Introduction

In India Groundwater is a vital resource that provides water to communities and industries. It is used for the preservation of the environment and human well-being (IPCC 2001; Arefin 2020). The steady growth in groundwater consumption has little or no regard for the environment's balance (Zhang et al. 2019; Nhamo et al. 2020). Around 44% of the world's groundwater is used for domestic and agricultural applications, and 27% for industrial use. In India, it is used for irrigation and drinking water (Patra et al. 2018a, b). It is also used for drinking water and irrigation. Around 90% of the groundwater in India is used for agricultural irrigation, including Sagar Island (Schneider 2018). Furthermore, groundwater use in this deltaic island gets incremented in recent years as a result of expanding socioeconomic change. Unfortunately, a lack of groundwater has posed a threat to the residents of this region (Gayen and Zaman 2013). As a result, identifying and assessing critical parameters is essential for determining groundwater potential (Patra et al. 2018a, b). Furthermore, as a result of industrial expansion and high population rise, freshwater demand on a global scale has surged during the last century(Arefin 2020; Zhang et al. 2019). To fulfill the water demand of an area, identification of ground water potential zone is of supreme importance (Rana et al. 2022). Sagar Island in India is located in one of the coastal locations, recognized as an erosion-prone saline area (Gopinath and Seralathan 2005). As a result, officials' primary concern is assessing the existence of groundwater. Some conventional methods, such as field survey including drilling and pumping test, hydrogeological study, resistivity survey, are common ways for identifying and delineating groundwater. These methods also demand a significant amount of time, funds, and resources for exploration, as well as the participation of professionals. (Arefin 2020; Shao et al. 2019a; Achu et al. 2020; Nhamo et al. 2020; Kamruzzaman et al. 2020; Shao et al. 2020). Modern technologies such as “Geographic Information Systems” (GIS) and “Remote Sensing” (RS) are economical, efficient, and easily accessible.

Geospatial technology has now evolved as an effective tool in the area of hydrogeological study (Saravanan et al. 2021). Integrating RS and GIS in groundwater research constitutes a paradigm revolution in groundwater studies that allow improved assessment, proper monitoring, management, and conservation of the groundwater resources (Silwal and Pathak 2018). The technologies applied here can be beneficial in groundwater mapping research. The amalgamation of RS and GIS for groundwater potential assessment and mapping allows easy storing, handling, and analysis of the generated data in a variety of formats and scales(Abijith et al. 2020; Shao and Cai 2018; Ahmad et al. 2020; Shao et al. 2018; Yin et al. 2018; Siddi Raju et al. 2019). The composite layers of the data that are used in these techniques share a general coordinate system, that may be used to construct thematic or distribution maps over the whole research region. The use of RS with groundwater flow reckonings is an effective way to calculate a lake, river, and reservoir’s boundaries, seepage areas, evapotranspiration zone. In an advanced GIS system, RS data can also be used to create thematic maps, apply proper weights, and achieve convergence. This ensures and increases the accuracy of anticipating the location of promising locations as well as their groundwater capabilities (Abijith et al. 2020; Silwal and Pathak 2018; Shao et al. 2019b). Combining RS with GIS is an important hydrogeologic map data source, including RS for the management and mapping of water resources, for hydro-geolocation, for identifying and traceability of salinization, for decisive biomass, land use, and crops choice, similarly as for environmental protection and management(Achu et al. 2020; Mahmoud and Alazba 2016; Arefin 2020).

Therefore, it is highly beneficial to utilize RS data with GIS to detect GWPZs (Mahmoud and Alazba 2016) as the characteristics of the groundwater are intensifying. Satellite data provide timely and precise guidance and information on several elements that directly and/or indirectly impact the occurrence of groundwater and movement (Jasmin and Mallikarjuna 2011). In addition, GIS is an excellent workstation to manage extensive and complicated spatial–temporal data effectively (Patra et al. 2018a, b). RS is an acceptable technique for regions with poor or absent geology, hydrogeological, and field data (Yousif et al. 2018). Much previous research (Mahmoud and Alazba 2016; Oikonomidis et al. 2015; Mohammadi-Behzad et al. 2018; Shao et al. 2019b; Pinto et al. 2017; Yousif et al. 2018; Siddi Raju et al. 2019) had used RS and GIS methodologies to measure the potential zone of groundwater storage (Table 1).

Table 1 Index of spatially reactive detectors for groundwater potentiality investigations (Crétaux et al. 2011; Carleer and Wolff 2006; Gaber et al. 2010; Hall et al. 2008; Lavender and Groom 1999; Turk and Miller 2005; Song and Park 2014)
Full size table

