agreement between raters beyond what would be expected by chance alone (Yilmaz and Demirhan, 2023),

$$\kappa \equiv \frac{{w}_{o}-{w}_{e}}{1-{w}_{e}}$$

Where, ${w}_{\text{o}}$ In Cohen's kappa statistic (k) represents the number of observed agreements between raters, while we denotes the number of agreements expected by hypothetical probability. These parameters play a crucial role in quantifying the level of agreement between raters, beyond what would be expected by chance alone. The classification of kappa values (Landis and Koch, 1977), offers a framework for interpreting the degree of agreement. These classifications range from no agreement (κ < 0) to perfect agreement (κ = 1.0), with intermediate categories including slight (0.00-0.20), fair (0.21–0.40), moderate (0.41–0.60), and substantial (0.61–0.80). By employing these classifications, researchers can assess the reliability of their data and draw meaningful conclusions about the consistency of judgments among raters.

4.1. Multicollinearity Assessment

The findings from the multicollinearity investigation are summarized in (Table 1). The analysis indicates that across all thematic layers, the Variance Inflation Factor (VIF) values remain below 10, while the tolerance values exceed 0.1. These results suggest the absence of collinearity among the ten selected thematic input layers utilized in the study. Consequently, the model outcomes remain unaffected by multicollinearity, ensuring the reliability and accuracy of the obtained results. This robustness in the model’s performance underscores the soundness of the methodology employed and enhances confidence in the validity of the study’s conclusions.

Table 1

Statistics on collinearity among factors influencing groundwater potential zones.
Thematic map	${\varvec{T}}_{\varvec{p}}$	$\varvec{V}\varvec{I}{\varvec{F}}_{\varvec{p}}$
LULC	0.631	1.58
Soil Texture	0.583	1.71
Geomorphology	0.758	1.32
Lithology	0.731	1.37
Rainfall	0.886	1.13
Drainage Density	0.781	1.28
Lineament Density	0.819	1.22
Slope	0.948	1.05
NDVI	0.785	1.27
DEM	0.533	1.87

[Table 1 near here]

4.2. Description of thematic layers used for mapping

[Fig. 3 near here]

4.2.1. Land Use Land Cover (LULC)

The concept of Land Use and Land Cover (LULC) is closely linked with groundwater management, providing reliable and precise data. The manner in which land is utilized affects processes such as infiltration and permeability, which in turn are influenced by human activities. Farmland, areas with abundant vegetation, and surface water sources facilitate groundwater infiltration, whereas built-up areas and wastelands impede it (Muduli et al., 2023). In the study area, LULC is categorized as follows: Waterbodies/wetland (3.01%), agricultural land (71.58%), forest land (9.04%), built-up area (6.62%), wastelands (5.75%) and other classifications (5.75%) (see Table 2). Agricultural land covers a significant portion of the watershed (1658.50 km²), distributed throughout except for the eastern dense forest area (Fig. 3a). Its prominence is due to the loose soils that enhance subsurface infiltration. Natural vegetation and scrub forests are prevalent in the northeast. Forest cover, known for its water storage capacity, ranks second after agricultural land. The reclassification map illustrates LULC fuzzy values ranging from 0.22 to 0.84 across the study area see (Fig. 4a).

Table 2

Reclassification of Factors with Rankings.
No.	Thematic layer	Influence	Individual feature Class	Scale value (Each have importance 0 to 5) Rating	Area in Km²	Area in %
1	LULC (km²)	12%	Agricultural Land	9	1658.50	71.58
			Built Up	2	83.91	3.62
			Forest	8	209.49	9.04
			Other Area	1	132.76	5.73
			Wastelands	4	133.11	5.75
			Waterbodies	9	69.70	3.01
			Wetlands	7	29.43	1.27
2	Soil Texture (km²)	5%	Clayey loam	5	660.30	28.53
			Fine clay	4	1497.43	64.69
			Very Fine clay	3	101.45	4.38
			Unclassified	0	55.55	2.40
3	Geomorphology (km²)	12%	Alluvial Plain	8	1258.95	54.33
			Coastal Plain	5	112.88	4.87
			Pediplain	7	587.02	25.33
			Plateau	6	188.93	8.15
			Water Bodies	0	67.67	2.92
			Other	0	101.32	4.36
4	Lithology (km²)	5%	Sandstone with Shale	7	238.16	10.28
			Basalt	6	537.79	23.21
			Gravel, sand and Silt	4	1371.84	59.21
			Built-up	0	94.40	4.07
			Canal	0	5.05	0.22
			Sea	0	6.92	0.30
			River	0	57.24	2.47
			Waterbodies	0	5.38	0.23
5	Rainfall (mm)	10%	1150	2	214.95	9.15
			1200	4	583.24	24.84
			1250	6	383.66	16.34
			1300	8	700.62	29.84
			1373	9	465.67	19.83
6	Drainage Density (km/km²)	15%	0-2.04	3	1429.32	61.55
			2.05–4.08	5	669.96	28.85
			4.09–6.12	7	170.29	7.33
			6.13–8.15	8	49.37	2.13
			8.16–10.20	9	3.32	0.14
7	Lineament Density (km/km²)	15%	0-0.11	3	460.23	19.58
			0.12–0.23	5	1222.55	52.02
			0.24–0.34	7	426.38	18.14
			0.35–0.46	8	124.19	5.28
			0.47–0.56	9	116.76	4.97
8	Slope (%)	8%	0–1	9	1029.22	44.02
			1–5	8	1036.74	44.34
			5–11	6	204.75	8.76
			11–22	4	55.45	2.37
			> 22	2	11.89	0.51
9	NDWI (unitless)	10%	(-0.501)- (-0.303)	1	7.22	0.31
			(-0.302)- (-0.106)	2	851.81	37.68
			(-0.105)- (0.0879)	2	1242.53	54.96
			(0.088)- (0.285)	7	158.96	7.03
10	Elevation (m)	8%	0–25	9	1892.28	80.55
			26–63	7	297.16	12.65
			64–110	5	100.55	4.28
			120–180	3	52.74	2.25
			190–310	1	6.44	0.27

4.2.2. Soil

Soil characteristics, including soil type and texture, exert a significant influence on the spatial distribution of groundwater (P. Wang et al., 2023). Groundwater recharge is affected by various factors such as runoff, water retention capacity, soil depth, porosity, among others. A thematic map depicting soil textures highlights the diverse range of soil types present in the district, categorized into three main groups: black clayey loam, fine clay, and very fine clay. In the higher elevation areas of the east, clayey loam predominates covering an area of 660.30 km². Fine clay, covering the majority of the watershed (64.69%), is widespread except in the east. Very fine clay is found in the central to western regions, comprising 4.38% of the total area (Fig. 3b). The permeability of clayey loam receives the highest rating of 5, whereas very fine clay is rated the lowest at 3 (see Table 2). The soil texture FM values for maximum and minimum soil textures are recorded at 0.22 and 0.83 respectively refer to(Fig. 4b). Additionally, the composition and texture of the soil can significantly impact groundwater dynamics. For instance, clayey soils tend to have lower permeability, which may affect groundwater recharge rates. Moreover, areas with finer soil textures might exhibit different rates of water infiltration compared to coarser soils. Understanding these soil properties is crucial for effective groundwater management and resource planning.

