An integrated application of remote sensing, geographic information system, and analytical hierarchy process for identification of potential groundwater recharge zones across Potohar Plateau, Pakistan Open Access

Water is a vital part of daily life and has become crucial to human civilization due to anthropogenic and climatic changes (Abuzied et al. 2015; AL-Shammari et al. 2021; Iqbal et al. 2022; Masood et al. 2023). But the changing climate and urbanization have a significant impact on surface water and groundwater. For instance, untimely and unpredictable rainfall patterns have given rise to intense weather events, leading to prolonged droughts and episodes of flooding, accurate estimation of these phenomena is imperative for effective risk management in agriculture and the preservation of life (Mahmood et al. 2022). Since the world's population is expanding, there is an increasing need for water. Nearly 30% of the freshwater is underground in the form of aquifers (Echogdali et al. 2023; Zaheer et al. 2024a). In both rural and urban regions, groundwater is a crucial natural resource for the reliable and affordable supply of water (Pande et al. 2017). Given the fast degradation of surface water resources in terms of both quantity and quality, groundwater has become one of the primary sources of freshwater worldwide (Balkhair & Ur Rahman, 2021; Zaheer et al., 2024b).

Of the country's total water reserves, groundwater in Pakistan makes up 40% (Rehman et al. 2019). 43% of groundwater is utilized for irrigation purposes (Qureshi 2020). Arid climate, deforestation, increasing population, unplanned use of irrigation water, poor water management policy, and urbanization resulted in the depletion of the groundwater table (Arshad et al. 2020; Mahmood et al. 2024). Pakistan is the world's fourth-largest groundwater extractor (Qureshi 2020). Its annual sustainable groundwater reserves are estimated to be worth 55 billion cubic meters (BCM) and its annual groundwater extraction is estimated to be 65 BCM (Qureshi 2020). Focusing on the Potohar Plateau region, the situation is not significantly different despite the growth in population, which was estimated to be around 12 million in the year 2017. The groundwater resources, however, are substantially unchanged throughout the 1990s. Because of the swift population growth, groundwater supplies have been overused. As a result of exorbitant removal, such as 1.7 m per year in Islamabad, the amount of groundwater in the Potohar region is diminishing at a concerning pace (Khan et al. 2020). Very few studies have been carried out on groundwater recharge (GWR) in Pakistan. Moreover, no study has been done on GWR potential sites in the Potohar Plateau. Groundwater depletion is a common issue in most of the parts of the country (Pakistan). GWR is much less than the use of groundwater. Given the concerns about declining groundwater levels and an expanding population, the Potohar region's decision to pursue sustainable groundwater management makes sense.

Markov GWR suitability map was created by integrating and manipulating parameter layers in ArcGIS 10.5. GWR suitable sites can simply be identified using the raster data. The weighted overlay tool was utilized to classify the recharge suitability map into five groups: (1) unsuitable, (2) poor, (3) moderate, (4) suitable, and (5) most suitable. The classification of the final suitability map into pre-specified five groups is individual. The basic idea beyond that categorization was to give several options to the water resource engineers, managers, and planners to delineate appropriate areas for GWR.

