1 Introduction

Groundwater (subsurface water) plays a vital environmental role in continuous river flows, societies, and the ecosphere globally. Additionally, it is a immense source of fresh water on the planet (UNESCO, 2018). In arid and semi-arid areas, subsurface water is the prime available water resource for many purposes, including domestic and irrigation (Dimple et al., 2022). Of the freshwater available on the earth, nearly 30% is present in subsurface water, and < 1% is present in lakes and rivers. The CGWB (2017) report states that the total yearly groundwater recharge is 432 Billion Cubic Meters (BCM), and the yearly exploitable groundwater resources are 393 BCM. The yearly extraction of groundwater for all applications is 249 BCM, of which 221 BCM (89%) is used for agriculture and 25 BCM (10%) for domestic needs (CGWB, 2017). Recent studies have reported that approximately 20% of globally groundwater is used for the purposes of irrigation (Adimalla & Venkatayogi, 2018; Saeid et al., 2018). Almost 2 billion people depend on groundwater for drinking water (Abe & Ersado, 2022). Because it has a greater potential for availability and is less likely to become contaminated than surface water, groundwater is an important source of water supplies in all climatic regions. This includes both urban and rural areas of both developed and developing nations. This is because climatic conditions threaten the reliability of surface water. Groundwater is an ever-changing resource influenced by various factors, including geomorphology, lithology, topography, slope, precipitation, soil, drainage pattern, LULC, and hydrological conditions (Acharya, 2017; Murmu et al., 2019; Singh et al., 2011). Several regions of the country’s water level have dropped quickly during the last two decades (Jasrotia et al., 2016).

Traditional groundwater research techniques, such as drilling, geophysical, geological, and hydro-geological approaches, are expensive and time-consuming, necessitating the deployment of large human and financial resources. Groundwater potential mapping (GPM) is a cost-effective method for identifying spatial differences in groundwater recharge potential with fewer sources (Arabameri et al., 2019), whereas nowadays RS and GIS are important tools that may be used to predict groundwater resources at a minimal cost quickly and can be used efficiently for groundwater research before resorting to more extensive and costly surveying procedures (Priya et al., 2022). Groundwater potential zone development is typically influenced by a variety of factors, including denudation, structural characteristics, terrain, drainage, climate, and the type, thickness and structural makeup of the underlying rocks. The effect of these factors on groundwater recharge varies by location. Linear structures such as fractures and faults, for example, can operate as conduits or barriers for groundwater movement. As a result, they are powerful variables in determining potential recharge zones. Groundwater recharging often favors buffer zones of 300–500 m near the lineament (Ahmed et al., 2021). However, these aspects must be considered to outline and identify an area’s groundwater recharge potential (GWRP).

Presently, groundwater is decreased at a rate of 800 km3 per year globally. Irrigated agricultural lands are a significant source of groundwater depletion (Burek et al., 2016). Due to a lack of surface water, water scarcity in arid and semi-arid areas has considerably grown (Castillo et al., 2022). According to studies, groundwater resources offer more than 70% of the water supply (Sahu et al., 2022; Suliman et al., 2022) and are overexploited, with a rate of depletion of 545 km3/year (Makonyo & Msabi, 2021). The estimated annual groundwater usage is almost 230 km3, and India is a largest user of groundwater globally (World Bank, 2010). In addition, more than 60% of the groundwater is used for irrigation, and 85% is used for domestic supplies. Due to this, groundwater levels have declined by more than 4 m in different parts of the country (CGWB, 2014). Groundwater used for irrigation is approximately 245 × 109 m3 in India (CGWB, 2014). Groundwater productivity potential in semi-arid hard rock aquifers remains restrained by shallow, fractured, and weathered formations (Machiwal & Jha, 2014). Because of the consequences of climate change and population increase, among other variables, which impose severe stress on accessible surface water sources in these locations, groundwater supplies provide the majority of household, agricultural and industrial water demands in these hard rock terrains (Edmunds et al., 2003). Shallow hard rock aquifers account for roughly 73% (2.386 million \({\text{km}}^{2}\)) of India’s total land area (Machiwal & Singh, 2015).