Multifactor analysis, single-factor analysis, fuzzy clustering, GIS information fusion, brittle rock proportion, and Fuzzy-AHP indices have all been used to assess a GWPZ (Chen et al. 2010; Ahmad et al. 2020; Siddi Raju et al. 2019; Singaraja et al. 2015; Pinto et al. 2017). The analytic hierarchy process (AHP) developed by Saaty (1990), the fuzzy model by Zadeh (1965), and the RS–GIS methodologies are utilized to define the GWPZ in this research. Furthermore, ten criteria has been employed to define the GWPZs that are elevation, slope, distance from the creek, drainage density, rainfall, LULC, plane and profile curvature, topographic wetness index (TWI), and soil type. Multi-criteria decision analysis (MCDA) is considered an important method for formalizing and addressing environmental concerns and the decisions to be made (Steele et al. 2009). MCDA's general aim is to choose a resource from an existing collection of resource options. A variety of geographical difficulties have been confronted with the GIS–MCDA method. The multi-criteria methods has been applied to several hydrological variables to estimate geographically potential regions for groundwater storage. For illustration, Adiat et al. (2012), investigated the suitability of the GIS–MCDA technique as a structural tool for modeling groundwater potential zones in Malaysia's Kedah and Perlis states. Which examined the influence of criterion consistency on the efficiency of a prediction tool and found that the approach is successful at producing accurate and consistent predictions. On the other hand, Singh et al. (2017), estimated rainwater harvesting effectiveness and befitting sites, as well as simulated recharge structures used in the Canal Command in the East part of India. Murthy and Mamo (2009) also used the MCDA approach to identify GWPZs in the Moyale–Teltele sub-basin of South Ethiopia. Furthermore, Ajibade et al. (2019) had described a method for the selection of appropriate disposing zone and management of solid wastes by the application of GIS and MCDA techniques in Akure, Ondo State, Nigeria. Multiple scholars have altered and enhanced this theory to make it more user-friendly (Liu and Pedrycz 2009; Cagman et al. 2011).

In the present study much effort has been put into studying the groundwater system in Sagar Island using various hydrological and geological properties. The objective of the study is (a) to propose a technique of identification and demarcation of distinct GWPZ based on numerous possible variables using Fuzzy-AHP; (b) to analyse the connections between the different conditioning factors (c) to validate the performance of the model using geophysical survey.

Materials and methods

Study area

Sagar Island is the deltaic part of the lower Gangetic plain and falls under the tract of Indian Sundarban. It is the biggest deltaic island of Indian Sundarban and is bounded on the north and west by the Hooghly River, the east by the Muriganga River, and the south by the Bay of Bengal. The coastline alterations on this island are mostly controlled by the Hooghly and Muriganga rivers. This island's attitudinal and longitudinal extents ranges from 21°37′21 to 21°52′28′′N and 88°10′25′′E to 88°12′17′′E, respectively (Nandi et al. 2016). Sagar Island is 30 km long in the direction of north to south and 12 km wide at its widest point (Fig. 1).

Fig. 1
figure 1

Location of Sagar Island, India

Full size image

Each year this island suffers from a number of natural disasters, including cyclones, floods, earthquakes, and tidal fluctuations. The dynamicity of this island has changed due to the variable pace of erosion and accretion. The rapid shift of the island's shoreline has had an impact on human habitation and the coastal ecology.West Bengal's shoreline is part of the Amero trailing edge coast, while the opposite continental coast is a collision coast (Nandi et al. 2016). All the main rivers in the Sundarbans deltaic region flow in the direction of north to south. Sagar Island has a tropical oceanic climate. Summer temperatures vary from 28 to 36 °C, while winter temperatures vary from 10 to 25 °C in this deltaic island. The yearly rainfall in the area varies between 1500 and 2400 mm (Nandi et al. 2016). The lower region of the Bengal Basin in the form of the delta extended towards the sea and is surrounded by “Pre-Cambrian crystalline” rocks on the west, north, and north–east, and the “Tertiary hill ranges of the Assam-Burma arc” on the east. Intertidal mangroves are submerged twice daily during high tides by a complicated network of tidal streams. Sagar Island's primary mangrove species are (I) A. alba (II) A. marina (III) Bruguieragymnorrhiza (IV) S. Griffith (V) Aegicerascorniculatum (Das 2006; Nandi et al. 2016).

Components that influence the potential of groundwater

The distribution and rejuvenation of groundwater are influenced by a variety of factors. Because of the various elements that influence groundwater occurrence and recharge, the groundwater condition in any given place might vary dramatically (Mahmoud and Alazba 2016; Huq et al. 2018). The components are often assessed on different scales and shown on maps. In multifactor decision-making, component maps are built on equal levels, classifying the factor maps according to the scale variances. The relevant classes are then utilized as the root for intelligible mapping with every aspect included. In the following study, ten parameters were used for GWPZ mapping. The elements are organized into ten thematic layers: elevation, slope, rainfall, drainage density, plane and profile curvature, land use/land cover (LULC) classification, TWI, and the soil type. The above-mentioned components are interrelated and treated as a single unit when categorizing the GWPZ. Four assumptions have been established to improve groundwater potential to pick the most significant aspects: (1) increase groundwater recharge (penetration rates), (2) high soil and rock penetrability (geological & lithological unit), (3) intensified drainage density, (4) flat slopes.

Preparation of prospective areas of groundwater and data source

The RS satellite data have been used in the present study to prepare the thematic maps (Layers), which show several elements that affect groundwater potential in research areas. Groundwater recharge is a plan of action, where injecting water in an unsaturated area from a saturated area is done for recharging and balancing water tables. The pattern of drainage density and slope was obtained from the Shuttle Radar Topography project using satellite data. In groundwater storage, structural elements such as slope and drainage density play an essential part as a route for groundwater. A lot of surface rush but not much infiltration can be managed in high drainage densities zones. In the mild slope areas, water penetration is greater.

LULC, which is derived from Sentinel-2A satellite data that has been processed. Any terrain's curvature can have concave or convex upward profiles (Nair Hema et al. 2017). The profile and planform curvature of the Sagar Island were calculated in the GIS domain using a DEM as an input. Profile curvature emphasizes the circumferential change (convexity and concavity) in the direction of maximal slope gradient. In general, the concavity of the surface is denoted by positive values of profile curvature (therefore, the escalation of flow), and the convexity is denoted by negative values (deceleration of flow).