4.2.3. Geomorphology

The geological formation and topography of the area significantly influence groundwater distribution (H. Wang et al., 2023). Geomorphic mapping aids in identifying different landforms, geology, and topographical features, offering insights into the behavior of materials, methods, and geological factors affecting groundwater potentials, as emphasized by (Mosaad et al., 2024). The region is classified into six categories based on sub-classification, with each subcategory ranked according to its importance in groundwater occurrence. For instance, the alluvial plain, Pedi plain, and plateau rank higher compared to other geomorphological segments see Table 2). The alluvial plain covers the largest area (54.33%), followed by other subclasses such as Pedi plain, plateau, built-up areas/canals/rivers/water bodies, coastal plain, and structural hills (Fig. 3c). Higher FM values are concentrated in the central and western parts of the watershed, particularly in the alluvial plain areas, with values ranging from 0.29 to 0.91 refer to (Fig. 4c). Understanding these geomorphological features is crucial for assessing groundwater potential and implementing effective resource management strategies.

4.2.4. Lithology

The lithological makeup of a region is closely linked to its groundwater potentiality, as different rock types and sediment formations have varying capacities for storing and transmitting water (El Ayady et al., 2023). Understanding the distribution and characteristics of the lithological units is essential for assessing groundwater potential and developing effective management strategies within the watershed (Kumar Gautam et al., 2023). The lithological composition of the KRBC area watershed is crucial in understanding groundwater potentiality. The area is primarily characterized by three main lithological classifications: Sandstone with Shale covering 238.16 km² (10.28%), Basalt occupying 537.79 km² (23.21%), and Gravel, sand, and Silt encompassing 1371.84 km² (59.21%). Additionally, there are smaller areas of Built-up land (94.40 km² or 4.07%), Canal networks (5.05 km² or 0.22%), Coastal regions (6.92 km² or 0.30%), River courses (57.24 km² or 2.47%), and various Waterbodies (5.38 km² or 0.23%) (Fig. 3d). The Lithology FM values for maximum and minimum soil textures are recorded at 0.18 and 0.77 respectively (Fig. 4d).

4.2.5. Rainfall

Precipitation serves as the essential mechanism for replenishing surface water and recharging groundwater through weathered and fractured zones, as outlined by (Pandey et al., 2023). This relationship between rainfall and groundwater is bolstered by the facilitation of water percolation through runoff, elucidating a strong positive correlation between the two, as noted by (Nourani et al., 2024). Over the span of two decades, rainfall data has been systematically collected from the SWDC, Gandhinagar, and utilized to generate a geographical distribution map employing the Inverse Distance Weighted (IDW) tool within the ArcGIS platform. Annual rainfall in the district is categorized into five distinct classes, ranging from 1150 to 1373 mm/year, with varying extents represented in each class (Fig. 3e). Subsequently, sub-class ratings were assigned based on rainfall levels, where higher values denote a greater likelihood of water percolation and vice versa. In the study region, rainfall FM values spanned from 0.20 to 0.92, highlighting the variability in precipitation patterns across the area (Fig. 4e). Understanding these rainfall dynamics is crucial for evaluating groundwater recharge potential and implementing sustainable water management practices.

4.2.6. Drainage network density

The potential for groundwater occurrence is significantly impacted by water percolation and the lithological composition, which can be evaluated through the analysis of drainage network density (Abdullahi et al., 2023). Higher drainage network density implies a lower capacity for water infiltration, as demonstrated by the inverse relationship between drainage density and permeability (Sahoo et al., 2024). The area is categorized into five distinct groups based on drainage network density (Fig. 3f), ranging from 0-2.04 to 8.16–10.20 km/km² (Table 2). Fuzzy membership values, representing the degree of belongingness to each category, range from 0.16 to 0.84 see (Fig. 4f). The largest area, approximately 1429 km², exhibits a drainage network density of 0-2.04 km/km², whereas the highest categories, exceeding 8.16 km/km², cover the least area, totalling 3.32 km². Regions with drainage density greater than 8.16 km/km² are predominantly concentrated in the central portion, with few patches scattered across other areas. Understanding the relationship between drainage network density and groundwater potentiality is essential for effective water resource management and planning.

4.2.7. Lineament density

Another significant hydrological factor influencing groundwater potential is lineament density, particularly evident in regions with hard rock terrain where linear features and fractures indicate favourable conditions for groundwater storage (Kamaraj et al., 2023). Lineament density typically varies between 0 and 0.56 km/km². In our district, lineament density is categorized into five classes, with rankings assigned based on ascending lineament density values, ranging from 0-0.11 to 0.47–0.56 km/km² (Table 2). The majority of the area, encompassing 60.15%, falls within the 0.12–0.23 km/km² class (Fig. 3g). Fuzzy membership (FM) values associated with the lineament density map range from 0.07 to 0.74, indicating the degree of membership to each density class see (Fig. 4g). It’s important to note that areas with higher lineament density may present better groundwater storage potential, highlighting the relevance of lineament analysis in groundwater resource assessment and management planning.

[Fig. 3 near here]

[Fig. 4 near here]

4.2.8. Slope

Slope gradient, often overlooked in groundwater studies, plays a crucial role, especially in regions with relatively flat terrain (Guyo et al., 2024). The inclination of the land surface directly affects rainfall infiltration, with steeper slopes leading to rapid surface runoff, limiting water penetration and recharge to the saturated zone (Yu et al., 2023). Categorizing the slope into five subclasses based on percentage facilitates understanding, ranging from very gentle (0–1%) to very steep (> 22%) (Table 2). These classifications are ranked according to their water percolation capacities, with very gentle slopes receiving the highest rank (9) and very steep slopes the lowest (2). Notably, in the Surat district, characterized by predominantly flat terrain, the majority of the area (50.51%) falls under the very gentle slope category (Fig. 3h). However, in the northeastern region, adjacent to mountainous areas such as Dang and Tapi districts, very steep slopes are observed, influencing groundwater dynamics differently. The slope's fuzzy membership (FM) values range from 0.35 to 0.85 across the district see (Fig. 4h), underscoring the varying degrees of slope influence on groundwater recharge and flow patterns. Understanding these slope characteristics is essential for comprehensive groundwater management and sustainable resource utilization strategies.

[Table 2 near here]

4.2.9. NDWI

Normalized Difference Water Index (NDWI) values, ranging from − 1 to 1, serve as critical indicators of water presence in remote sensing data (Chandramohan et al., 2024). Higher NDWI values typically indicate the presence of water bodies, while lower values signify non-water features (Hu et al., 2024). In the study area, NDWI values range from − 0.501 to 0.285 (Fig. 3i), with higher NDWI areas assigned higher scores gradually decreasing to 1. Correlating NDWI values with groundwater potential zones provides valuable insights into areas conducive to groundwater recharge, facilitating comprehensive water resource management strategies (Yi et al., 2023). By integrating NDWI analysis with groundwater potential assessments, suitable locations for groundwater extraction can be identified, recharge measures can be implemented, and water-related risks can be mitigated, thereby fostering sustainable management of both surface and subsurface water resources. Fuzzy values, ranging from 0.05 to 0.83, further refine this assessment process (Fig. 4i).

