Delineation of groundwater potential zones for sustainable development and planning using analytical hierarchy process (AHP), and MIF techniques

Groundwater plays a vital role in the sustainable development of agriculture, society and economy, and it's demand is increasing due to low rainfall, especially in arid and semiarid regions. In this context, delineation of groundwater potential zones is essential for meeting the demand of different sectors. In this research, the integrated approach consisting of analytical hierarchy process (AHP), multiple influence factors (MIF) and receiver operating characteristics (ROC) was applied. The demarcation of groundwater potential zones is based on thematic maps, namely Land Use/Land Cover (LULC), Digital Elevation Model (DEM), hillshade, soil texture, slope, groundwater depth, geomorphology, Normalized Difference Vegetation Index (NDVI), and flow direction and accumulation. The pairwise comparison matrix has been created, and weights are assigned to each thematic layer. The comparative score to every factor was calculated from the overall weight of two major and minor influences. Groundwater potential zones were classified into five classes, namely very poor, poor, moderate, good and very good, which cover an area as follows: 3.33 km2, 785.84 km2, 1147.47 km2, 595.82 km2 and 302.65 km2, respectively, based on AHP method. However, the MIF groundwater potential zones map was classified into five classes: very poor, poor, moderate, good and very good areas covered 3.049 km2, 567.42 km2, 1124.50 km2 868.86 km2 and 266.67 km2, respectively. The results of MIF and AHP techniques were validated using receiver operating characteristics (ROC). The result of this research would be helpful to prepare the sustainable groundwater planning map and policy. The proposed framework has admitted to test and could be implemented in different in various regions around the world to maintain the sustainable practices.


Introduction
Water is a significant natural resource to mankind and has been using and controlling from the moment of creation to sustain life. Groundwater for daily use in various purposes like drinking, cooking and grooming has provided about half the feasible freshwater. Irrigation and industrial production are depending on the groundwater supply (Ndatuwong et al. 2014;Takase et al. 2019;Luker et al. 2019). The regime of groundwater resources is difficult for researchers, planners and decision-makers in massive population and water shortage countries with surface and groundwater quality are affected by climate change (Al-Bakri et al. 2013;Zanini et al. 2019). In many regions, the groundwater level decreases, and the unavailability of freshwater to agriculture and irrigation uses. Therefore, this is also declining, which affects crops productivities (Agarwal et al. 2016). Urban water demands are continuously growing (Gauthier et al. 2019;Ghosh et al. 2019;Ray 2019). The remaining irrigated land has fast amplified from 28.7% in 1950 to 62.3% in 2013, respectively, in India. Climate change also affects water's demand and supply chain (Machiwal et al. 2011;Sahoo et al. 2015;Chinchmalatpure et al. 2019;Rudra 2019). In this context, groundwater development must be a top priority in any country (Suhag 2019;Thirumurugan et al. 2019).
The occurrence and groundwater distribution in geological rock formation depend primarily on the permeability of the formation. Rainwater quickly percolates and adds to groundwater by interconnecting cracks, joints, faults and shear zones or solution cavities (Ganapuram et al. 2009;Elmahdy et al. 2013;Khalid 2019). The hard rock area has been found in the central region of India. Surface runoff is very high in this area, and rainwater is not easily conserved in the fracture rock and aquifers. Most of the groundwater has been presented and conserved in the cracked and weathered rock areas (Saha et al. 2017;Díaz-Alcaide et al. 2019). Systematic aquifer mapping is required for sustainability growth but has not received much attention (Yu et al. 2019;Bierkens et al. 2019;Arabameri et al. 2019).
Remote sensing (RS) and geographical information system (GIS) have benefited from joining together primary and numeral data-set (Burrough 1986;Shailaja et al. 2019). Remote sensing has played a vital role to map and analysis at synoptic scales (Mahato et al. 2019;Chen et al. 2019;Gueretz et al. 2019). A combination of geospatial technologies datasets was processed in the RS and GIS software and used to delineate and interpret the suitable groundwater areas (Mahato et al. 2018;Qadir et al. 2019). Many scientists have used advanced techniques for the preparation of thematic factors, including slope, surface runoff, rainfall, soil types, drainage map, geomorphology, geological formation, land use, flow direction, groundwater level, NDVI and many others (Pradhan et al. 2019;Rani et al. 2019;Maity et al. 2019). It delivers correctness and decreases the probabilities of human mistakes by using a numeral of essential parameters like geology, lineament, lineament structure, stream density, slope, land use, and soils types. Previously, various approaches have been used for interpreting the groundwater suitable sites using weighted overlay analysis method, multiinfluence factor (MIF) and AHP techniques (Kanagaraj et al. 2018;Thapa et al. 2018;Ghimire et al. 2019), RS, GIS and hydrological data were grouped (Singh et al. 2010;Selvarani et al. 2016), statistical approaches (Umar et al. 2014;Rajaveni et al. 2015;Sener et al. 2018) analytical hierarchy procedure and groundwater flow modeling (Murmu et al. 2019;Sashikkumar et al. 2017).
The Mula river basin is situated under the semi-arid condition of Maharashtra in India. In this area, more rainfall was observed only during monsoon seasons, and in particular in the last two months of monsoon, more rain was reported in the entire basin. In this context, an accurate groundwater potential zones map can play an essential role in maintaining the balance supply of water to human, industry and ecology development. Furthermore, changeable rainfall during the monsoon and freshwater shortage starting during December to June, upcoming year sustainable watershed development and planning should be required in these regions. This paper has provided more accurate techniques and information for groundwater potential maps that are useful for decreasing the groundwater issues and storing water in the aquifer.
Therefore, the purpose of this study delineated a correct groundwater potential map by using RS and GIS, the map is created using analytical hierarchy process (AHP) and multiinfluence factor (MIF) techniques and is validated using ROC. The overall references of MIF and AHP approaches are listed and described in Table 1.