The topographic wetness index (TWI) evaluates the influence of topography on hydrological processes (Mokarram et al. 2015). The syntax used to calculate the TWI for the selected study area through the GIS arena is as follows: ln (flow accumulation + 0.001)/ (slope in percent rise/100) + 0.001. The value 0.001 is added to flow accumulation and the gradient of the slope, which ensures maximum figures of cells for computation of the TWI of the receptacle. TWI is an indicator by which the assessment of the potential of wetland, where the positive readings indicate strong potential and negative values indicate weak potential which increases the risk of erosion.

Altitude has a significant impact on climatic conditions and results in a different classification of plants and soil in each place (Jothibasu and Anbazhagan 2016; Aniya 1985). Aside from that, it has an indirect influence on GW potential; for example, lower altitudes have lower slopes, and the rate of infiltration increases appropriately (Althuwaynee et al. 2012, 2014). The Sagar Island elevation map was derived from the “SRTM-DEM.” Creeks are the primary source of GW recharge in coastal island areas. As a result, one of the key hydrological factors influencing GW potential is the distance from creeks. This layer was created using the “Euclidean distance function”. The Soil and Land-use Survey of India and Global Precipitation Measurement provide soil characteristics and precipitation data, which were used to extract data on soil, and rainfall. Following that, thematic layers were created using the factors applied to the RS and GIS dataset, as the dataset included a variety of attributes such as geographic resolution, conservation, and structure. As a result, all the datasets were transformed to assured equally distributed features, grid formatted, and re-envisaged using the WGS-84 method. The complete methodology is shown in Fig. 2.

Fig. 2
figure 2

Schematic flowchart of GWPZ map

Full size image

Fuzzy logic model

Zadeh (1965) proposes the fuzzy model. It's a methodology for modeling and simulating complicated systems. In spatial planning, the execution of the spatial objects as set numbers is widely used for the preparation of the map. Zadeh (1965) explained the level of membership functions of each item in the fuzzy set model which is indicated by the membership value. The Triangular Fuzzy Number (TFN) M is shown in Fig. 3.

Fig. 3
figure 3

Triangular fuzzy number (TFN)

Full size image

The main purpose of the fuzzy model is the ambiguity of the data. The fuzzy model accepts both mathematical and programming functions. For the mapping of GWPZ, the fuzzy model allows the idea of imperfect site membership and appraises several groups. In these conceptual circumstances, the analysis of Geographic diversity and patterns of a particular area is conducted by the fuzzy membership function (FMF) which contributes to creating fuzzy boundaries in all possible locations. In the present study, for each potential zone, FMF’s were assigned to their variation as their course appeared from fuzzy borders. Five distinct levels were chosen and weighted and each level had a range of fuzzy numbers indicating a distinctive assessment in the fuzzy dialect (Table 2). For example, in the FMF, the low level is represented by the numbers 0, 3, 3, and 5 and the high level by 5, 7, 7, 10. All these fuzzy number ranges are given by analyzing previous literature.

Table 2 Five levels with fuzzy numbers
Full size table

Normalization and weight assignments

The primary objective of the AHP method is to carter for expert knowledge, but it can't be completely reliant on human judgment. Therefore, to deal with the problems related to the fuzzy hierarchical process the Fuzzy-AHP fusion was designed. This research employs the Fuzzy-AHP concept to estimate fuzzy weights through pairwise comparisons by fuzzification of the hierarchical process with fuzzy numbers. When utilizing Fuzzy-AHP to evaluate the weights of provided criteria, the following procedures were taken into account.

Step I: A pairwise comparison matrix was created in a hierarchical framework using all of the criteria. For paired assessments, philological rules were applied, and in all cases, one of the two parameters was more important.

Pairwise comparison by AHP can be used to create judgment matrices (A):

\(A=\left[\begin{array}{cccc}\begin{array}{c}{A}_{11}\\ {A}_{21}\\ {A}_{31}\end{array}& \begin{array}{c}{A}_{12}\\ {A}_{22}\\ {A}_{32}\end{array}& \cdots & \begin{array}{c}{A}_{1n}\\ {A}_{2n}\\ {A}_{3n}\end{array}\\ \vdots & \vdots & \ddots & \vdots \\ {A}_{n1}& {A}_{n2}& \cdots & {A}_{nn}\end{array}\right]\) = \(\left[\begin{array}{cccc}\begin{array}{c}{A}_{1}/{A}_{1}\\ {A}_{2}/{A}_{1}\\ {A}_{3}/{A}_{1}\end{array}& \begin{array}{c}{A}_{1}/{A}_{2}\\ {A}_{2}/{A}_{2}\\ {A}_{3}/{A}_{2}\end{array}& \cdots & \begin{array}{c}{A}_{1}/{A}_{n}\\ {A}_{2}/{A}_{n}\\ {A}_{3}/{A}_{n}\end{array}\\ \vdots & \vdots & \ddots & \vdots \\ {A}_{n}/{A}_{1}& {A}_{n}/{A}_{2}& \cdots & {A}_{n}/{A}_{n}\end{array}\right]\)

where An indicate the nth indicator element along with the judgment matrix component Aij.