4.2.10. Elevation

The elevation within the study area is divided into five subclasses, ranging from 0–25 m to 190–310 m, as indicated in (Table 2). Approximately 80.55% of the total area falls within the subclass category of 0–25 m, indicating a predominantly low-lying terrain. Conversely, the higher elevations, covering 364 km², are concentrated in the northeast portion of the district. Areas with lower elevations typically exhibit favourable conditions for groundwater potential due to extended water retention periods, while higher elevation regions tend to have lower occurrences of groundwater. Additionally, water tends to flow from higher relief areas to lower relief areas (Fig. 3j). Consequently, the regions falling under elevations of less than 25 m are assigned the highest ranking, with other elevation categories depicted accordingly in (Table 2). The spatial distribution map of the Digital Elevation Model (DEM) fuzzy membership within the district reveals a range of values between 0.04 and 0.97 (Fig. 4j). This fuzzy membership assessment provides additional insight into the varying degrees of elevation influence on groundwater potentiality across the study area.

4.3. Groundwater Potential zones mapping

In this section, the discussion centres on the application of the AHP and FOA method for groundwater potential zone mapping. A careful selection of 10 factors was made to comprehensively evaluate groundwater potential. Through the integration of AHP and FOA method, these factors underwent thorough evaluation to delineate the spatial distribution of groundwater potential zones. The AHP method facilitated the assessment of the relative significance of different thematic layers and their respective groups influencing groundwater, as depicted in (Table 2) illustrating the weights assigned to each thematic layer. Consistency Index (CI) and Consistency Ratio (CR) values, calculated as CI = 0.08 and CR = 0.06 respectively (Table 3), fell below the threshold of 0.1, indicating a high level of consistency in the analysis. Subsequently, rankings assigned to sub-criteria were integrated into the final Weight of Importance (WOI) for mapping Groundwater Potential Zones (GPZs). The resulting classification into five categories (very poor, poor, moderate, good, and very good) offers a comprehensive understanding of groundwater potential across the study area (Fig. 5), facilitating informed decision-making and resource management strategies.

Table 3

Pairwise comparison matrix.
Thematic map	LULC	ST	GM	LI	R	DD	LD	SL	NDWI	EL
LULC	1.00	0.33	1.00	2.00	3.00	1.00	2.00	0.33	3.00	1.00
ST	3.00	1.00	1.00	3.00	5.00	1.00	7.00	0.50	8.00	1.00
GM	1.00	1.00	1.00	4.00	2.00	1.00	2.00	0.33	3.00	0.33
LI	0.50	0.33	0.25	1.00	0.50	0.50	0.50	0.20	2.00	0.17
R	0.33	0.20	0.50	2.00	1.00	0.20	2.00	0.14	2.00	0.13
DD	1.00	1.00	1.00	2.00	5.00	1.00	4.00	2.00	6.00	1.00
LD	0.50	0.14	0.50	2.00	0.50	0.25	1.00	0.11	2.00	0.25
SL	3.00	2.00	3.00	5.00	7.00	0.50	9.00	1.00	6.00	0.50
NDWI	0.33	0.13	0.33	0.50	0.50	0.17	0.50	0.17	1.00	0.17
EL	1.00	1.00	3.00	6.00	8.00	1.00	4.00	2.00	6.00	1.00
Maximum Eigen value (λ max) = 10.8.
n = 10.
Consistency index (CI) = (λ max - n)/(n − 1) = 0.08.
Random index (RI) = 1.33
Consistency ratio (CR) = CI/RI = 0.06.

[Table 3 near here]

To evaluate groundwater potentiality in the KRBC watershed using the FOA method, ten influencing criteria were employed. Each criterion was subclassified and ranked based on its significance in determining groundwater potentiality. Subsequently, reclassified layers were assigned Fuzzy Membership (FM) function values ranging from 0 to 1. Ten FM layers were then integrated using the "Fuzzy Overlay" function within the ArcGIS 10.6 environment. The Overlay operation was processed using the GAMMA operator with a value of 0.9, effectively combining AND, OR, algebraic product, and algebraic sum functionalities. The resulting groundwater potential zone map ranged from 0.20 to 0.83, where lower values indicated lower groundwater potentiality and vice versa. This methodology provides a robust framework for assessing groundwater potential, enhancing understanding and aiding in decision-making processes related to water resource management.

According to the computed values in (Table 4), the entire study area is divided into five sub-categories viz; very good, good, Moderate, Poor, and very poor. The highest area covered by moderate potential zone by both methods i.e., AHP (54.20%) and FOA (38.06%) The comparison of groundwater potential zones between the AHP and FOA reveals notable differences in the allocation of area for each zone category. While AHP tends to designate smaller areas to categories like Very Poor and Very Good, the FOA assigns larger areas to these categories. For instance, the Very Poor zone covers 15.00 km² (0.63%) according to AHP, compared to 419.5 km² (17.59%) as per the FOA. Similarly, the Moderate zone encompasses 1293.02 km² (54.20%) in AHP, while it occupies 908 km² (38.06%) in the FOA. These disparities suggest variations in methodologies and criteria, warranting further examination to ascertain the accuracy and reliability of groundwater potential mapping results and their implications for management decisions.

Table 4

Comparison of Groundwater Potential Zones.
Class	AHP		FOA
Class	Area in Km²	Area in %	Area in Km²	Area in %
Very Poor	15.00	0.63	419.5	17.59
Poor	231.58	9.71	210	8.80
Moderate	1293.02	54.20	908	38.06
Good	819.59	34.36	420	17.61
Very Good	26.34	1.10	428	17.94

[Table 4 near here]

[Fig. 5 near here]

4.4. Exploring Sensitivity: The Feature Elimination Approach

A sensitivity study was conducted to assess the significance of influencing parameters on groundwater potentiality within the KRBC Watershed. This study involved systematically removing each layer one by one from the groundwater potential zone map in the ArcGIS platform, allowing for the evaluation of the impact of individual parameters. The results of the sensitivity study, presented in (Table 5), revealed higher standard deviation (SD) values associated with certain layers, indicating their greater sensitivity in assessing groundwater potential zones. According to the AHP method, the layers that exhibited higher SD values upon removal were NDWI (0.2262), Lineament density (0.1692), and LULC (0.1267). Conversely, the FOA method indicated higher SD values for Soil Texture (0.0773), Lineament Density (0.0515), and NDWI (0.0466). This suggests that these specific criteria are particularly sensitive in assessing groundwater potential within the watershed. Figures 6 and 7 depict the potential zones resulting from the removal of each layer using the AHP and FOA method, respectively. These figures provide visual representations of the sensitivity analysis conducted to assess the impact of individual layers on groundwater potentiality within the study area. By systematically removing each layer and observing the resultant changes in potential zones, insights into the relative importance of different parameters in groundwater potential assessment are gained. These visualizations serve as valuable tools for understanding the dynamics of groundwater potential and guiding decision-making processes related to water resource management strategies.