Groundwater recharging has emerged as one of the effective options to maintain the groundwater tables. Rainwater harvesting can help to improve the natural replenishment of groundwater. Similarly, in metropolitan locations with limited space for surface recharge structures, injection wells and bore well recharging are commonly used techniques. Check dams and percolation ponds allow water to infiltrate into the ground. Wetlands act as natural sponges, storing water during wet periods and gradually releasing it into the groundwater. Similarly, reconnecting rivers to their floodplains allows floodwaters to spread over a larger region, enhancing infiltration and GWR. Planting trees and vegetation improve water infiltration and GWR. Urban green structures like permeable pavements also promote GWR and allow water to infiltrate. The technique of GWR is composed of three processes: collect, store, and use rainwater for various purposes (Xu et al. 2021). GWR has become a major issue in water supply and management due to the lack of natural resources (Allafta & Opp 2021). In addition to its contribution to the available water resources, GWR practices decrease the peak flow rate, annual volume of rainfall, stormwater runoff, and pollutant loadings (Pathmanandakumar et al. 2021). The effective management and utilization of groundwater should be carried out using modern methods and technologies (Chowdhury et al. 2008). To conserve water resources, maximize water availability, and increase land productivity, the use of GWR is a crucial step (Phong et al. 2019; Das et al. 2021; Mandal et al. 2021; Pande et al. 2021). The success of GWR is mostly dependent on the choice of location and suitable design for water recharging structures (Fauzia et al. 2021). The most important factor in GWR is to identify potential zones. The suitability map assists the water resource engineers in delineating zones for GWR and exploring appropriate recharge bodies, e.g., dams, reservoirs, and injecting wells. Drilling, as well as hydrological and geological procedures, are conventional ways to investigate GWR but are time- and resource-intensive. Remote sensing (RS) and geographic information system (GIS) approaches have proven to be more affordable, responsive, and practical (Hassan et al., 2022). These techniques are beneficial because they allow the collection of data over large areas, even where direct sampling may be challenging or impractical (Rana et al. 2018; Sharma et al. 2020; Vasistha & Ganguly 2022). Satellite data provides information about a variety of factors, such as soil types, drainage patterns, geomorphology, land slope, geology, land use, and lineament density, that directly or indirectly control the occurrence and movement of groundwater (Krishnamurthy et al. 1996; Meijerink 1996; Madrucci et al. 2008). Each of these elements serves as a predictor of the likelihood that groundwater will exist in an area. These indicators can then be combined using several techniques, such as the analytical hierarchy process (AHP) and the analytical network process, to produce a common index showing the degree of likelihood that the resource will occur. AHP is a decision-supporting method, useful in dealing with complicated problems (Saaty 1977; Saaty & Vargas 1980; Crawford & Williams 1985). The AHP process involves the configuration of special factors of a specific problem with the starting order of criteria, sub-criteria, and alternatives at successive levels. The AHP includes the exploitation of the collected spatial data and experts' opinions for determining the appropriate output. In this research, the GWR suitability map is developed through sensitivity analysis.

Due to the lack of management of groundwater resources in Pakistan, artificial groundwater recharging is of great importance as it can help to satisfy agricultural and domestic water needs in a particular manner. For regions with poor natural resources management such as Pakistan, the adoption of GWR practices is thus very important. Hence, in the present research, a suitability map for the identification of GWR potential sites in the Potohar Plateau is developed using GIS-based AHP and RS. Sensitivity analysis was carried out to assess the influence of each parameter on the distribution of suitability classes of the developed suitability map.

RS and GIS methods have been used extensively worldwide to pinpoint possible groundwater replenishment zones at various scales. Alikhanov et al. (2021) used RS and GIS methods for the evaluation of factors that impact the potential for groundwater in regions of Uzbekistan. Also, Xu et al. (2021) developed a method using GIS and multi-criteria decision making (MCDM) to find prospective locations for artificial recharge in China by including variables such as the depth of the groundwater, the distance to the canal, the density of the drainage system, the quality of the source water, the hydraulic conductivity of the aquifer, and the separation from sensitive areas. By combining influencing factors, including slope, lithology, land use and land cover (LULC), geomorphology, lineament density, rainfall, soil types, and drainage density using AHP, Zghibi et al. (2020) identified the potential recharge zones in Korba aquifer in north-east of Tunisia to reduce groundwater decline. AL-Shammari et al. (2021) linked GIS and RS approaches to find aquifer recharge regions and economically viable GWR structures using the multi-criteria decision method. However, variables such as runoff, soil texture, drainage, slope, groundwater depth, land use, and geology were taken into consideration.

In this study, numerous RS techniques for monitoring and assessing groundwater supplies are explored. LANDSAT VIII satellite imagery is used in this research to extract LULC data. This imagery is useful for studying the spatial distribution of various land cover types, which influences GWR. The Shuttle Radar Topography Mission (SRTM) Digital Elevation Model (DEM) is used to develop slope and elevation maps. These maps help in understanding the topography of the research region, which affect the flow of water and potential recharge areas. RS data are integrated with GIS to simulate hydrological processes. This method enables the generation of spatial maps that delineate groundwater potential zones based on several hydrological and environmental factors. These RS techniques are crucial in the spatial analysis and mapping of groundwater resources, allowing researchers to identify potential zones for GWR and manage water resources more effectively. Hence, in the present research, a suitability map was developed for the identification of potential GWR sites in the Potohar Plateau. This map was created by categorizing it into predefined suitability classes using the spatial analysis tool of weighted overlap. A sensitivity analysis is then conducted to assess the influence of each parameter on the distribution of suitability classes of the map thus developed.