Study area
The Mula river flows from northwest to southeast direction. The basin is situated between 19° 34′ 12" N latitude and 74° 59′ 24" E longitude (Fig. 1). The minimum and maximum elevation lies between 555 and 650 m above mean sea level (Source: DEM). The total basin area is 2275.86 km 2 . The area has lowered from sources in a bottomless valley with a deep mountain, shortly in the northeast and for another 30 km enters the plains in the same direction. The entire longitude is around 2.5 m from the Mula river basis to convergence with the Pravara river. Many hills, the Mula basin, join Rahuri in a bottomless bed on steep rocky sides. In this area, sugarcane is the main crop, and secondary crops are onion, wheat and vegetables crops, etc., which indicates Land cover, groundwater depth, slope, drainage, flow direction, flow accumulation and geomorphology

WOA, GIS and MIF
The MIF method has demarcated the groundwater potential zones mapping based on the satellite data and software Not performed Mehra et al., (2016) crops based on the dam water and groundwater in the entire basin area. The Mula Dam is constructed and established on the Mula river in western region of Rahuri block. The maximum water storage volume is 26 TMC. Total annual rainfall is 345-850 mm (Source: Indian Metrological Department (IMD)) in the whole river basin area; it is very low rainfall compared to other regions of the state. The overall area is under the basaltic rock and Deccan Plateau. The river Mula is a tributary of the Bhima. Mula River has a dendritic drainage pattern type. In this view, the groundwater potential map can play a significant role in the groundwater level improvement and future challenges to face water demand and other related issues. The minimum and maximum temperatures recorded was 24 °C and 42 °C (Source: IMD) in Ahmednagar District.

Selection of the data layers
For this study, different factors, namely land use/land cover (LULC), soil types, geomorphology, digital elevation model (DEM), flow direction, hillshade, slope, normalized difference vegetation index (NDVI), aspect and groundwater depth of aquifer, were used. Basin level geological and hydrological information has been play an important role in the spatial groundwater circulation. The capacity for water holding for discharge/renew of rock groups and aquifer deposit was greatly affected by characteristics of aquifers (permeability, hydraulic conductivity, transmissivity and porosity), interconnectivity and forms of rock (Yeh et al. 2016;Mukherjee et al. 2020). The NDVI shows the canopy density calculated by Sentinel-2 data, which is helpful as a factor considering the high canopy bulk increases the probability of groundwater harvesting in poor aquifer zone Archibald et al. 2019). The adopted methodology has been used for obtaining the potential groundwater maps for the basin area, and the adopted techniques are presented in Fig. 2.

Data collection and preparation of thematic layers
For the current analysis, ten thematic layers have been considered. The data sources and layers of information are  Table 2. Satellite imagery was collected from National Remote Sensing Center (NRSC) and U.S. Geological Survey earth explorer, accessible through http:// bhuvan. nrsc. gov. in and pre-processed and https:// earth explo rer. usgs. gov by image processing software. Geology was prepared from the Geological Survey of India (GSI, Nagpur) map available at a 1: 50,000 scale. A soil texture map has been prepared from National Bureau of Soil Survey and Land Use planning (NBSS & LUP), Nagpur. The groundwater level data were collected from observation wells of Central Groundwater Board (CGWB), India and used for validation. AHP and MIF results were validated with receiver operating characteristic curve (ROC). The basic approach for assessing the reliability of the analytical assessment is determined by the ROC (Egan 1975). False-positive values along with the x-axis, true positive along with the y-axis rate are plotted for ROC. The ROC curve is a balance of the two values (Negnevitsky 2002). The complete procedure used in this analysis is shown in Fig. 2.