Table 5

Sensitivity Analysis- Map Elimination Method.
Removed Thematic layer	Standard Deviation		Standard deviation difference in potential zones after removing thematic layer
Removed Thematic layer	AHP	FOA	AHP	FOA
LULC	0.5872	0.2644	0.1267	0.0088
Soil	0.7419	0.3329	0.0287	0.0773
Geomorphology	0.676	0.2658	0.0379	0.0102
Lithology	0.7347	0.276	0.0208	0.0204
Rainfall	0.6741	0.2932	0.0398	0.0376
Drainage Density	0.6523	0.2895	0.0616	0.0339
Lineament Density	0.8831	0.3071	0.1692	0.0515
Slope	0.676	0.2674	0.0379	0.0118
NDWI	0.9401	0.3022	0.2262	0.0466
Elevation	0.7608	0.2662	0.0469	0.0106

[Fig. 6 near here]

[Fig. 7 near here]

The heightened sensitivity of NDWI in both methods underscores the significance of water body dynamics in influencing groundwater potential. Similarly, the sensitivity of Lineament density highlights the importance of geological structures in groundwater flow patterns. The varying sensitivity observed between methods and parameters underscores the need for a comprehensive understanding of the hydrogeological context when assessing groundwater potential. Moreover, the identification of less sensitive layers underscores the relative stability of certain parameters in groundwater potential assessment. This information can guide resource allocation and management strategies by prioritizing interventions in areas where influential parameters exhibit higher sensitivity. Hence, the sensitivity study provides valuable insights into the relative importance of influencing parameters in groundwater potential assessment, facilitating more informed decision-making processes for sustainable water resource management within the KRBC Watershed.

[Table 5 near here]

4.5. Validation of Groundwater potential zone map

In this study, a comprehensive approach was adopted, integrating fieldwork data with existing groundwater depth records to validate findings regarding groundwater potential zones. A total of 49 well-points were meticulously surveyed, with approximately 30 well points measured manually and the remaining collected from the local groundwater authority, ensuring a comprehensive dataset for analysis. Each well point was documented with its corresponding latitude and longitude coordinates. Emphasizing pre-monsoon periods for data collection allowed capturing groundwater conditions during challenging periods. The recorded well depths in the KRBC area during April through June ranged notably, from shallow depths of 2 m to depths as deep as 25 m. To effectively interpret these variations, wells were categorized into distinct groups based on observed water depths, ranging from very good to very poor conditions.

To assess the reliability of findings, two methodologies were employed: the AHP and the FOA method. Analysis of results using an error matrix revealed fair agreement (AHP) and substantial agreement (FOA) between identified groundwater potential zones and observed well data. The AHP method demonstrated fair agreement, while the FOA method showed substantial agreement, with overall accuracy rates of 59.18% and 77.5%, respectively. These findings were reinforced by Cohen’s kappa values of 0.336 (fair agreement) and 0.625 (substantial agreement), respectively, affirming the robustness and reliability of the study's conclusions across both methodologies.

[Table 6 near here]

Table 6

Confusion Matrix for Potential zone validation.
	Category		No. of Groundwater well depth observed						Overall Accuracy	Kappa Coefficient
	Category		Very Good (0–3 m)	Good (3–6 m)	Moderate (6–9 m)	Poor (9–12 m)	Very poor (> 12 m)	Total
Groundwater Potential Zones	AHP Method	Very Good	1	1	1	2	0	5	59.18%	0.336 (fair agreement)
		Good	0	7	2	1	0	10
		Moderate	2	0	9	2	1	14
		Poor	0	1	2	9	0	12
		Very poor	0	0	1	4	3	8
		Total	3	9	15	18	4	49
	FOA Method	Very Good	3	1	0	0	0	4	77.55%	0.625 (substantial agreement)
		Good	0	8	2	0	0	10
		Moderate	4	0	8	1	0	13
		Poor	0	0	1	13	0	14
		Very poor	0	0	0	2	6	8
		Total	7	9	11	16	6	49

Furthermore, Table 6 presents the Confusion Matrix for potential zone validation, providing a systematic assessment of the accuracy of potential zone delineation by comparing predicted potential zones against observed conditions. Such validation measures play a critical role in ensuring the accuracy and robustness of groundwater potential assessments, thereby enhancing the credibility of decision-making processes in water resource management.

4.6. Discussion

Several studies have underscored the AHP method as the preferred approach for delineating groundwater potential zones due to its effectiveness in decision-making under uncertainty and hierarchical structuring of criteria (Arumugam et al., 2023; Ikirri et al., 2023; Muavhi and Mutoti, 2023; Atar Singh et al., 2024; Upwanshi et al., 2023; Warghat et al., 2023). However, the success of the FOA method in the KRBC region suggests an alternative perspective. The GIS integrated FOA method offers a versatile approach applicable across various domains, such as groundwater potential zone mapping (Jesiya and Gopinath, 2020; Singha et al., 2021), identification of recharge site (Raja Shekar and Mathew, 2023), assessment of land susceptibility (Nwazelibe et al., 2023), and analysis of drinking water quality (Mallik et al., 2021; Tiwari et al., 2024), among other fields. This method, proven through empirical validation and theoretical substantiation, stands as a robust technique for spatial analysis and decision-making processes. Its inherent flexibility and capability to accommodate diverse datasets and parameters render it indispensable for addressing complex geospatial challenges across various sectors. The FOA method has demonstrated notable effectiveness in this specific context, prompting a re-evaluation of its merits. Its adaptability to complex decision-making scenarios and robust performance in delineating groundwater potential zones merit recognition. Additionally, regional characteristics such as geological diversity and hydrological dynamics profoundly influence the suitability and performance of different methodologies. Therefore, comprehensive sensitivity analysis and robustness testing are essential when applying these methods in diverse hydrogeological settings.

Within the region, groundwater potential is influenced by various factors, including lithology, land use land cover, topography, and hydrological parameters. While water bodies act as crucial recharge sources, high drainage density can increase surface runoff, impeding infiltration rates and affecting groundwater recharge. Elevated terrain hinders water percolation due to increased gravitational potential energy, whereas agricultural land in the central watershed enhances groundwater potential through enhanced recharge processes and reduced surface runoff. Flat terrain and gentle slopes promote groundwater storage, characterizing areas with Good to Very Good potential zones due to favourable conditions for groundwater infiltration and storage.

The concentration of alluvial and Pedi plains in the central region in the study area suggests the highest likelihood of elevated groundwater potential, owing to their high hydraulic conductivity and significant groundwater storage capacities (Fig. 3c and Fig. 4c). Lineament density, varying from 0.12 to 0.23 km/km², correlates with increased potential, particularly in the Good and Moderate zones, indicating the importance of fracture networks and structural controls on groundwater flow and storage (Fig. 3g and Fig. 4g). Additionally, areas rich in clayey loam and fine clay exhibit higher potential due to their low permeability and high water retention capacities (Fig. 3b and Fig. 4b), emphasizing the importance of soil composition in groundwater dynamics and recharge processes.