Islamabad and Rawalpindi are the two large cities of the Potohar Plateau region. Although the Potohar area is arable, it is not irrigated. Most of the region experiences a semi-arid environment with yearly rainfall of less than 25 cm. The northern part of the Plateau receives more than 100 cm of annual rainfall. Temperature changes with changes in elevation in the Potohar Plateau. Rawalpindi has 14.8 °C minimum and 28.9 °C maximum mean temperature. The Potohar Plateau is enriched with minerals like oil and gas.

Two steps are then identified, described below: Step 1 – The preliminary analysis is a leading phase in which the accessible GWR methodologies, e.g., in situ GWR, macro-catchments GWR, and micro-catchments GWR are examined, and the methodologies which satisfy the research region are employed. Step 2 – Processes the chosen parameters and the pairwise comparison table using AHP. The desired data, such as hydrological, geological, and climatic information, are gathered and examined. The influential parameters are selected based on the expert's opinions, collected data, and geological conditions of the study area. A pair-by-pair table is created using professional opinions.

The weights of all the parameters are calculated by applying a pairwise comparison approach. GIS analysis is carried out after the weights have been calculated. In this step, ArcGIS 10.5 is used to process all the data (geological, hydrological, and climatic) and generate parameter weights to create thematic layers of the selected parameters. To create the final suitability map, all the produced raster layers are combined in the final stage using ArcGIS 10.5 software.

The selection of appropriate parameters for GWR is an important stage. Largely effective parameters are chosen based on the condition of the research region. As many explored zones in the literature, geomorphologic, climate, meteorological, and hydrologic conditions are important to select the GWR area.

The suitability level for parameters is prepared from the expert's judgments on a general rating scale of criteria, as shown in Table 2. These levels depend on the conditions of the study area. In the scale, one shows equal preference between two parameters, and five shows a specific parameter that is strongly preferred over the other. Each variable is contrasted with the others and graded in order of importance. In the stage of the expert's opinion for parameter weights, a wide range of values was collected as each expert was consulted separately. Fifty experts were consulted individually from top-ranked universities of Pakistan, NUST, UET Taxila, and public administrative departments such as Capital Development Authority, Islamabad. These experts include the dean, professors, government officials not below the rank of deputy directors, associate professors, assistant professors, heads of departments, laboratory engineers, research assistants, and some doctorate, and master's research students. They all were from the field of water resources engineering and basically belong to the NUST Institute of Civil Engineering (NICE) and the Institute of Environmental Sciences and Engineering (IESE). Drainage density plays the most influential role in the runoff and ultimately GWR process. Similarly, rainfall is one of the main sources of recharge. It has a very important role in the hydrological cycle. The magnitude of rainfall affects the groundwater resources. Another crucial element that defines the rainfall-runoff connection for the plateau is the CN. Table 3 shows the ranks and ranges of the parameters.

In this research, AHP was employed to demarcate GWR sites. GWR potential sites were delineated using integrated RS and GIS techniques. Various thematic layers such as drainage density, CN, rainfall, LULC, geology, evapotranspiration, and slope were prepared. Before calculating weights, experts were consulted individually. The collected expert's opinions are utilized in calculating the weights of criteria through AHP. The principal eigenvector can be acquired by taking an average across the rows. The sum of all elements in the eigenvector is also one. Each parameter is compared with another, and an eigenvector is computed. The eigenvector demonstrates relative weights between the parameters that are compared. An Excel spreadsheet is employed to calculate the final weights, consistency index (CI), and consistency ratio of the parameters, The pairwise comparison matrix and the normalized pairwise comparison matrix are shown in Tables 4 and 5, respectively.

The number of parameters in this equation is n, and the components of the normalized comparison matrix are denoted by C_kj and k is the value of the consistency vector.