Nowadays, the integration of RS and GIS techniques is practiced as a faster, more precise and more cost-efficient way to identify the many factors relevant to the groundwater potential zones (GWPz) (Fagbohun, 2018; Lentswe & Molwalefhe, 2020; Vishwakarma et al., 2021). Additionally, RS and GIS technology development has supported the delineation of groundwater potential zones in vast regions quickly. These methods have become very useful and cost-effective through high-resolution satellite images, mostly in inaccessible locations. Many researchers have reported many studies on using geographic information systems in groundwater management and monitoring, like delineation of GWPz (Bera et al., 2020; Bhattacharya et al., 2020; Das et al., 2017; Kumar et al., 2016; Machiwal & Singh, 2015; Pande et al., 2019; Rajasekhar et al., 2020). Most of these studies have relied on integrating the relative weights of various thematic layers, such as geology, geomorphology, lineament density, drainage density, post-monsoon water table depth, relative slope position (RSP), topographic wetness index (TWI), slope, etc., within a GIS environment (Ahmed et al., 2021; Akbari et al., 2021; Yeh et al., 2009). Many multi-criteria decision analysis (MCDA) models are used by researchers for groundwater potential zoning like multi-influencing factors (MIF) (Ahmed et al., 2021), full consistency method, best–worst method (Akbari et al., 2021), analytical networking process (ANP) (Mir et al., 2021; Singha et al., 2019), fuzzy analytic hierarchy process (Chaudhry et al., 2019; Jesiya & Gopinath, 2020; Singh et al., 2021; Sresto et al., 2021), weight of evidence (WoE) and frequency ratio (FR) (Rane & Jayaraj, 2021); the AHP (Castillo et al., 2022) technique has the potential to determine groundwater prospect zones (Arulbalaji et al., 2019). Delineation of GWPz through proper modeling techniques is important to mitigate the water scarcity problem in the semi-arid region (Bera et al., 2020). No such studies have been done in the NSC regarding the delineation of GWPz mapping due to a lack of knowledge about the aquifer parameters, factors which influence the groundwater and integrated approach of RS, GIS, AHP, etc. Because agriculture is the primary occupation in this area, a precise groundwater potential map for this catchment is firmly needed. There is a need to map the aquifers to know the quantum and quality of groundwater resources. Therefore, the present work aimed to identify the GWPz in the NSC, using GIS and AHP techniques for managing and planning groundwater resources. Geospatial technology is a novel concept that integrates RS, GIS, and AHP, and it has proved to be an effective tool for delineating GWPz. AHP was introduced by (Saaty, 1980) to solve complex decision-making through pairwise comparisons. AHP is a helpful method for prospecting GWPZs (Sahu et al., 2022; Sajil-Kumar et al., 2022). Shekhar and Pandey (2015) reported that utilizing AHP in combination with GIS is a time-saving and effective way of managing spatial data. Various researchers have exercised AHP to identify the groundwater potential zones by calculating the weights of various TLs and their respective classes (Machiwal et al., 2011; Manap et al., 2014; Rahmati et al., 2015; Singh et al., 2018). The literature suggests that GIS and AHP techniques could be utilized for groundwater potential zoning studies. The study’s novelty was the evaluation of field datasets by combining multi-criteria decision problems to delineate GWPz in the catchment. Therefore, the current study utilized geospatial tools and the AHP method to identify groundwater potential sites in the Nand Samand Catchment to improve groundwater resource utilization. Understanding this valuable water resource is critical to ensuring its long-term development and management. The research may contribute to creating, planning, and managing sustainable groundwater resources.

1.1 Study area

Nand Samand Dam is located in Rajsamand district of Rajasthan state, India, between 24°0′0.5″ and 26°0′0.5″ N and 72°59′59.50″ and 73°59′59.50″ E coordinates. The location map is shown in Fig. 1. The research area comes under an arid to semi-arid climate with an annual average rainfall (2001–2011) of 640.45 mm (CGWB, 2013). The district’s central and eastern areas are largely flat, forming the foothill of the Aravalli ranges. The general slope of the terrain is toward the east. The main river of the district is Banas, with its tributaries, i.e., Khari and Chandrabhaga generating an excellent drainage system in the area (GWDR, 2013). The total study area is 865.18 \({\text{km}}^{2}\), with a maximum elevation (1318 m) and the minimum elevation (570 m) above mean sea level.

Fig. 1
figure 1

Nand Samand Catchment

Full size image

2 Materials and methods

In this study, we have used primary and secondary data for mapping the GWPz of the study area. After reviewing published literature, ten criteria for geospatial mapping of groundwater potential zones were chosen for the current study. Thematic layers like well recharge, geomorphology, lineament density, slope, topographic elevation, transmissivity, post-monsoon water level, soil, rainfall and LULC were created using RS and conventional/existing data in ArcGIS 10.1 environment. Data description is summarized in Table 1. The current study created a digital elevation model (DEM) of the research area using Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) elevation data (30 m spatial resolution). The research area’s base map was created using SOI (Survey of India) topographic sheets (45G-12, 45H-5, 6, 9, 10, 13, 45NG-9) of 1:50,000 scale. These SOI topographic maps were georeferenced in ArcGIS 10.1 using the World Geodetic System (WGS) 84 datum and the Universal Transverse Mercator (UTM) zone 43N projection. Soil data of the study area were collected from the National Bureau of Soil Survey (NBSS) and Land Use Planning (LUP), Udaipur, Rajasthan. The unsupervised classification was used to extract the LULC categories within the catchment. Digitally processed raw satellite images included geometric correction, image enlargement, image interpretation and multispectral categorization. Additionally, unsupervised image classification was used to categorize land use types. Satellite data were analyzed using visual interpretation techniques to identify and describe various objects associated with urban areas. After detailed analysis, the result was compared and corrected by data collected from different locations of the area. The methodology developed is given in Fig. 2. The Landsat image was used to map the lineament density, and the DEM was used to generate the hillshade map. Figure 3 depicts the study’s methodological flowchart.