This comprehensive understanding underscores the complex interaction of environmental factors in shaping groundwater potential zones, emphasizing the need for advanced hydrogeological modelling techniques and integrated management approaches to ensure sustainable water resource utilization in the study area and similar hydrogeological settings.

The identification of zones with high groundwater potential is of paramount importance for the sustainable management of groundwater resources. In this study, the AHP and FOA approaches, integrated with GIS-RS, were utilized to model groundwater potentiality in the KRBC watershed area. Prior to analysis, multicollinearity tests were conducted on ten thematic layers, revealing no collinearity between layers (tolerance value > 0.1). Thematic layers of influencing parameters were prepared using GIS software. The study categorizes the area into five groundwater potential zones, with the moderate zone covering the largest proportion: 54.20% according to AHP and 38.06% with FOA. Notably, the Very Poor zone encompasses 15.00 km² (0.63%) in AHP, compared to 419.5 km² (17.59%) with FOA. Similarly, the Moderate zone occupies 1293.02 km² (54.20%) in AHP, contrasting with 908 km² (38.06%) in the FOA method. These differences highlight the need for further examination to validate mapping results and inform management decisions. The sensitivity study revealed higher standard deviation (SD) values for specific layers, indicating their increased sensitivity in assessing groundwater potential zones. According to the AHP method, the layers showing higher SD values upon removal were NDWI, Lineament density, and Land Use Land Cover (LULC). Conversely, the FOA method identified Soil Texture, Lineament Density, and NDWI as the layers with the highest SD values. To validate the results, a comprehensive approach was taken in this study, integrating fieldwork data with existing groundwater depth records. A total of 49 well-points were surveyed. Data collection focused on pre-monsoon periods, capturing depths ranging from 2m to 25m in the KRBC area. The reliability of findings was assessed using both the AHP and FOA, demonstrating fair and substantial agreement, respectively, with overall accuracy rates of 59.18% and 77.5%. These results were further supported by Cohen’s kappa values, affirming the study's robustness across both methodologies.

In the study area, the FOA method surpassed the AHP method in assessing groundwater potential due to its superior ability to handle complexity and uncertainty. By incorporating fuzzy logic and accommodating vague information, the FOA method provided a more precise representation of groundwater dynamics, integrating multiple data layers and capturing spatial relationships effectively. Its adaptability and nuanced analysis make it a preferred choice for groundwater assessment in regions with heterogeneous hydrogeological conditions. This study underscores the efficacy of fuzzy logic in defining groundwater potential zones, offering a methodology that can be applied beyond the study area with appropriate adjustments. Additionally, effective policymaking for groundwater management necessitates collaborative efforts among policymakers, researchers, and local communities to ensure sustainable practices and resource utilization.

Despite the extensive scientific analysis conducted in this study, there are limitations that need addressing. The thematic layers employed for groundwater potential assessment may necessitate the incorporation of additional data layers, tailored to site-specific conditions, to optimize the accuracy of delineating local groundwater potential zones. Adjusting and incorporating more layers relevant to the local hydrogeological context can enhance the precision and reliability of the model's predictions. Furthermore, the incorporation of geophysical surveys for verification purposes in future studies would further enhance the robustness and applicability of the FOA model to diverse geographical regions.

Funding

The authors extend their appreciation to the Researchers Supporting Project number (RSPD2024R848), King Saud University, Riyadh, Saudi Arabia.

Conflict of Interest

There is no conflict of interest for this article

Author Contribution: Fenil R. Gandhi: Analysis, Writing, Figure Preparation, Table Creation;

Jaysukh C Songara: Conceptualization, Supervision, Analysis, Writing; Indra Prakash: Drafting,

Editing, Reviewing; Hamad Ahmed Altuwaijri: Writing; Discussion, Editing.

Consent to participate: All of the authors consent to participate in this research work.

Consent for publication: All of the authors consent to publish this work.

Ethical approval: We declare herein that our paper is original and unpublished elsewhere and that this manuscript complies to the Ethical Rules applicable for this journal