Traditional AHP makes a logical decision but is surrounded by ambiguity. As a result, this model is inadequate to adequately mirror human thought. fuzzy logic, on the other hand, is a sophisticated logical approach based on the traditional AHP procedure. TFN represents the matrix element in the normalization procedure. TFN is capable of avoiding ambiguity, inaccuracy, and uncertainty. Weights were used with the same Saaty grading system to compare the outcomes of the two procedures. For the fuzzy scaling ratios, the numbers 1/2, 1, 3/2, 2, 5/2, 3, 7/2, 4, 9/2, and 5 were utilized, indicating the potential of one component to influence other components using intermediate values rather than frangible integers:

$${A}_{ij}=\left(l,m,u\right)=(l,\frac{{A}_{i}}{{A}_{j}}, u)$$
(1)

Here \(l, m, \mathrm{and} u\) represent the lowest potential, moderate probable, and maximum conceivable values, respectively. In this equation, Aij is the decision matrix component from Eq. (1).

Step II: Determine the normalized weight (Eq. 2).

$${W}_{n}=\left(\frac{GM}{\sum_{n=1}^{{N}_{f}}{GM}_{n}}\right).$$
(2)

In the equation 3, GM represents the geometric mean of the ith line in the decision matrices and is defined as

$$\mathrm{GM}=\sqrt[\mathrm{Nf}]{{A}_{1n}{A}_{2n}\dots ..{A}_{n\mathrm{Nf}}}$$
(3)

The weight normalization and the elimination of individual weight have been done using the characteristic vector technique. By calculating the CR of various categories of the subject variables, the uniformity of the normalized data was checked. In contrast to the randomly generated matrix, the continuity level of the matrix is shown by its CR. In the matrix from Table 3 the consistency ratio was 0.03 and is acceptable (Saaty 1990).

Table 3 Fuzzy-AHP normalized comparison matrix in pairs for possible groundwater areas
Full size table

Step III: The precision of paired comparisons was deliberated using the consistency index (CI). On the other hand, the consistency of the discernment was validated using the CR. The equilibrium condition, based on the assigned weights, is maintained by the CR. The weights assigned to the parameters are considered based on field experience but they should be as accurate as possible rather than exaggerated. The maximum CR value that may be used is 0.10. The CR equation is as follows:

$${\text{CR}}\, = \,{\text{CI}}/{\text{RI}}{.}$$
(4)

The following equation is used to compute the Consistency Index (CI) for a matrix:

$${\text{CI}}\, = \,\left( {\lambda_{\max } - n} \right)/\left( {n - 1} \right),$$
(5)

where λmax = the maximum (weight) value and n = the number of criterion. The Random Index (RI) values in Table 4, developed by Coyle (2004) give the sum of random values in which the top row matches the random assessment consistency index, and the lower row corresponds to the random matrix order (n values).

Table 4 Average random consistency indices for various n
Full size table

Groundwater potential zone

GWPZ map has been created in combination with the compatible thematic layers and overlaid GIS data (Weighted). In such experiments, where a weighted overlapping method is used, the vector and raster data are applied equally. The selected elements were integrated and the value of each conditioning factor has been calculated in the aggregated findings. Furthermore, a weighted linear factor combination (WLC) is also used to produce the GWPZ map.

Figure 4 illustrates the methods for delineation of GWPZ using RS, GIS, and Fuzzy-AHP integration. It demonstrates the generation of subject layers, including elevation, slope, plane, and profile curvature, TWI, drainage density, precipitation, LULC, distance from the creek, and soil type, as the initial phase. Fuzzy-AHP has been applied to calculate the weight of the factors in the second stage. The thematic layers have been merged into the GIS environment via the overlay technique. GWPZ has been calculated using the following equation:

Fig. 4
figure 4

a Slope and b drainage density map of the Sagar Island, India

Full size image
$$\mathrm{GWPZ}=\sum_{i=1}^{n}*\sum_{j=1}^{m}{w}_{i}{x}_{j}$$
(6)

where GWPZ = groundwater potential zone map, wi = normalized weight of the ith thematic variable (Table 5), xj = normalized weight of the variable's jth class (Table 3), m = the total number of variables classes, and n = total number of variables.

Table 5 Assignment and normalization of weights for ten thematic layers or maps
Full size table

Model sensitivity analysis

The term sensitivity analysis for any type of model is used to identify the effectiveness of the output based on the insert criteria. The model analyzes the exerted result efficiency and validated the impacts of the parameters or factors that are adopted. The influence of the data used validating weighted values and the weights applied to individual elements, sensitivity analysis has been utilized. Equation 8 has been used to identify the effectiveness of every weight:

$$\mathrm{Effective weight}=\frac{{\mathrm{Theme}}_{w}*{\mathrm{Theme}}_{s}}{\mathrm{GWPI}}$$
(7)

where w = the weight of the factors given to each grid and s = scale value of the factors. GWPI refers to the groundwater potential index, as calculated from Eq. 8.

Conditioning factors

Based on the influence on groundwater potential in the Sagar Island, 10 thematic layers are prepared using RS satellite data. The thematic factor maps using geological and environmental information were constructed in the fixed spatial scale of specified parameters using conventional data/maps and RS data. As a result, it does not indicate the specific features of the inhabitants. Elevation, slope, plane and profile curvature, TWI, drainage density, precipitation, LULC, distance from the creek, and soil type are the layers. Though, the thematic maps, or in another word the layers utilized to categorize GWPZ differ within the research region, and the weights attributed to the features or the classes of the layers used for mapping are chosen at randomly (Table 5).