Table 1 Data collected to identify GWPz in this work are detailed
Full size table
Fig. 2
figure 2

Methodology flow diagram of the present study

Full size image
Fig. 3
figure 3

Unsupervised classification of Landsat image

Full size image

2.1 AHP weight assignment and normalization

The MCDA methodology for delineating the Nand Samand Catchment GWPz was produced by integrating various GIS tools and the AHP method. The most significant aspect of the integrated analysis is the weight assigned to all classes. As a result, it is highly dependent on the assignment of suitable weight (Murmu et al., 2019). The current research employed the AHP approach, which Saaty (1980) introduced. The benefit of hierarchy is that it enables us to focus our judgment independently on each of the numerous attributes required to make solid decisions (Saaty, 1990). The relative importance values for all thematic layers and their associated characteristics were calculated using a standard Saaty’s 1–9 scale, where ‘1’ indicates “Equal Importance (EI)” for two themes and ‘9’ indicates “extreme importance” for one theme relative to the other TLs, as presented in Table 2.

Table 2 Saaty’s scale (26)
Full size table

The following procedure was exercised to calculate the final weights assigned to each theme layer (Murmu et al., 2019):

  1. (1)

    Using Eq. (1), add the data in the individual column of the pairwise matrix (PWM) (Table 3).

    $$L_{j} = \mathop \sum \limits_{i,j = 1}^{n} C_{ij}$$
    (1)

    where \(L_{j}\) = the pairwise matrix’s total data in each column and \(C_{ij}\) = at the ith row and jth column, the number assigned to each condition.

Table 3 Pairwise comparison matrix of ten thematic layers
Full size table
  1. (2)

    To construct a normalized PWM, divide each element in the matrix by the sum of its columns (Table 4)

    $$X_{ij} = \frac{{C_{ij} }}{{L_{j} }}$$
    (2)

    where \(X_{ij}\) = the value that appears in the ith row and jth column of the normalized PWM.

  1. (3)

    Using the following equation, divide the sum of the normalized row of the matrix by the number of criteria used to calculate standard weights (N) to obtain the normalized row of the matrix.

    $$W_{i} = \frac{{\sum X_{ij} }}{N}$$
    (3)

    where \(W_{i}\) = standard weight.

Table 4 Normalized pairwise matrix and weight of each thematic layer
Full size table

2.2 Consistency analysis

Multiply the consistency vector by the pairwise comparison matrix values of theme layers (Murmu et al., 2019).

  1. (4)

    Principal eigenvalue (\(\lambda_{{{\text{max}}}}\)) computed by the following formula

    $$\lambda_{\max } = \sum \left( {C_{ij} X_{ij} } \right)$$
    (4)
  2. (5)

    The consistency ratio (CR) was computed as (Saaty, 1980):

    $$C_{{{\text{ratio}}}} = \frac{{C_{{{\text{index}}}} }}{{C_{{{\text{random}}}} }}$$
    (5)

    where \(C_{{{\text{random}}}}\) = random consistency index (RI), the values of which were acquired from Saaty’s standard table (1980).

  1. (6)

    The following formula was used to calculate the consistency index (CI):

    $$C_{{{\text{index}}}} = \frac{{\lambda_{\max } - 1}}{n - 1}$$
    (6)

    where n = number of thematic variables.

If the CR value is ≤ 0.1, the inconsistency is permissible; otherwise, the subjective assessment must be revised (Saaty, 1990). Table 5 shows the RI values for n criterion (Murmu et al., 2019).

Table 5 Random inconsistency values
Full size table

2.3 Determination of groundwater potential zones

The groundwater potential index (GWPI) is a dimensionless number that aids in defining areas with high groundwater potential (Kumar & Krishna, 2018). The total weights of various polygons in the integrated layer were calculated using the following equation.

$${\text{GWPI}} = \left( \begin{gathered} {\text{GM}}_{{\text{w}}} {\text{GM}}_{{{\text{wi}}}} + S_{{\text{w}}} S_{{{\text{wi}}}} + {\text{SL}}_{{\text{w}}} {\text{SL}}_{{{\text{wi}}}} + {\text{TE}}_{{\text{w}}} {\text{TE}}_{{{\text{wi}}}} + {\text{LU}}_{{\text{w}}} {\text{LU}}_{{{\text{wi}}}} + \hfill \\ {\text{WD}}_{{\text{w}}} {\text{WD}}_{{{\text{wi}}}} + {\text{RE}}_{{\text{w}}} {\text{RE}}_{{{\text{wi}}}} + {\text{TR}}_{{\text{w}}} {\text{TR}}_{{{\text{wi}}}} + {\text{RF}}_{{\text{w}}} {\text{RF}}_{{{\text{wi}}}} + {\text{Ld}}_{{\text{w}}} {\text{Ld}}_{{{\text{wi}}}} \hfill \\ \end{gathered} \right)$$

where GM denotes the geomorphology, S for soil type, SL for slope, TE for topographic elevation, LU for land use, WD for post-monsoon groundwater depth, RE for groundwater recharge, TR for transmissivity, RF for rainfall, Ld for lineament density, w for the normalized weight of a theme and wi for the normalized weight of the individual features of a theme.