Abd-Elmaboud ME, Saqr AM, El-Rawy M, Al-Arifi N, Ezzeldin R (2024) Evaluation of groundwater potential using ANN-based mountain gazelle optimization: A framework to achieve SDGs in East El Oweinat. Egypt J Hydrol Reg Stud 52:101703. https://doi.org/10.1016/j.ejrh.2024.101703
Abdullahi A, Jothimani M, Getahun E, Gunalan J, Abebe A (2023) Assessment of potential groundwater Zones in the drought-prone Harawa catchment, Somali region, eastern Ethiopia using geospatial and AHP techniques. Egypt J Remote Sens Sp Sci 26:628–641. https://doi.org/10.1016/j.ejrs.2023.07.005
Al-Rashidi A, Sabarathinam C, Samayamanthula DR, Alsabti B, Rashid T (2023) Groundwater Management for Agricultural Purposes Using Fuzzy Logic Technique in an Arid Region. Water 15:2674. https://doi.org/10.3390/w15142674
Ali R, Hussain A, Nazir S, Khan S, Khan HU (2023) Intelligent Decision Support Systems—An Analysis of Machine Learning and Multicriteria Decision-Making Methods. Appl Sci 13:12426. https://doi.org/10.3390/app132212426
Aloui S, Zghibi A, Mazzoni A, Elomri A, Al-Ansari T (2024) Identifying suitable zones for integrated aquifer recharge and flood control in arid Qatar using GIS-based multi-criteria decision-making. Groundw Sustain Dev 25:101137. https://doi.org/10.1016/j.gsd.2024.101137
Arumugam M, Kulandaisamy P, Karthikeyan S, Thangaraj K, Senapathi V, Chung SY, Muthuramalingam S, Rajendran M, Sugumaran S, Manimuthu S (2023) An Assessment of Geospatial Analysis Combined with AHP Techniques to Identify Groundwater Potential Zones in the Pudukkottai District, Tamil Nadu, India. Water 15:1101. https://doi.org/10.3390/w15061101
Asadi Shamsabadi E, Salehpour M, Zandifaez P, Dias-da-Costa D (2023) Data-driven multicollinearity-aware multi-objective optimisation of green concrete mixes. J Clean Prod 390:136103. https://doi.org/10.1016/j.jclepro.2023.136103
Atagün AO, Kamacı H (2023) Strait fuzzy sets, strait fuzzy rough sets and their similarity measures-based decision making systems. Int J Syst Sci 54:2519–2535. https://doi.org/10.1080/00207721.2023.2233971
Batliwala B, Patel J, Porey P (2014) Modernization of Kakrapar Right Bank Main Canal. IOSR J Mech Civ Eng 11:01–04. https://doi.org/10.9790/1684-11450104
Benjmel K, Amraoui F, Boutaleb S, Ouchchen M, Tahiri A, Touab A (2020) Mapping of Groundwater Potential Zones in Crystalline Terrain Using Remote Sensing, GIS Techniques, and Multicriteria Data Analysis (Case of the Ighrem Region, Western Anti-Atlas, Morocco). Water 12:471. https://doi.org/10.3390/w12020471
Bertoni C, Lofi J, Micallef A, Moe H (2020) Seismic Reflection Methods in Offshore Groundwater Research. Geosciences 10:299. https://doi.org/10.3390/geosciences10080299
Cammerino ARB, Ingaramo M, Piacquadio L, Monteleone M (2023) Assessing and Mapping Forest Functions through a GIS-Based, Multi-Criteria Approach as a Participative Planning Tool: An Application Analysis. Forests 14:934. https://doi.org/10.3390/f14050934
Chandramohan K, Elayapillai P, Vijayalakshmi G, Kaliraj S (2024) Evaluating the relation of NDVI, NDWI, SMI, and LAI to land and soil degradation processes — a case study of Virudhunagar district, Tamil Nadu, India. Remote Sensing of Soils. Elsevier, pp 689–697. https://doi.org/10.1016/B978-0-443-18773-5.00040-5
Chen W, Li Y, Tsangaratos P, Shahabi H, Ilia I, Xue W, Bian H (2020) Groundwater Spring Potential Mapping Using Artificial Intelligence Approach Based on Kernel Logistic Regression, Random Forest, and Alternating Decision Tree Models. Appl Sci 10:425. https://doi.org/10.3390/app10020425
Choudhury A, Mahammood V, Rao KHVD (2023) Ground water potential zone assessment using geospatial technique and identification of artificial recharge sites: a case study of Mahabubnagar District of Telangana. Model Earth Syst Environ 9:1709–1723. https://doi.org/10.1007/s40808-022-01575-4
Dandapat K, Chatterjee U, Das S, Patra A, Manna H, Ghosh A, Pal SC, Islam T, Costache ARM, Alam R, Islam E, M.K (2024) Assessment of groundwater potential zones in data-scarce regions using GIS-based multicriteria decision making approach. Geocarto Int 39. https://doi.org/10.1080/10106049.2024.2311202
Danso SY, Ma Y (2023) Geospatial techniques for groundwater potential zones delineation in a coastal municipality. Ghana Egypt J Remote Sens Sp Sci 26:75–84. https://doi.org/10.1016/j.ejrs.2022.12.004
Dhaoui O, Agoubi B, Antunes IM, Tlig L, Kharroubi A (2022) Groundwater quality for irrigation in an arid region—application of fuzzy logic techniques. Environ Sci Pollut Res 30:29773–29789. https://doi.org/10.1007/s11356-022-24334-5
Doke AB, Zolekar RB, Patel H, Das S (2021) Geospatial mapping of groundwater potential zones using multi-criteria decision-making AHP approach in a hardrock basaltic terrain in India. Ecol Indic 127:107685. https://doi.org/10.1016/j.ecolind.2021.107685
El Ayady H, Mickus KL, Boutaleb S, Morjani E, Ikirri ZEA, Echogdali M, Bessa FZE, Abdelrahman AZ, Id-Belqas K, Essoussi M, Fnais S, Abioui MS, M (2023) Investigation of groundwater potential using geomatics and geophysical methods: Case study of the Anzi sub-basin, western Anti-Atlas, Morocco. Adv Sp Res 72:3960–3981. https://doi.org/10.1016/j.asr.2023.08.044
Farhat B, Souissi D, Mahfoudhi R, Chrigui R, Sebei A, Ben Mammou A (2023) Environ Monit Assess 195:285. https://doi.org/10.1007/s10661-022-10845-8. GIS-based multi-criteria decision-making techniques and analytical hierarchical process for delineation of groundwater potential
Gandhi FR, Patel JN (2022) Groundwater potentiality deciphering and sensitivity study using remote sensing technique and fuzzy approach. Acta Geophys 70:265–282. https://doi.org/10.1007/s11600-021-00711-5
Gopinath G, Jesiya N, Achu AL, Bhadran A, Surendran UP (2023) Ensemble of fuzzy-analytical hierarchy process in landslide susceptibility modeling from a humid tropical region of Western Ghats, Southern India. Environ Sci Pollut Res. https://doi.org/10.1007/s11356-023-27377-4
Grošelj P, Zandebasiri M, Pezdevšek Malovrh Š (2023) Evaluation of the European experts on the application of the AHP method in sustainable forest management. Environ Dev Sustain. https://doi.org/10.1007/s10668-023-03859-w
Guyo RH, Wang K, Saito M, Onodera S, Shimizu Y, Moroizumi T (2024) Spatiotemporal shallow and deep groundwater dynamics in a forested mountain catchment with diverse slope gradients, western Japan. Groundw Sustain Dev 25:101150. https://doi.org/10.1016/j.gsd.2024.101150
Hu Y, Lee J, Paik K (2024) Combining topography and reflectance indices for better surface water detection. J Hydro-environment Res 52:38–49. https://doi.org/10.1016/j.jher.2024.01.001
Humphreys WF (2008) Rising from Down Under: developments in subterranean biodiversity in Australia from a groundwater fauna perspective. Invertebr Syst 22:85. https://doi.org/10.1071/IS07016
Ikirri M, Boutaleb S, Ibraheem IM, Abioui M, Echogdali FZ, Abdelrahman K, Id-Belqas M, Abu-Alam T, Ayady E, Essoussi H, Faik S, F (2023) Delineation of Groundwater Potential Area using an AHP, Remote Sensing, and GIS Techniques in the Ifni Basin, Western Anti-Atlas, Morocco. Water 15:1436. https://doi.org/10.3390/w15071436
Jesiya NP, Gopinath G (2020) A fuzzy based MCDM–GIS framework to evaluate groundwater potential index for sustainable groundwater management - A case study in an urban-periurban ensemble, southern India. Groundw Sustain Dev 11:100466. https://doi.org/10.1016/j.gsd.2020.100466
Kamaraj P, Jothimani M, Panda B, Sabarathinam C (2023) Mapping of groundwater potential zones by integrating remote sensing, geophysics, GIS, and AHP in a hard rock terrain. Urban Clim 51:101610. https://doi.org/10.1016/j.uclim.2023.101610
Kang YO, Yabar H, Mizunoya T, Higano Y (2024) Optimal landfill site selection using ArcGIS Multi-Criteria Decision-Making (MCDM) and Analytic Hierarchy Process (AHP) for Kinshasa City. Environ Challenges 14:100826. https://doi.org/10.1016/j.envc.2023.100826
Khilchevskyi V, Karamushka V (2021) Global Water Resources: Distribution and Demand. pp. 1–11. https://doi.org/10.1007/978-3-319-70061-8_101-1
Kumar Gautam V, Pande CB, Kothari M, Kumar Singh P, Agrawal A (2023) Exploration of groundwater potential zones mapping for hard rock region in the Jakham river basin using geospatial techniques and aquifer parameters. Adv Sp Res 71:2892–2908. https://doi.org/10.1016/j.asr.2022.11.022
Landis JR, Koch GG (1977) An Application of Hierarchical Kappa-type Statistics in the Assessment of Majority Agreement among Multiple Observers. Biometrics 33:363. https://doi.org/10.2307/2529786
Li M, Gao Q, Yu T (2023) Using appropriate Kappa statistic in evaluating inter-rater reliability. Short communication on Groundwater vulnerability and contamination risk mapping of semi-arid Totko river basin, India using GIS-based DRASTIC model and AHP techniques. Chemosphere 328:138565. https://doi.org/10.1016/j.chemosphere.2023.138565
Mallik S, Mishra U, Paul N (2021) Groundwater suitability analysis for drinking using GIS based fuzzy logic. Ecol Indic 121:107179. https://doi.org/10.1016/j.ecolind.2020.107179
Mandal KK, Ranjan A, Dharanirajan K (2021) Delineation of Groundwater Potential Zones (GWPZ) of Port Blair, Andaman Islands, India, using Multi Influencing Factors (MIF) method and geospatial techniques. Remote Sens Appl Soc Environ 24:100631. https://doi.org/10.1016/j.rsase.2021.100631
Mansour A, Mrad D, Djebbar Y (2024) Advanced modeling for flash flood susceptibility mapping using remote sensing and GIS techniques: a case study in Northeast Algeria. Environ Earth Sci 83:60. https://doi.org/10.1007/s12665-023-11324-0