In this equation, RI is the random index based on the elements which are being compared (Garfì et al. 2009). The final weights of the parameters and their percentage weights, which are computed from Equation (1) through Equation (3), are shown in Table 6. Saaty & Vargas (1980) proposed that matrices having values CR ≤ 0.1 are acceptable while ignoring matrices having values larger than 0.1. If CR = 0.1, then it shows that the judgments had been given randomly and are 10% inconsistent. These final weights of the parameters are utilized to develop the suitability map. In our case, the value of λ is 7.40, the consistency index is 0.07, and the consistency ratio is 0.05 which is perfect as it is less than 0.1.

Markov GWR suitability map was created by integrating and manipulating parameter layers in ArcGIS 10.5. GWR suitable sites can simply be identified using the raster data. The weighted overlay tool was utilized to classify the recharge suitability map into five groups: (1) unsuitable, (2) poor, (3) moderate, (4) suitable, and (5) most suitable. The classification of the final suitability map into pre-specified five groups is individual. The basic idea beyond that categorization was to give several options to the water resource engineers, managers, and planners to delineate appropriate areas for GWR.

The results are presented and discussed in this section, followed by a general discussion of the methodology.

The results show that around 41.8% of the area has good and excellent sites for GWR, whereas most of the study area (57.52%) has moderate recharge capability, to fulfill the area's demand for groundwater. When making plans to fulfill local water needs and GWR, local authorities can take this into account. To prevent any more harm to the already vulnerable water state of the plateau, policymakers and planners should take mitigation measures and develop strategies to conserve the majority of this 41.8% of the land. Through effective management and protection of the groundwater assets in the area, the information derived from this final zone of the groundwater potential map can assist in resolving the long-standing water shortage challenges in different areas of the Potohar Plateau and its neighboring locations. In contrast to the earlier studies (Adeyeye et al. 2019; Gedam & Dagalo 2020; Khalid 2020; Springer et al. 2023; Yaw & Ma 2023), the present study fills a research gap by utilizing the approach in a non-coastal location and changes it by using DEM data with a higher resolution and thematic maps with a bigger spatial scale to improve the methodology's accuracy. The one-time data and output map were used in all previously published research. They, however, do not accurately reflect the GWR. This is why the examined datasets may vary over time, necessitating the usage of additional datasets to get around this restriction. Because all datasets in this analysis were seasonal or time-varying, the annual mean was chosen for all of them. Second, prior research studies (Popoola 2020; Alam et al. 2022) used datasets with limited influence that might not accurately reflect the study's condition. All the significant aspects were examined in the current study to take the full scenario into account. As a result, the model provides trustworthy actual outcomes. In addition, this study considers more contributing aspects than the other studies (Abijith et al. 2020; Qureshi 2020; Ouchar et al. 2021; Benjmel et al. 2022; Ifediegwu 2022), which will further improve the output's accuracy. The study offers a comprehensive strategy that produces noticeably better outcomes and may be used in other places as required.

In order to examine the ground conditions, a number of site visits have been carried out. Only a few suitable sites are needed for any proposed GWR project, the identified zones provide ample options for this purpose. Mostly, good and excellent areas are observed located in the regions having slopes from 0 to 3%, in the areas having high rainfall, and good soil properties. There are three large dams constructed in the Potohar Plateau, namely, the Rawal dam, the Khanpur dam, and the Simly dam. There are also many small dams in the Jhelum district. These dams are considered to be several structures that have been suggested in relation to the practice of GWR.