2.4 Validation of groundwater potential zones

Before explaining a model, whether stochastic or deterministic, it must first be validated. The model’s scientific significance is determined by its validation (Al-Manmi et al., 2021). Eventually, to assess the accuracy of the proposed technique, the existing wells location map in the study has been used. Sixteen-well yield data resulting from the pumping test have been used. To verify the accuracy of the current work in various groundwater potential zones, the well yield points were superimposed on the final groundwater potential zones map. Many studies have used a similar validation method (Ahmed et al., 2021; Akbari et al., 2021; Al-Manmi et al., 2021; Arulbalaji et al., 2019; Priya et al., 2022; Trabelsi et al., 2022). Also, the ROC method (Castillo et al., 2022; Jhariya et al., 2021; Kaewdum & Chotpantarat, 2021; Melese & Belay, 2022; Pandey et al., 2022) was used for the validation of the AHP technique.

3 Results and discussion

3.1 Hydrologic TLs

Groundwater occurrence and movement are influenced by many factors, including geological structure, topography, slope, rainfall, soil, soil porosity, drainage density and pattern, LULC, and hydrological conditions of the area, as well as the interdependence of these parameters (Murmu et al., 2019). Groundwater in the hard rocks (consolidated) occurs under water table conditions in the zone of weathering and fracturing, joints, fissures and bedding planes; correct identification of the GWPz is important to prevent economic loss, wasting time and effort. In the present study, ten thematic layers were identified as the influencing factors, such as recharge, geomorphology, lineament density, slope, topographic elevation, transmissivity, post-monsoon water level, soil, rainfall and LULC to delineate the GWPz. The number of thematic layers selection depends on the data available in a given area. Each theme layer has specific information about the occurrence of groundwater. The groundwater prospect map was created by merging ten themed maps into GIS environment, utilizing remote sensing and conventional data. Out of these thematic maps, ASTER data were used to generate topographic elevation, lineament density and slope maps, whereas conventional data were used to generate soil map. Thematic maps were prepared in the ArcGIS environment to understand baseline information of different groundwater-related parameters. Each factor evaluating the GWPz in the study area is classified into four classes. These thematic maps were converted to raster data sets with the same pixel size (resolution), and different weightage was assigned using the AHP method. Each map was reclassified based on the weight values generated.

3.1.1 Geomorphology

The geomorphology of an area consists of many landforms and structural elements that are important in determining groundwater potential and prospects because they influence groundwater movement. In the present study, analog-type districts maps were extracted from the Geology survey of India (https://bhukosh.gsi.gov.in/Bhukosh/Public) and clipped the selected study area geomorphology in the ArcGIS software to generate a thematic layer. The geomorphological features of the research area can be divided into four categories, which are as follows: (i) waterbodies, (ii) alluvial plain, (iii) denudational hills and valleys, and (iv) structural hills (Fig. 4). Waterbodies encompassing an area of 16.49 km2 (1.91% of the total area), alluvial plain encompassing an area of 14.35 km2 (1.66% of the total area), denudational hills and valleys encompassing an area of 460.01 km2 (53.17% of the total area), and structural hills encompassing an area of 374.32 km2 (43.27% of the total area). Denudational hills and valleys dominate the study area. Denudational origins are characterized by gently sloping plains covered in weathered material conducive to groundwater recharge (Murmu et al., 2019). Authors have used geomorphology as one of the most influencing factors in the GWPz (Anand et al., 2021; Arulbalaji et al., 2019). The maximum weight is given for valley fill, water bodies, coastal plain of fluvial origin and minimum weight assigned to lower lateritic plateau and denudational hills (Arulbalaji et al., 2019). Table 6 shows the final geomorphological feature weights. The weights allocated to all the thematic layers and the AHP procedure were used to derive the normalized weights as given in Tables 3 and 4. Consistency analysis calculation:

$$\begin{gathered} 1 * 0.30\, + \,2 * 0.15\, + \,3 * 0.10\, + \,3.5 * 0.09\, + \,4 * 0.08 \hfill \\ + \,4.5 * 0.07\, + \,5 * 0.06\, + \,5.5 * 0.06\, + \,6 * 0.05\, + \,7 * 0.04\, = \,3.05. \hfill \\ 0.5 * 0.39\, + \,1 * 0.19\, + \,2 * 0.10\, + \,3 * 0.06\, + \,3.5 * 0.06 \hfill \\ + \,4 * 0.05\, + \,4.5 * 0.04\, + \,5 * 0.04\, + \,5.5 * 0.04\, + \,6 * 0.03\, = \,1.95. \hfill \\ 0.33 * 0.38\, + \,0.5 * 0.25\, + \,1 * 0.13\, + \,2 * 0.06 \hfill \\ + \,3 * 0.04\, + \,3.5 * 0.04\, + \,4 * 0.03\, + \,4.5 * 0.03\, + \,5 * 0.03\, + \,5.5 * 0.02\, = \,1.25. \hfill \\ 0.29 * 0.31\, + \,0.33 * 0.27\, + \,0.5 * 0.18\, + \,1 * 0.09\, + \,2 * 0.04\, \hfill \\ + \,3 * 0.03\, + \,3.5 * 0.03\, + \,4 * 0.02\, + \,4.5 * 0.02\, + \,5 * 0.02\, = \,0.89. \hfill \\ 0.25 * 0.27\, + \,0.29 * 0.23\, + \,0.33 * 0.20\, + \,0.5 * 0.13\, + \,1 * 0.07\, \hfill \\ + \,2 * 0.03\, + \,3 * 0.02\, + \,3.5 * 0.02\, + \,4 * 0.02\, + \,4.5\, + \,0.01\, = \,0.66. \hfill \\ 0.22 * 0.23\, + \,0.25 * 0.21\, + \,0.29 * 0.18\, + \,0.33 * 0.15\, + \,0.50 * 0.10\, \hfill \\ + \,1 * 0.05\, + \,2 * 0.03\, + \,3 * 0.02\, + \,3.5 * 0.01\, + \,4 * 0.01\, = \,0.52. \hfill \\ 0.20 * 0.21\, + \,0.22 * 0.19\, + \,0.25 * 0.17\, + \,0.29 * 0.15\, + \,0.33 * 0.12\, \hfill \\ + \,0.50 * 0.08\, + \,1 * 0.04\, + \,2 * 0.02\, + \,3 * 0.01\, + \,3.5 * 0.01\, = \,0.41. \hfill \\ 0.18 * 0.19\, + \,0.20 * 0.17\, + \,0.22 * 0.15\, + \,0.25 * 0.14\, + \,0.29 * 0.12\, \hfill \\ + \,0.33 * 0.10\, + \,0.50 * 0.07\, + \,1 * 0.03\, + \,2 * 0.02\, + \,3 * 0.01\, = \,0.34. \hfill \\ 0.17 * 0.17\, + \,0.18 * 0.16\, + \,0.20 * 0.14\, + \,0.22 * 0.13\, + \,0.25 * 0.11\, \hfill \\ + \,0.29 * 0.10\, + \,0.33 * 0.09\, + \,0.50 * 0.06\, + \,1 * 0.03\, + \,2 * 0.01\, = \,0.29. \hfill \\ 0.14 * 0.17\, + \,0.17 * 0.14\, + \,0.18 * 0.13\, + \,0.20 * 0.12\, + \,0.22 * 0.11\, \hfill \\ + \,0.25 * 0.10\, + \,0.29 * 0.08\, + \,0.33 * 0.07\, + \,0.50 * 0.05\, + \,1 * 0.02\, = \,0.24. \hfill \\ \end{gathered}$$
Fig. 4
figure 4

Thematic layer of geomorphology

Full size image
Table 6 Normalized weights of all thematic layer classes for groundwater potential zoning
Full size table

The consistency vector (\(\lambda_{{{\text{max}}}}\)) value is determined as 9.59. From Eqs. (5) and (6), consistency ratio was determined as (− 0.03) and found acceptable.

3.1.2 Slope

The slope significantly defines groundwater’s presence and recharge situations in a particular region (Bagyaraj et al., 2013). Slope gradient directly impacts rainfall infiltration and is thus an indicator of appropriateness for groundwater prospecting. Because of the high surface run-off, the steeper the slope, the lesser will be recharge, allowing inadequate time for the water to infiltrate through the surface and replenish the saturated zone (Ghosh et al., 2016).

The slope of the study area was classified into five categories as very low (0–3%), low (3–8%), moderate (8–15%), high (15–30%) and very high (> 30%) presented in Fig. 5. The highest weight is assigned for flat and gentle slopes, and the lowest weight is assigned for the steep and very steep slope. Arulbalaji et al. (2019) also similarly assigned the weight to the slope layer. The majority of the research area falls in the 3–8 percent slope class (35.16% of the total area). As a result, the maximum portion of the study area has a sloping slope which is favorable for water retention. Using a scale of 1–9, the ratings were assigned to distinct slope classes. Higher ranks were given to slope classes with a smaller percentage of slopes due to the flat terrain’s ability to retain more groundwater, while lower ranks were assigned to steeper slopes with more run-off and less infiltration (Kumar & Krishna, 2018; Nag & Ghosh, 2013). Table 6 shows the final weights for the slope classes.

Fig. 5
figure 5

Slope map

Full size image

3.1.3 Soil

The capacity of a soil to hold and transmit soil moisture in profile depends on its texture, porosity and soil structure. For example, heavy soils provide effective capillary tubes for moisture movement due to their fine texture and porosity. Due to large particles and pores, with minimum fine capillary pores, there is minor soil moisture movement in loose sandy soils. Soil texture has a significant impact on groundwater recharging and surface run-off. The soil type, infiltration rate, percolation and permeability of the soil affect groundwater recharge (Jasrotia et al., 2016). The soil map of the study area was collected from the NBSS and LUP, at 1:250,000 scale. ArcGIS software was used to scan and digitize the soil map. The research area comprises four soil types: loamy skeletal, coarse loamy, fine loamy and rocks outcrops. It is seen from Fig. 6 that most of the research area is dominated by fine loamy to rock outcrops soil which is about 91.48% of the total research area. Loamy/loamy skeletal soils have a high positive association with groundwater recharge, having maximum weightage values, signifying maximum groundwater recharge potential (Bera et al., 2020; Doke et al., 2021), while rock outcrops soils were negatively associated with groundwater recharge; therefore, the lesser score is assigned. Table 6 shows the final weights for the soil classes.