Mao Y, Chen L, Nanehkaran YA, Azarafza M, Derakhshani R (2023) Fuzzy-Based Intelligent Model for Rapid Rock Slope Stability Analysis Using Qslope. Water 15:2949. https://doi.org/10.3390/w15162949
Mathewos Y, Abate B, Dadi M, Mathewos M (2024) Evaluation of the groundwater prospective zone by coupling hydro-meteorological and geospatial evidence in Wabe River Catchment Omo Gibe River Basin, Ethiopia. Water Cycle 5:37–58. https://doi.org/10.1016/j.watcyc.2024.01.002
Melese T, Belay T (2022) Groundwater Potential Zone Mapping Using Analytical Hierarchy Process and GIS in Muga Watershed, Abay Basin, Ethiopia. Glob. Challenges 6. https://doi.org/10.1002/gch2.202100068
Moharir KN, Pande CB, Gautam VK, Singh SK, Rane NL (2023) Integration of hydrogeological data, GIS and AHP techniques applied to delineate groundwater potential zones in sandstone, limestone and shales rocks of the Damoh district, (MP) central India. Environ Res 228:115832. https://doi.org/10.1016/j.envres.2023.115832
Mosaad S, Kotb ADM, Basheer AA (2024) Groundwater potentiality mapping: A case study in Baba and Sidri watersheds, South Sinai. Egypt J Afr Earth Sci 210:105145. https://doi.org/10.1016/j.jafrearsci.2023.105145
Muavhi N, Mutoti MI (2023) Using Geospatial Techniques and Analytic Hierarchy Process to Map Groundwater Potential Zones. Groundwater 61:494–509. https://doi.org/10.1111/gwat.13257
Muduli A, Chattopadhyay PB, Pal U (2023) Mapping of heterogeneity on groundwater level and potential zones along expeditiously urbanizing tropical coastal regions. Groundw Sustain Dev 23:101002. https://doi.org/10.1016/j.gsd.2023.101002
Mukate SV, Panaskar DB, Wagh VM, Baker SJ (2020) Understanding the influence of industrial and agricultural land uses on groundwater quality in semiarid region of Solapur, India. Environ Dev Sustain 22:3207–3238. https://doi.org/10.1007/s10668-019-00342-3
Mukherjee S, Patel AK, Kumar M (2020) Water Scarcity and Land Degradation Nexus in the Anthropocene: Reformations for Advanced Water Management as Per the Sustainable Development Goals. pp. 317–336. https://doi.org/10.1007/978-981-32-9771-5_17
Nepal S, Neupane N, Belbase D, Pandey VP, Mukherji A (2021) Achieving water security in Nepal through unravelling the water-energy-agriculture nexus. Int J Water Resour Dev 37:67–93. https://doi.org/10.1080/07900627.2019.1694867
Nguyen HD (2023) Spatial modeling of flood hazard using machine learning and GIS in Ha Tinh province. Vietnam J Water Clim Chang 14:200–222. https://doi.org/10.2166/wcc.2022.257
Nguyen PT, Ha DH, Jaafari A, Nguyen HD, Van Phong T, Al-Ansari N, Prakash I, Le H, Van, Pham BT (2020) Groundwater Potential Mapping Combining Artificial Neural Network and Real AdaBoost Ensemble Technique: The DakNong Province Case-study, Vietnam. Int J Environ Res Public Health 17:2473. https://doi.org/10.3390/ijerph17072473
Nourani V, Ghaffari A, Behfar N, Foroumandi E, Zeinali A, Ke C-Q, Sankaran A (2024) Spatiotemporal assessment of groundwater quality and quantity using geostatistical and ensemble artificial intelligence tools. J Environ Manage 355:120495. https://doi.org/10.1016/j.jenvman.2024.120495
Nourani V, Maleki S, Najafi H, Baghanam AH (2023) A fuzzy logic-based approach for groundwater vulnerability assessment. Environ Sci Pollut Res 31:18010–18029. https://doi.org/10.1007/s11356-023-26236-6
Nwazelibe VE, Unigwe CO, Egbueri JC (2023) Testing the performances of different fuzzy overlay methods in GIS-based landslide susceptibility mapping of Udi Province, SE Nigeria. CATENA 220, 106654. https://doi.org/10.1016/j.catena.2022.106654
Pandey A, Padhya V, Ganguly A, Chakra S, Deshpande RD (2023) Surface water groundwater interaction in water-stressed semi-arid western India: Insights from environmental isotopes. J Arid Environ 208:104879. https://doi.org/10.1016/j.jaridenv.2022.104879
Paria B, Pani A, Mishra P, Behera B (2021) Irrigation-based agricultural intensification and future groundwater potentiality: experience of Indian states. SN Appl Sci 3:449. https://doi.org/10.1007/s42452-021-04417-7
Qin Y, Yang G, Lu K, Sun Q, Xie J, Wu Y (2021) Performance Evaluation of Five GIS-Based Models for Landslide Susceptibility Prediction and Mapping: A Case Study of Kaiyang County, China. Sustainability 13:6441. https://doi.org/10.3390/su13116441
Raja Shekar P, Mathew A (2023) Assessing groundwater potential zones and artificial recharge sites in the monsoon-fed Murredu river basin, India: An integrated approach using GIS, AHP, and Fuzzy-AHP. Groundw. Sustain. Dev. 23, 100994. https://doi.org/10.1016/j.gsd.2023.100994
Ray SK (2024) Identifying the groundwater potential zones in Jamsholaghat sub-basin by considering GIS and multi-criteria decision analysis. Int J Environ Sci Technol 21:515–540. https://doi.org/10.1007/s13762-023-04923-8
Sahoo S, Ramole MM, Dahiphale P, Awasthi S, Pateriya B (2024) Geospatial technology based morphometric analysis and watershed prioritization of lower Satluj basin in India for groundwater recharge potential. Trop Ecol 65:43–58. https://doi.org/10.1007/s42965-023-00307-8
Saravanan S, Saranya T, Abijith D, Jacinth JJ, Singh L (2021) Delineation of groundwater potential zones for Arkavathi sub-watershed, Karnataka, India using remote sensing and GIS. Environ Challenges 5:100380. https://doi.org/10.1016/j.envc.2021.100380
Şener E (2022) Appraisal of groundwater pollution risk by combining the fuzzy AHP and DRASTIC method in the Burdur Saline Lake Basin, SW Turkey. Environ Sci Pollut Res 30:21945–21969. https://doi.org/10.1007/s11356-022-23651-z
Shekar PR, Mathew A (2023) Flood susceptibility mapping of the Peddavagu River Basin using GIS-AHP techniques. pp. 125–141. https://doi.org/10.1016/B978-0-443-18640-0.00011-0
Singh A, Kumar R, Kumar R, Pippal PS, Sharma P, Tanuja, Sharma A (2024) Delineation of groundwater potential zone using geospatial tools and analytical hierarchy process (AHP) in the state of Uttarakhand, India. Adv Sp Res 73:2939–2954. https://doi.org/10.1016/j.asr.2023.12.041
Singh A, Patel S, Bhadani V, Kumar V, Gaurav K (2024) AutoML-GWL: Automated machine learning model for the prediction of groundwater level. Eng Appl Artif Intell 127:107405. https://doi.org/10.1016/j.engappai.2023.107405
Singha S, Das P, Singha SS (2021) A fuzzy geospatial approach for delineation of groundwater potential zones in Raipur district, India. Groundw Sustain Dev 12:100529. https://doi.org/10.1016/j.gsd.2020.100529
Sulaiman WH, Mustafa YT (2023) Geospatial Multi-Criteria Evaluation Using AHP–GIS to Delineate Groundwater Potential Zones in Zakho Basin, Kurdistan Region. Iraq Earth 4:655–675. https://doi.org/10.3390/earth4030034
Tan Y, Wang X, Liu X, Zhang S, Li N, Liang J, Xu D, Yang Q (2023) Comparison of AHP and BWM methods based on ArcGIS for ecological suitability assessment of Panax notoginseng in Yunnan Province, China. Ind Crops Prod 199:116737. https://doi.org/10.1016/j.indcrop.2023.116737
Tiwari A, Kumar M, Thakur S (2024) Groundwater quality assessment for drinking purpose using traditional and fuzzy-GIS-based water quality index in Gurugram District of Haryana, India. Environ Geochem Health 46:52. https://doi.org/10.1007/s10653-023-01783-6
Ullah Z, Rashid A, Ghani J, Nawab J, Zeng X-C, Shah M, Alrefaei AF, Kamel M, Aleya L, Abdel-Daim MM, Iqbal J (2022) Groundwater contamination through potentially harmful metals and its implications in groundwater management. Front Environ Sci 10. https://doi.org/10.3389/fenvs.2022.1021596
Upwanshi M, Damry K, Pathak D, Tikle S, Das S (2023) Delineation of potential groundwater recharge zones using remote sensing, GIS, and AHP approaches. Urban Clim 48:101415. https://doi.org/10.1016/j.uclim.2023.101415
Venkatesh S, Parimalarenganayaki S (2023) Multi-collinearity Based Parameter Optimization and Comparison among Multi-criteria Decision Analysis to Map Groundwater Potential Zones. J Geol Soc India 99:1158–1164. https://doi.org/10.1007/s12594-023-2441-7
Wang H, Wang F, Sun J, Cheng Z, Wang Y, Cao Y (2023) New strategy for evaluating the spatiotemporal distribution of groundwater resource quantity under seasonal freeze/thaw in mountainous areas. J Hydrol 616:128850. https://doi.org/10.1016/j.jhydrol.2022.128850
Wang P, Su X, Zhou Z, Wang N, Liu J, Zhu B (2023) Differential effects of soil texture and root traits on the spatial variability of soil infiltrability under natural revegetation in the Loess Plateau of China. CATENA 220, 106693. https://doi.org/10.1016/j.catena.2022.106693
Warghat SR, Kulkarni SV, Das S (2023) Groundwater potential zones identification using integrated remote sensing and GIS-AHP approach in semiarid region of Maharashtra, India. Case Studies in Geospatial Applications to Groundwater Resources. Elsevier, pp 67–90. https://doi.org/10.1016/B978-0-323-99963-2.00013-4
Yi F, Lu Q, Li Y, Wang Z, Yao B, Yang Q, Wang J (2023) Ecological vulnerability assessment of natural oasis in arid Areas: Application to Dunhuang, China. Ecol Indic 149:110139. https://doi.org/10.1016/j.ecolind.2023.110139
Yilmaz AE, Demirhan H (2023) Weighted kappa measures for ordinal multi-class classification performance. Appl Soft Comput 134:110020. https://doi.org/10.1016/j.asoc.2023.110020
Yu D, Huang Q, Kang X, Liu Y, Chen X, Xie Q, Guo Z (2023) The unsaturated seepage process and mechanism of internal interfaces in loess-filled slopes during intermittent rainfall. J Hydrol 619:129317. https://doi.org/10.1016/j.jhydrol.2023.129317