In order to determine the impact of each criterion on the distribution of five suitability classes on the suitability map, sensitivity analyses have been performed. Sensitivity analysis has a very important role in identifying the influential criteria that affect the GWR suitability map. Each criterion is significant in its place; however, important parameters can be determined during a sensitivity analysis. Based on the expert's judgment data, a set of ranges and criteria weights has been derived, as described in Section 3.3. Sensitivity analysis is performed using two sets of expert-derived ranges and criteria weights. In this analysis, the most significant criteria are drainage density, rainfall, CN, and slope. Table 7 shows the two sets of weights, whereas Table 3 shows the parameter ranges. The combination of ranges and weights forms four input categories. The weights are presented in Table 6, while ranks of parameters are shown in Table 3. Table 8 shows how ranges and weights affect the distribution of suitability classes. The excellent class area significantly increases particularly by 95% with weight set 1. In contrast, the increase in the excellent class under set 2 is modest, not exceeding 25%. Moderate and good classes generally decrease across all combinations of weights. Similarly, a decrease in moderate and good classes is observed in both sets of weights. The area classified as unsuitable increases by 2.2 times its original value across all combinations, limiting the areas of other classes. The sensitivity analysis identified set 1 as the most favorable weight set, leading to a significant increase of about 95% in the excellent class for GWR. A significant difference in the parameter selection has been identified when comparing this study's results with other studies found in the literature. In previous studies, for example, Khan et al. (2020), Balkhair & Ur Rahman (2021), and Mondal et al. (2022), the number of parameters has varied. The parameters used in these studies are lithology, rainfall, lineament density, slope, topographic wetness index, LULC, drainage density, etc. This research showed that the catchment hydrological condition, selection of parameters, weights of parameters, and their ranges have substantial roles in selecting the suitable site.

• The technique, which linked RS, GIS, AHP, and the field data, was used for the delineation of potential zones for GWR, and the recharge suitability map was developed for the study area.

• This suitability map was enough to employ GWR practice at any site based on the analysis which demonstrated that the majority (41.9%) of the area can be classified as good and excellent for GWR.

• The recharge suitability map may be utilized as a guide to construct GWR structures at appropriate sites where the water may be recharged potentially (Mandal et al. 2021; Sresto et al. 2021; Agyemang 2022).

• As GWR projects need a rather limited amount of space per site, the recharge suitability map demonstrates that recharging can be carried out all over the research region while considering moderate, good, and excellent sites.

• These sites are not only intended to maximize GWR but also help reduce flood risks in the region, and increase water supply with a view to meet growing demand while contributing to an even distribution of water resources across the whole region (Aslan & Çelik 2021; Dar et al. 2021; Gebresilasie & Gebrie 2021).

• Sensitivity analysis demonstrated that drainage density, rainfall, CN, and slope are the critical parameters in the research region. A significant difference in the distribution percentage areas of suitability classes has been created by the introduction of a combination of expert weights. The sensitivity analysis showed that the expert weights of the criteria have a significant impact on the percentage of areas per suitability class.

• Data acquisition is itself a challenge during this study. The required data were not readily available or accessible, leading to delays or constraints in the study. When data were collected from numerous sources, integrating it into a consistent format was difficult.

• As compared to observed data, satellite data could be less time-consuming and -effective. Satellite data provides comprehensive coverage across large areas, capturing data from entire regions or even globally in a single pass. This makes it more effective for studies requiring a broad overview, such as land-use change analysis.

• A comprehensive methodology for the selection of GWR sites has been presented in this research. This approach is shown in the case study of the Potohar Plateau of Pakistan.

• GWR potential areas delineated by employing the recharge suitability map will serve as a guide for the water resources engineers and planners for efficient planning of hydro resources of the Potohar Plateau.

• Moreover, the possibility of building recharge bodies is not only an engineering, geological, etc., decision, but must take into account social, cultural, and economic aspects in order to produce multiple benefits.

• In this sense, an appropriate stakeholder engagement should be planned, in order to involve different needs, those of academic research in the broadest sense, with engineering technologies, with requests related to the constraints on the territory imposed by public administration studies, and with what is suggested by the operators in the agricultural world based on the different cultivation practices, by non-governmental associations and by citizens, who are all concerned with protecting the balance of nature.

The following recommendations are made for more studies based on the research methods and the findings: (1) GWR structure size study, (2) effect of the identified recharge zone on the water distribution of the entire area, (3) soil sedimentation and research, (4) inclusion of the GWR process in the resources management of the area, (5) evaporation study, and (6) socioeconomic studies. Different hydraulic structures or small dams can provide a large number of services. This can lower flood risks, increase GWR, and promote reliable access to water.

All authors were involved in the intellectual elements of this paper. S.Z., M.A.U.R.T., M.L.U.R.S., and P.M. designed the research. S.Z., M.A.U.R.T., and P.M. conducted the research and wrote the manuscript. S.Z., M.A.U.R.T., and M.L.U.R.S. helped in the data arrangement and analysis. All authors have read and agreed to the published version of the manuscript.

This research was conducted without any external funding or financial support.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.