Slope

Groundwater replenishment or recharge largely depend on the slope aspect. Uniform or shallow slopes are significant signals for groundwater potential zonation as they provide better renewable water levels than steeper slopes (Mumtaz et al. 2019; Nayyer et al. 2019). The amount of water held and the precipitation infiltrated is affected by the slope. Surface runoff particulate material transport and erosion are the major factors or functions of this system. Water penetrating capacity is a path feature, which means that when the path is at its lowest, maximum penetration occurs (Patra et al. 2018a, b; Sarker et al. 2020a). Hilly areas with a greater degree of slopes have a high water flow and low infiltration. On the contrary, flat low-lying regions may contain more precipitation and a greater groundwater recharge capacity. However, the development of a digital elevation model is the key to a slope map (DEM). A DEM with a resolution of 30 m is adopted to prepare a slope map. The values of the slope were computed from the DEM using the Arc-GIS slope tool. The elevation of Sagar Island is between 0 and 8 m (Fig. 4a). The slope map is divided into 5 categories: < 1, 1.1–2.0, 2.1–30, 3.1–4.0 and > 4, as shown in Fig. 4a. The lowest slope group gets the weight mostly because of less runoff and considerable infiltration. The highest slope class is allocated the least weight value as a very low groundwater yield is probable (Agarwal and Garg 2016). As a result, The locations where the degree of slope is lower are seen to be ideal for rich groundwater potentiality, and the opposite cases are seen in the areas with a higher degree of slope. Areas of 1 (about flat) fall into the excellent category because of the flat topography, which has a very high perforation ratio and a generally strong GWP zone. The regions having a slope range of 1.1–2.0 (gradual sloping) are believed to be a source zone of high groundwater potential due to the gently rolling terrain and low surface runoff. Moderate GWP areas have a slope of 2.1–4.0 (steep-to-steep slope) and have mild to no runoff. The regions having slopes of > 4.0 (highly steep) are reviewed as poor GWPZ because of steep terrain, where high runoff and low infiltration are observed.

Drainage network density

This phenomenon is directly is related to the soil's infiltration, happening naturally. It is also the major hydrogeological research parameter. Drainage and density are significant indications of the hydrogeological characteristics of an area. Climate, soil lithology, roughness, penetration efficiency, the intensity of vegetation, and drainage intensity all have a part to play (Sarker et al. 2020c; Mahmoud and Alazba 2016). The reduced soil-flow potential is caused by high-surface drainage's low recharge ratios. After generating a map of drainage density from the DEM, the intensity of drainage is computed using the ArcGIS "density analysis tool" (Fig. 4b). In general, the closeness of stream channel spacing determines drainage density. The drainage density of a basin is calculated by dividing the entire river network length by the area of the basin:

$${D}_{N }D=\sum \frac{L}{A}$$
(8)

DND signifies the drainage network's density, L stands for the stream's length (total), and A signifies the total area.

The range of the drainage density varies from 0 to > 2 km/km2 and is divided into five categories: < 0.5 km/km2 (excellent), 0.51–1 km/km2 (good), 1.01–2 km/km2 (moderate), and > 2 km/km2 (poor). Because it is inversely related to topography perviousness, low intensity of drainage areas is a good predictor of water reserving zones. Similarly, high runoff is influenced by high drainage intensity, indicating a low possibility of groundwater. Due to this nature, low drainage density zones are given a higher value of weightage for low surface runoff and high infiltration characteristics (Agarwal and Garg 2016; Sarker et al. 2020b). In the center of Sagar Island, there is a high level of drainage intensity. Low drainage quality affects the majority of the study region. Rainwater infiltration is reduced when there is less permeable rock. In comparison to a high-drainage-intensity area, a low-drainage-intensity area causes more infiltration and less surface runoff. Therefore, as the areas with higher drainage are not appropriate for groundwater potentiality as significant surface runoff and low infiltration observed (Mohammadi-Behzad et al. 2018; Achu et al. 2020; Yin et al. 2018; Adnan et al. 2020; Mumtaz et al. 2019).

Rainfall

Effective rainfall recharges the aquifers and groundwater in most cases. Based on the ground parameters, such as soil lithology, hydro-geomorphology, etc., the amount of rainfall can be used to estimate the capacity of a groundwater reservoir. Strong groundwater potentiality is directly correlated with higher rainfall as it is an important component of water resources and a major contributor to groundwater recharging (Shao et al. 2020; Minh et al. 2019). Rainwater flow and its topographical distribution have a high influence on hydro-geology. Meanwhile, the intensity of precipitation and the combination of several favorable situations help to detect GWPZ. In case of high rainfall, there is a substantial chance of additional groundwater. In contrast, if the precipitation is minimal, the potential of groundwater is lower. Not only temporarily but spatially, rainfall differs. For this reason, the impact of rainfall in the identification of GWPZ must be determined. The Inverse Distance Projection Weighting (IDW) technology in the ArcGIS environment develops the geographical distribution of the average annual precipitation map. Satellite data has been used to calculate rainfall. Sagar Island has an oceanic climate, which means that rainfalls predominantly in the summer due to oceanic influence. The Sagar Island region receives 150 mm–170 mm of rainfall per year (Fig. 5a), with the northern regions of the province receiving the most. There are five categories on the precipitation map: < 155 mm, 155.1–158 mm, 158.1–161 mm, 161.1–164 mm, and > 164 mm. Among those categories, the foremost weight value is given to higher rainfall classes for better recharge rate, leading to greater groundwater potentiality and lower weight given to the less rainfall rate classes.