Download PDF

Reviewers agreed at journal
13 May, 2024
Reviewers invited by journal
13 May, 2024
Editor assigned by journal
05 May, 2024
First submitted to journal
04 May, 2024

You are reading this latest preprint version

Comparative Assessment of Analytical Hierarchy Process (AHP) and Fuzzy Overlay Analysis (FOA) Models in Groundwater Potential Zone Mapping Using Sensitive Analysis: A GIS-RS Integrated Approach

Status:

Version 1

Abstract

Figures

1.Introduction

2. Study area

3. Material and Methodology

3.1. Data acquisition

3.2. Parameter Optimization using Multicollinearity Assessment

3.3. AHP Method

3.4. Fuzzy Overlay Analysis (FOA)

3.5. Groundwater Potential Zone Evaluation and Comparison

3.6. Investigation of Sensitivity and Influence of Different Factors on Potential Zones

3.7. Validation of a model

4. Results and discussion

4.1. Multicollinearity Assessment

4.2. Description of thematic layers used for mapping

4.2.1. Land Use Land Cover (LULC)

4.2.2. Soil

4.2.3. Geomorphology

4.2.4. Lithology

4.2.5. Rainfall

4.2.6. Drainage network density

4.2.7. Lineament density

4.2.8. Slope

4.2.9. NDWI

4.2.10. Elevation

4.3. Groundwater Potential zones mapping

4.4. Exploring Sensitivity: The Feature Elimination Approach

4.5. Validation of Groundwater potential zone map

4.6. Discussion

5. Conclusions

Declarations

References

Supplementary Files

Status:

Version 1