Weighting and ranking of every layer (AHP)
Analytical hierarchy process (Saaty 1986(Saaty , 2008) was used to calculate the weight of thematic layers. Every layer's effect was not equal; hence a relative weight, as per Saaty's scale from 1 to 12, was added to each thematic layer with its difficulties and capacity for water holding. The relative weights were allocated depending on past reports on other geographical areas with ground knowledge. The comprehensive stages have been identified of the procedure (Mu et al. 2017).
The pairwise comparison matrix has been created for thematic layers with strength judgments (Table 3). Based on their formation/categories, the slope, geomorphology, soil types and land use maps were categorized. AHP technique supports multi-parameter calculation (Saaty 1986;Bhatla et al. 2019). The slope map has divided into two sub-classes depending on the digital elevation model (Elewa et al. 2011). The NDVI thematic map was classified into four sub-classes using spatial analysis tools. However, a groundwater depth map has been prepared into four sub-classes based on the interpolation method in the ArcGIS platform. One sub-class was determined with a slope, hillshade and Digital Elevation Model (DEM) values of 0 and one class were found by the Jenks natural disruptions sorting technique as an extensively used technique for the creation of maps of the slope, hillshade and DEM sub-classes (Jothibasu et al. 2016;Arulbalaji et al. 2019). Additionally, some investigators have implemented the Jenks natural interruptions classification technique to categorize various maps (Mallick et al. 2019). A comparative rank on a 0-12 scale (Table 7) (Saaty's) was allocated to every sub-class of thematic layers depending on relative influences groundwater availability (Tables 4 and  5). The finding on the pairwise contrast and sub-classes of thematic maps was verified by estimating the consistency ratio (CR) by Eqs. (1 and 2) (Saaty 1990).
The following equation was used to measure the consistency ratio (CR) ( Table 3).    where n is the number of parameters helpful in the study, where RI is the random index whose value by influencing parameter; and CI is consistency index which is estimated by the following Equations: If CR value ≤ 0.10, it is acceptable for analysis. If CR = 0, it is perfect level of consistency.

Overlay analysis method
Here the weights have been assigned to every thematic layer and ranks were allocated to every sub-class of various layers. The groundwater potential zones can be reported using multiple overlying maps using the weighted overlay technique (spatial analyst tool) in ArcGIS 10.5 using Eq. (3).
Every factor and their particular classes have assigned the grades and weights depending on their proportional possible influence ( Table 5).
The following equation was applied to calculate GWPZ.
(2) CR = CI RI  where GWPZ illustrations the potential areas identified using Eq. 4. The various thematic maps such as soil texture (

Multi-influencing factors (MIF)
The important factors revealed interrelationship between factor and their effects are specified in Table 6. If an impact occurred, viewing a robust correlation between factors, after 1.0, a score is assigned. If a minimum effect occurred, revealing poor relations between elements, and those parameters score 0.5. The comparative score to each parameter was measured from the overall weight of two major and minor influences in the definition of groundwater potential zones. The relative score is extra functional for calculating the planned score of individually impacting parameters, as specified in Eq. 5. The arrangement of weighted influencing parameters is assigned into sub-classes to delineate potential groundwater areas (Tables 7 and 8). where D represents the suggested mark of each inducing factor, A positions for the major relationship both factors, and B is view points for the minor interrelationship with both parameters. Each group of single themes has been given a rank by weighted overlay investigation (Ghosh et al. 2015). By groundwater and hydrogeology information, the required scores have been assigned under the multi-influencing factor (MIF) of that specific factor (Manap et al. 2014). The methodology flowchart and the weighting particulars allocated to the different maps and their characteristics in the current study are shown in Tables 8 and 2. The single influence parameters' planned score was separated and assigned to each reclassified sub-parameter (Table 8).

Sensitivity analysis
The sensitivity analysis based on the map removal technique has been applied to identify to most sensitive layer in producing the GWPZ map. The study was performed using the 'raster calculator' of map algebra function available in the spatial analyst tool of ArcGIS.