Fig. 5
figure 5

a Rainfall, b Soil type, c Land use Land cover and d Distance from drainage map of the Sagar Island, India

Full size image

Soil type

As the water passes through the soil through infiltration, percolation, and seepage so the type of soil and its characteristics have a significant role on aquifer condition or groundwater storage. As a result, the quality of soil influences various recharge components and the quantity of research water inflow into the subsurface (Ahmad et al. 2020). The primary determinant of water infiltration and soil porosity is soil texture, which assists in the definition of the GWPZ. Infiltration is influenced by two main factors: soil porosity and surface runoff. Surface runoff occurs when soil is less permeable, allowing for less penetration, whereas porous soil aids in water recharge (Siddi Raju et al. 2019). The National Bureau of Soil Survey (NBSS) and Land Use Planning (LUP) maps were used to create the soil map. The research area is categorized into five main soil types (as indicated in Fig. 5b): clay (22.63%), clay loam (52.53%), loam (14.41%), sandy loam (2.44%), and sand (7.99%). Depending upon the soil quality and penetration rate, each soil unit received a varied weight. Sandy loam has a wider soil foundation, lighter texture, and greater soil nutrient quality than other soil types, but it only covers a small area, making it suitable for groundwater storage (Shao et al. 2020). As a consequence, the weight was the highest. In many places, loam is found, but it is quite widespread in distribution. Loam, which consists of coarse grain with a small amount of clay, is highly porous, and sensitive to drought, proved extremely effective for the storage of groundwater (Mumtaz et al. 2019). Due to high organic content, silty clay loam is easily getting dissolved by the infiltrating water and flows towards the consecutive bottom underlying layers. This type of soil has a moderate capacity for storing groundwater due to its high clay content. The lowest water permeability soil is clay, mostly found in the center of Sagar Island.

Land use/land cover (LULC)

LULC changes have emerged as one of the most important human functions affecting groundwater supplies. LULC has a considerable influence on groundwater occurrence and renovation in landscapes. To address human interference with groundwater, a specific element, such as LULC, was then explored. Groundwater restoration can also be aided by agricultural growth (Mumtaz et al. 2019). The LULC modifications might influence recharging rates that could have serious repercussions for groundwater quality, particularly in coastal zones, such as Sagar. They are a key factor in the recharge cycle of groundwater, because of their impact on evapotranspiration, drainage, and recharge of the groundwater system. The LULC map, which was created using Sentinel -2A remote sensing images, is shown in Fig. 5c. Vegetation (19.07%), Agricultural land (31.80%), waterbodies (13.30%), sand bar (1.66%), and built-up (34.17%) are the five categories on the LULC map. The largest LULC is occupied by built-up, while the lowest LULC is filled by a sand bar. Water bodies, Agricultural fields, and sand bar areas are all good places to store groundwater, since agricultural land allows for more water accumulation in the soil pores. In the meantime, land expansion create deficiency in soil pore spaces hence reducing the infiltration of water for groundwater recharge. The water penetration in parts of a high and impermeable surface (constructed areas) is, therefore, decreased, and the weight for constructed areas is lower, since the settled areas are considered less important. However, water bodies, agricultural land, and sand bar were given the most weight, while urban areas and vegetation were given the least weight, according to the GWPZ (Table 5).

Topographic wetness index (TWI)

The topographic wetness index (TWI) is often used for calculating topographic control on water processes and indicates the possible infiltration of groundwater produced by topographical impacts, hence, greater TWIs signify a larger potential value for groundwater. The index depends both on the slope and on the upstream area per orthogonal breadth per unit in the direction of the flow. To Understand the topographical results on the placement and size of saturated supply areas for generating runoff, TWI is one of the widely used methods (Beven 1997; Beven and Freer 2001). It has the following definition:

$${\text{TWI}}\, = \,In\left( {\alpha /\tan \beta } \right),$$
(9)

where α = Upslope contributing area; β = Topographic gradient (Slope).

The TWI of the present study area ranged from < −0.70 to > 4.7. There are five sets of results derived and the set ranges are < −0.70, −0.6 to 1.1, 1.2 to 2.9, 3.0 to 4.7, and > 4.7, respectively. A higher value of weight had been given to high TWI and contrariwise. The TWI map of Sagar Island is shown in Fig. 6b.

Fig. 6
figure 6

a Elevation, b TWI, c plane curvature and d profile curvature map of the Sagar Island, India

Full size image

Curvature

Curvature is a quantitative indication of the character of a surface profile. It is possible to have concave upward or convex downward profiles. Platforms with gentle slopes typically have a concave downward profile. Water tends to decelerate in the convex profile and accumulate in the concave, downward profile, which is also true for groundwater (Nair Hema et al. 2017). The profile curvature and plane curvature of the Sagar area are recategorized and split into three groups: concave, flat, and convex. A higher value of weight had been given to high curvature values and vice versa. The profile and plane curvature map of the study area is displayed in Fig. 6c, d.

Elevation

The roughness of the topography performs a significant role in defining groundwater potential zones and is reflected in the elevation of a given location. Mostly the groundwater recharging is done through the low elevation zones, where the rainwater gets stored for a longer period and slowly gets percolated in the underlying soil layers (Thapa et al. 2018; Patra et al. 2018a, b). Similarly, the areas having high elevation retain rainwater for shorter periods and more runoff. As a result, lower elevation locations were given more weight than higher elevation places.