Fig. 6
figure 6

Thematic layer of soil

Full size image

3.1.4 Topographic elevation

The DEM of the present research area was generated using Advanced Spaceborne Thermal Emission and Reflection Radiometer-30 m (ASTER) data (Source: https://earthexplorer.usgs.gov). The topographic data are available in the GeoTIFF raster file format with a pixel resolution of 30 m. The topography file was imported to ArcGIS, and DEM was created. The most effective recharge site would be lower than the river’s level, allowing the water to be easily diverted by gravity. Given that topography (land surface height) is one of the factors affecting groundwater potential, the present study included a topographical elevation map as one of the TLs. The higher is elevation, the lesser is groundwater available. Figure 7 depicts the topographic map. The study area has been classified into five topographic elevation classes based on its topographic elevation: (i) 570–690 m, (ii) 690–794 m, (iii) 794–890 m, (iv) 890–1000 m and (v) 1000–1318 m. Northern areas of the research area have the highest topographic elevations, which results in the most run-off and hence reduced potential of rainwater infiltration. The study area is dominated by topographic elevations 794–890 m in 316.42 km2 areas, which is about 36.57% of the total area.

Fig. 7
figure 7

Topographic elevation map

Full size image

3.1.5 Land Use/Land Cover (LULC)

According to Shaban et al. (2006), plant cover promotes groundwater recharge in various ways, containing biological decomposition of the plants roots, which offers a conduit for water to percolate onto the surface by loosening the rock and soil. Also, the plants act as barrier and retards the surface run-off velocity, which gives more infiltration opportunity time to the running water and thus more groundwater recharge. A LULC thematic layer was distinguished into five different classes: (i) waterbodies, (ii) cropland, (iii) built-up area, (iv) barren land and (v) shrub as shown in Fig. 8. Different LULC cover classes were ranked according to their importance in groundwater recharge potential. The LULC classes, like forest and agricultural land, hold substantially more water than the built-up land, barren land and rocky surfaces (Arulbalaji et al., 2019). Crop areas were given a higher rating than other land uses since they require more irrigation water, which results in groundwater recharge. Because of the shallow soil, the surface run-off rate is higher on barren ground. Hence, it receives a lower score. Because the built-up terrain is concrete, the surface run-off is high, and the infiltration rate is low. As a result, built-up land is assigned the lowest score (Doke et al., 2021).

Fig. 8
figure 8

LULC map

Full size image

3.1.6 Post-monsoon groundwater depth (WD)

The WD in the research area ranges from 0.27 m Below Ground Surface (bgs) to 9.2 m bgs. Groundwater at maximum depths was assigned a higher weight. Figure 9 depicts WD map of the research area, revealing that the groundwater depth < 4 m bgs in 91.54%, 4–8 m bgs in 8.33% and 8–12 m bgs in 0.13% of the total research area. The deeper groundwater levels in the study area’s north, east parts, and the small and scattered patches are a result of excessive groundwater pumping for irrigation. Figure 9 shows that less variance occurs during the post-monsoon season because the underlying hard rock aquifer system has a low specific yield, allowing water levels to rise fast in response to rainy-season recharge (Machiwal et al., 2017).

Fig. 9
figure 9

Post-monsoon groundwater depth map

Full size image

3.1.7 Recharge

The study area’s annual net groundwater recharge varies between 0.27 and 42.27 cm. The recharge map was created using the Inverse Distance Weighted Moving Average (IDWMA) technique with point recharge values (95 locations). The recharge area has been divided into five recharging zones based on these recharge estimates: (i) < 10 cm/year (yr), (ii) 10–15 cm/yr, (iii) 15–20 cm/yr (iv) 20–30 cm/yr and (v) > 30 cm/yr as seen from Fig. 10. Figure 10 indicates that a net recharge rate of < 10 cm/year is predominant in the research area, covering approximately 50.05% of the study area. Net recharging at a rate of between 10 and 15 cm/year corresponds to approximately 25.32% of the study area. The northeast part of the research area has a maximum recharge rate (> 30 cm/yr). Increased rainfall in any given area equates to increased recharge possibilities (Machiwal et al., 2011, 2017; Murmu et al., 2019).

Fig. 10
figure 10

Recharge map

Full size image

3.1.8 Transmissivity

Transmissivity data were generated by conducting pumping test in the study area and were used to generate a point map of transmissivity. A transmissivity raster map was created after interpolating this point map using the interpolation technique, viz. Inverse Distance Weighting (IDW) technique in the ArcGIS environment. It is found that the transmissivity in the present study area ranges from 105.10 to 321.90 \({\text{m}}^{2} /{\text{day}}\). The area was classified into seven transmissivity classes based on these values, namely extremely poor (105–136 \({\text{m}}^{2} /{\text{day}}\)), very poor (136–167 \({\text{m}}^{2} /{\text{day}}\)), very poor (167–198 \({\text{m}}^{2} /{\text{day}}\)), poor (198–229 \({\text{m}}^{2} /{\text{day}}\)), poor (229–260 \({\text{m}}^{2} /{\text{day}}\)), moderate (260–291 \({\text{m}}^{2} /{\text{day}}\)) and good (291–322 \({\text{m}}^{2} /{\text{day}}\)) as given in Fig. 11. The transmissivity map shows that a low transmissivity (105–136 \({\text{m}}^{2} /{\text{day}}\)) is confined to small patches in the Northwest parts of the research area. The dominance of transmissivity from 167 to 198 \({\text{m}}^{2} /{\text{day}}\) is maximum, which is about 32.19% of the total research area. Anand et al. (2021) used transmissivity as one of the thematic layers in the Bhavani River basin; the transmissivity of the ranges from 48.14 to 455.42 m2/day.