The research area's topography is defined by two unique features: extensive uplands in the northern region. Low-elevation areas feature loosely structured porous medium, which allows for higher groundwater recharge. The research area's elevation (Fig. 6a) ranged from 0 to 9 m above sea level (above mean sea level). The elevation was then categorized into four groups based on natural terrain conditions and raster layer’s natural cell values: (a) < 2 m (2.44% area), (b) 2.1–4.0 m (16.87% area), (c) 4.1–6.0 m (58.95% area), (d) 6.1–8 m (15.75% area) and (e) > 8.0 m (5.99% area) above mean sea level.

Distance from the drainage

Groundwater levels are significantly affected by the distance from the river. This looks to be a beneficial effect, especially in flat terrain (Pham-Duc et al. 2020). The groundwater level near the creek is higher than in regions further away. When it comes to groundwater risk, the distance between water bodies might be subjective.

As a result, all locations within 500 m of the drainage line are classified as having a very high potential for groundwater recharge, whereas areas within 500, 1000, 1500, 2000, and 2500 m of the drainages are categorized as having a very high, high, moderate, low, and very low potential for groundwater recharge, respectively, as shown in Fig. 5d

Results and discussion

Modelling and identification of groundwater potential zone

The application of the Fuzzy-AHP was pioneered by Van Laarhoven and Pedrycz, integrating the benefits of both of these methods. For a broad range of studies, including an assessment of the quality of groundwater (Minh et al. 2019), the exposure of the ecosystem and the environment, ground sensitivity, landslides (Mallick et al. 2018), and the extraction of quality of water, FAHP has become the widest known multi-criteria decision-making approach. In present research the capabilities of the GIS-generated FAHP model have been used more effectively and efficiently. The empirical assessment of ten thematic layers was used to identify the geographic ranges of the GWP zones. Various spatial exploration techniques were used in the model to solve geographic issues in defining prospective groundwater potential zones. Potential groundwater zones for the research region were delineated using RS–GIS techniques based on a synthesis of numerous parameters such as elevation, slope, plane, and profile curvature, TWI, drainage density, precipitation, LULC, distance from the creek, and soil type. According to the findings of this study, the weighted overlap analysis technique, which employs a GIS-based FAHP approach, is the most potent strategy for analyzing possible groundwater areas. The interpreted layers were produced in ArcGIS utilizing the spatial analyst tool and the weighted overlay classification.'

The GWPZ map of the Sagar Island region (Fig. 7) depicts five distinct categories (zones) representing the region's very high, high, moderate, low, and very low groundwater probability zones. Very high and high GWPZ are typically connected with large groundwater levels, which are calculated using several parameters. Groundwater conservation is best suited in areas with flat, mild slopes, porous, and unconsolidated terrain. The area covered by the very high GWPZ is approximately 16.03 km2 (6.84%). The study region's high GWPZ encompasses the northern and central half. The high GWPZ is spread across the research region, accounting for 83.69 km2 (36.23%) area. Mainly the northern portion and the central part of the Sagar island are having the most significant GWPZ. The moderate potentiality (GWPZ) consists mainly of dense dispersed spots that are consistently distributed over the region. It covers 63.94 km2 or nearly 27.68% of the total land. However, the southern and most northern parts of Sagar Island fall under low and very low GWPZ, covering approximately an area of 40.49 km2 (17.53%) and 27.07 km2 (11.72%), respectively, due to its raised slope with unfavorable geological and geomorphological conditions. Low and Very low GWPZs often have low discharge ratios (volume/time) in comparison to excellent and moderate GWPZs.

Fig. 7
figure 7

Groundwater potential zones map of Sagar Island, India

Full size image

The present study is significant in providing a complete foundation for the development and management of groundwater resources in Sagar Island through policymaking. The approach used in the study arose from reasonable criteria of a general kind. This general approach to delineate GWPZ is intended to ensure long-term aquifer development management. The GWPZ strategy emphasizes groundwater conservation, the use of appropriate scientific plans, the regulation of domestic and industrial water preservation practices, and the development of long-term strategies in water resources management and groundwater evaluation. Furthermore, the execution of an adequate groundwater evaluation plan could result in the storage and quality management of groundwater resources.

Accuracy assessment of the model

The validation of the FAHP model has been done using geophysical survey. 30 vertical electric soundings (VES) data has been used for accuracy assessment of the logic-based FAHP model. VES data were acquired from a field study on Sagar Island, India, and the collected information from these analyses has been applied in this groundwater development study (Table 6), indicating that geological hazard mitigation and geological environment protection are highly supported. These areas have the necessary infrastructure to support important policy objectives, such as management of natural or environmental resources, water supply security, and building up a maintainable society. In VES survey, the data were collected up to a maximum depth of 296.2 m below the ground surface. This technique is validated by linking surveyed VES data with the map prepared for GWPZ in ArcGIS survey. According to the findings, 53.33% (16VES points) and 16.66% (5VES points) of total VES points are related with very high and high GWP regions, respectively, whereas 6.66% (2 VES points), 13.33% (4 VES points) and 10% (3VES points) are connected with Medium, Low and very Low GWP regions, respectively (Fig. 7). The analyzed data shows that the GWPZ model is very much compatible with the practical data, such as VES data, and that the majority of active wells are located in very excellent and good zones. This demonstrates the capacity of these technologies (RS and GIS) for understanding GWPZ of an area and might aid in locating suitable groundwater extraction sites. The validation results shows that the set of generated data and the methodologies used for indentification of potential zones provided with correct results by analyzing parameter fitness and comparing significant weights of the components.

Table 6 Model accuracy assessment using geophysical survey
Full size table

Sensitivity analysis

The statistical analysis provided in Table 7 (driven by Eq. (8)) specifies a substantial difference in the adopted factors or variables. This is primarily due to the occurrence of an inundated plain and mature flooded regions with creek systems across the area of the research. Groundwater can be stored in these territories (plain/flat and moderate slope, pervious, unconsolidated). Elevation, slope, rainfall, drainage intensity, curvature, TWI, distance from the creek, LULC, and soil quality are all factors that affect the GWPZ. This might be owing to the large statistical weights given to all of the variables. Furthermore, the drainage network condition and slope variation specify that the enclosing indicators have a significant impact on the GWPZ. Differences in sensitivity are caused by LULC, slope, drainage density, precipitation elevation, and, soil composition. Their respective mean values are 0.72, 1.0, 0.96, 0.73, 0.62, 0.46, 1.27, 0.62, 0.62 and 2.78. The percentage area of each GWPZs from the very low to very high GWPZs varies when each element is removed from the evaluation. Drainage density and slope are important influencing elements in defining GWPZs, since removing them raises the “Very high” GWPZ area by 83.5 and 80.4%, respectively (Table 8). In the same way, excluding soil type raises the “high” GWPZ region by 32.7%. Furthermore, the “Moderate” and “Low” GWPZ regions appear to be responsive to TWI omission, as their deduction expands the zone of these two variables by 38.5 and 70.6%, accordingly provided in Table 8.

Table 7 Sensitivity analysis through the elimination of parameters (one issue is eliminated at each time)
Full size table
Table 8 Changes in zone category under the exclusion of each factor
Full size table

Such studies can be used to examine possible groundwater zones in coastal regions as well as other parts of the world. For example, if the goal is to define GWPZ, a novel GWPZ methodology may be produced based on the geological and environmental conditions of any selected sites. After that, the current GWPZ technique may be compared to the proposed GWPZ methodology's baseline. Furthermore, based on a specific area the mechanism of choosing variables, giving weightage according to the importance may be updated.

Conclusions

A precise, scientific, and systemic approach has been at Sagar Island in India, except for real groundwater reservation, was used efficiently to determine and create GWPZ maps using the FAHP model and the RS–GIS methodology. However, using the FAHP model in combination with the RS–GIS method and seems that the outputs of this study will be pretty much useful in mapping groundwater potential zones. In comparison to previous intrusive approaches, this technology provides reliable computable information on groundwater resources economically. To create the final GWPZ map, ten factors/thematic layers (Elevation, slope, rainfall, drainage intensity, curvature, TWI, distance from creek, LULC, and soil quality) were added into the model with varied weights. Each element was divided into several classifications and characteristics that influence groundwater potentiality. To give weights to each component and normalize ratings, the AHP and fuzzy models were used. The GWPZ mapping was prepared using normalized data and the Kriging feature interpolation method. The estimated groundwater potential zones has been classified into five groups, distributed between extremely high storage to extremely low. According to the GWPZ map used in this study, 16.03 km2 (6.84%) area has been classified as having a very good potential of groundwater, 83.69 km2 (36.23 percent) area as having good potential, and the remaining 63.94 km2 (27.68%), 40.49 km2 (17.53%), and 27.07 km2 (11.72%) areas are categorized having moderate, poor, and very poor potential. Alluvial deposits, on the other hand, provide very significant potential or storage of groundwater. That is why the categories, very high and high GWPZ indicate the alluvial plains in most of the cases which have excellent groundwater storage. According to the results and output, the majority of the Island has a potential or storage of groundwater classified in between high to moderate. According to the study, the northern and the central part of the Sagar island have very high groundwater possibility zone, due to the presence of sedimentary/alluvial soil placement and flatter slopes with effective penetration capacity.

According to the findings of this study most areas with flat slopes and appropriate rainfall patterns have a high groundwater potential. GWPZ has also been compared to VES points, and the FAHP model in combination with the GIS–RS method was found in concordance with the field survey data, demonstrating its validity. The model developed also goes through the sensitivity analysis to get a better understanding of the justification of each element. A sensitivity analysis disclosed that vegetation and LULC had a substantial influence on the GWPZ assessment. To determine the yields of the groundwater from various wells in the areas with different GWPZs, the technique described through this study should be confirmed by comparing the data available on depth, discharge, and step-down drilling of different wells in various locations throughout Sagar Island. This approach might be used to find sustainable zones having an adequate amount of groundwater. The groundwater potential model is delicate in giving weights to components/parameters and their corresponding classes/properties, because the distribution technique of various weights is somewhat arbitrary. A future study might look at the other elements and characteristics that impact the discovery of new GWPZs, either directly or indirectly. This model approach and the mapping of potential zones of groundwater sources may suitable for initial assessment of the study area for local authorities and water policymakers when deciding where to bore freshwater. The relevance of this research is that it lays the groundwork for properly formulating a groundwater management strategy in Sagar Island, India. The identification and sustainable development of groundwater supply zones may assist in the balance of mining and allow future research on groundwater extraction. The method's flexibility, logical conditions, and general character can continue with or without changes in the weights of the given components and may thus be used in various places in India and around the world.