Fig. 11
figure 11

Transmissivity map

Full size image

3.1.9 Rainfall

Rainfall makes a significant contribution to groundwater recharge. In arid areas, rainfall is considered as one of the most influential factors in groundwater recharge capability (Ahmed et al., 2019). More rainfall in any given location provides more chances of recharge (Machiwal et al., 2011). In this study, 10-years (2011–2020) rainfall data were taken (https://crudata.uea.ac.uk/cru/data/hrg/) to make a rainfall map of the research area (Fig. 12; Table 6). The spatial distribution map of rainfall was prepared using IDW interpolation method. The average annual rainfall in the research area is divided into five categories, i.e., very low (585.43–603.86 mm), low (603.86–615.70 mm), moderate (615.70–625.97 mm), high (625.970–636.50 mm) and very high (636.50–652.56 mm). The maximum intensity of rainfall (636.50–652.56 mm) was found in the southern most part of the catchment area. Maximum intensity rainfall regions have a high weightage value, which covers approximately 134.21 \({\text{km}}^{2}\) (15.51%) area, signifying excellent groundwater potential (Bera et al., 2020). Rain with a high intensity and a short duration results in less infiltration and more surface run-off; rain with a low intensity and a long duration causes more infiltration than run-off (Arulbalaji et al., 2019).

Fig. 12
figure 12

Thematic map of rainfall distribution

Full size image

3.1.10 Lineament density (Ld)

Lineament characteristics like joints, cracks, and faults are particularly important hydrogeologically because they operate like a conduit for groundwater occurrence, resulting in greater porosity, and so serving as a GWPz (Mukherjee et al., 2012). Lineaments have the potential to play an important role in GWPz by maximizing soil infiltration capacity and ease to groundwater flow and movement (Das & Pardeshi, 2018). Lineament density and drainage density are inversely proportional to each other. Good GWPz are available where lineament density is maximum and drainage density is minimum with gentle slope (Ahmed et al., 2021; Arulbalaji et al., 2019). The research area’s lineament orientation is primarily along the North East–South West and North West–South East directions, as seen by a “rose diagram” (Fig. 13). As reported by Harinarayana et al. (2000), the normalized transmissivity around the lineaments is maximum, and there is a strong correlation between increased fracture density and higher well yields. As a result, locations with maximum Ld have a greater effect on GWPz. The Ld of the catchment ranges from 0 to 1.28 \({\text{km}}/{\text{km}}^{2}\) (Fig. 14) and the Ld was regrouped into four classes, viz., very high (0.96–1.28 \({\text{km}}/{\text{km}}^{2}\)), high (0.64–0.96 \({\text{km}}/{\text{km}}^{2}\)), moderate (0.32–0.64 \({\text{km}}/{\text{km}}^{2}\)) and low (0–0.32 \({\text{km}}/{\text{km}}^{2}\)). Groundwater development has a significant potential in areas with extremely high to high lineament density (Bera et al., 2020). Table 6 contains the final weights for the Ld.

Fig. 13
figure 13

Rose diagram of lineaments of the study area

Full size image
Fig. 14
figure 14

Lineament density map of the research area

Full size image

3.2 Delineation of GWPz

A better understanding of groundwater potential is important for planning and long-term sustainable development. This information is vital for designing and implementing corrective measures to improve groundwater recharge processes. The parameters considered here are lineament density, well recharge, geomorphology, soil, topographic elevation, slope, rainfall, transmissivity, the post-monsoon water level in the well and land use/land cover (LULC). Groundwater potential zones in the Nand Samand Catchment have been generated using the weighted overlay approach in ArcGIS environment. The resulting map is divided into ‘good (G),’ ‘moderate (M),’ ‘poor (P)’ and ‘very poor (VP)’ GWPz. Following the computation of the final normal weights of all thematic layers and their features, thematic layers were transformed to a raster format and then concatenated together with the help of the raster calculator in ArcGIS environment. After computing the final weights of all the TLs and their related features, thematic layers were transformed to a raster format and then concatenated together with the help of the raster calculator in ArcGIS software to produce GWPz. The total score calculated for the GWPz (Fig. 15a, b) was categorized into four zones: ‘G,’ ‘M,’ ‘P’ and ‘VP’ groundwater potential. The research findings revealed that the area covered by ‘good’ GWPz is about 21.23%. The southeast portion and some small patches of western side of the study area have ‘M’ groundwater potential, and it occupies 53.18% area. About 23.76% area shows poor GWPz. 1.82% of the study region has very poor groundwater potential. As seen from Fig. 15a, the poor and very poor GWPz occur in the steep slope, high drainage density and reserved forests. The findings from this study are comparable to the findings of other groundwater potential studies in many regions globally (Ahmed et al., 2021; Akbari et al., 2021; Al-Manmi et al., 2021; Arulbalaji et al., 2019; Castillo et al., 2022; Hagos & Andualem, 2021; Owolabi et al., 2020). Jhariya et al. (2021) used the AHP method to assess GWPz in the Raipur city by using nine thematic layers and revealed that the medium–high class has maximum percentage (33.92%) of area of potential zones. Pande et al. (2019) applied an integrated approach for the delineation of GWPz of Devdari watershed of Maharashtra. The study results revealed that the good zones are in 9.51 km2 and very good 14.665 km2 area of the total area of the watershed. Raju et al. (2019) integrated RS, GIS and multi-influence factor (MIF) techniques for predicting GWPz in Mandavi River basin. Result of the study showed the good class of potential zone having area of 21% out of total study area. Rane and Jayaraj (2021) studied the multi-influence factor (MIF), weight of evidence (WofE) and frequency ratio (FR) technique to evaluate groundwater potential zones in the Kadva river, a tributary of Godavari River. The results of validation showed that the groundwater potential delineated using MIF technique has a prediction accuracy of 81.94%, followed by WofE technique (76.19%) and FR techniques (71.43%). Priya et al. (2022) employed eight thematic layers for the delineation of groundwater potential zones using AHP technique in the Mymensingh district. The results showed that 11.51% of the study area is under a very high groundwater potential zone. Pandey et al. (2022) conducted a study in the district Mahoba, Uttar Pradesh, India, for the GWPz identification by using AHP and Dempster–Shafer model. Results revealed that the northwest side of the study area is characterized with very high potential zones. The findings of this study were comparable to those of other studies conducted in various parts of India. As a result, the findings of this study can be used as a guideline for further investigation of groundwater development projects.

Fig. 15
figure 15

a Groundwater potential zones and well locations of the research area b groundwater potential classes

Full size image

Validation of the delineated GWPz by using AHP method has been performed by using the existing 16 well discharge data (Table 7) of the study area (Fig. 15a) and ROC method. The discharge data of the wells are grouped into high (> 6.06 lps), medium (3.03–6.06 lps) and low yield (< 3.03 lps). Table 7 shows the actual yield data obtained from the pumping test, as well as the acceptance/rejection of values that represent open well deviation yield data between expected/real in the form of the agreement. The accuracy prediction [(13/16) × 100 = 81.25%] shows that the selected technique used in this study is considerably reliable and accurate. Such a validation method was used by many researchers like, Pande et al. (2019), Gebru et al. (2020), Al-Manmi et al. (2021), Hibi et al. (2021), Jhariya et al. (2021) and Priya et al. (2022). In the study, there are 16 samples; the greater the AUC, the more accurately the model predicts 0 classes as 0 and 1 classes as 1. The ROC plot obtained is depicted in Fig. 16. The area under the ROC curve is found 0.889 (88.9%) at a significance value of < 0.001. As a result, the high value of AUC validates the suggested approach’s efficacy in predicting low and middle/high yield locations. Jhariya et al. (2021) found the ROC curve (AUC) 0.857 (85.7%). Melese and Belay (2022) found the AUC = 82.9% which shows the very good accuracy of the model. Similarly, Castillo et al. (2022), Kaewdum and Chotpantarat (2021) and Pandey et al. (2022) also used the ROC method for the validation.

Table 7 Accuracy assessment of calculated groundwater potential map
Full size table
Fig. 16
figure 16

ROC curve of the validate result

Full size image

Trabelsi et al. (2022) used AHP technique for the delineation of groundwater potential zones in the Sfax Basin in eastern part of Tunisia. They used wells discharge data and curve trend of sensitivity classes theory for the validation of the applied technique in GWPz. The results showed that the excellent and good potentiality classes have high to very high productivity in the majority of cases.

4 Conclusions

The demand for fresh water is a serious concern, so delineating new GWPz cost-effectively is essential to meet freshwater demands. Many scientists have investigated various techniques for evaluating and forecasting the distribution of groundwater for land use planning and water resource management. The majority of agricultural fields rely on groundwater for irrigation. The study region used an integrated view combining geospatial tools and the AHP method to delineate the GWPz. The conclusive groundwater potential map was divided into 4 classes: ‘G,’ ‘M,’ ‘P’ and ‘VP.’ Results revealed that 21.23% out of the total research area comes under good GWPz, 53.19% of the area under the moderate zone, 23.77% area shows poor GWPz and 1.82% area shows very poor groundwater potential. According to the current study, the research area should focus on groundwater management and development activities. The overall findings indicated that geospatial technologies like RS, GIS and the AHP method have the potential to give a good platform for examining GWPz in hard rock regions and developing appropriate groundwater exploitation plans for many objectives. The integrated map may benefit various developmental activities, including ensuring sustainable groundwater development in the research region and prioritizing locations for water conservation schemes and programs.