Validation and accuracy of GWPZ
Total 35 dug and open wells data have overlapped to know the accuracy of the groundwater potential map. Receiver operating characteristic curve (ROC) method was applied to compare the accuracy of MIF-and AHP-based GWPZ. ROC helps to identify which method is the best for groundwater suitable zone mapping (Pourtaghi et al. 2014;Ozdemir et al. 2011). The ROC plot shows a range from 0.5 to 1.0. To test the accurateness of groundwater potential maps based on the ROC model (Nandi et al. 2009). The AUC is equal to 0.5 if models do not assess the amount of groundwater zone. ROC makes the graphical picture plotting the true positive (sensitivity) and untruthful positive (1-specificity) rates on the Y and X axes, respectively (Pourghasemi et al. 2012). The observation locations of wells were overlapped on the groundwater potential map and evaluated the correctness of the outcome using Eq. (6) in the numerous groundwater potential areas in the river areas (Murmu et al. 2019).

Results and discussion
The basin is under a hard basaltic rock and does not have more water in the fractured rock zones. The demarcated groundwater potential areas map exposed projecting dissimilarities in potential areas of the groundwater basing region. About 3.049 km 2 , 567.42 km 2 , and 3.33 km 2 , 785.84 km 2 of the river basin area have very poor and poor possible areas of groundwater by MIF and AHP methods, respectively. At same time, basaltic rock is under the extremely low potential of groundwater in the river basin area. The MIF potential map has shown 20.43% very poor and poor area, but the AHP groundwater potentiality map has demarcated 25.34% very poor and poor area. namely, good, moderate and very good groundwater possible areas, correspondingly using the AHP technique (Table 8).

Thematic layers
The data interrelated to different thematic layers created to potential zones maps are deliberated in below:

Aspect
Aspect replicates the moisture retention, vegetation, moisture air and attitude of rock bedding, which influenced soil structure's physical assets, sloped materials and potential groundwater zones. In the current investigation, the northern areas and northeast region of the river area have the positive weight for groundwater potential and south and southeast part of the study have higher steeper value of slope and gradient and shows negative weight for potential zones map of groundwater (Fig. 3).

Curvature
The most significant characteristic to be measured for observing the potential groundwater areas, water hydrology and instability assessment of terrain is based on the curvature of peak slopes, representing the morphology of the area topography. The dynamic of the subsurface of hydrology, formation and accumulation of soil is better precise by curvatures of a slope. The curvature value ranged between − 4.35 and 2.90. The positive value of curvature indicates that the surface is curved. However, the negative value shows that it dipped, and the zero is assigned to the linear surface (Fig. 3). The entire river basin area comes under the category of concave. It represents that increase there is a in the pore water pressure during medium and heavy rainstorms.

Elevation
The factor elevation is a key feature used to detect groundwater potential areas. In general, plain surface area has low amount of irruption compared of moderate to high elevated points. The elevation map was created from CAR-TOSAT-30 m DEM data in GIS environment and it was reported that, the range of elevation varies from 0 to 1512 amsl in basin area (Fig. 3).

Flow direction
Flow direction is stated about the intimacy of the arrangement of channels. The stream or flow direction is the opposite of the permeability factor. The highest flow direction shows more runoff, which reduces the amount of water penetration to the subsurface and vice versa. In the study, the flow direction ranged between 1 and 255 (Fig. 3).

Hill shades
A hillshade is a 3D-grayscale depiction of the surface, considering the comparative site of the sun for the coloring of the related. To define the way of the sun, this feature uses the properties of altitude and azimuth. The hillshade map was created using a 3D-spatial analysis tool in the ArcGIS software 10.5. In the map, the hillshade value ranges from 0 to 180 (Fig. 3). This map is very important for the understanding of the surface topographic and hilly places entire basin area.

NDVI
The normalized difference vegetation index is another important parameter that plays a vital part in the constancy of grades by hydrological procedures in the eco-system. In the study, NDVI value is divided into four sub-classes: good, very good, very poor and moderate. The NDVI value varies from − 0.21 to 1. The preparation of NDVI value of − 0.21 to 0.16 was very poor, 0.17-0.36 was good, 0.37-0.54 was moderate, and 0.55-1 was very good (Fig. 3). For the potential groundwater areas, high weightage has been assigned for a higher value of NDVI, and low weightage has been given for a lower value.

Slope
The slope is the significant criterion to estimate the groundwater potentiality. It influences the penetration amount of surface water. The lower and higher value of slope angle indicates the flatter and steeper terrain, respectively. Many researchers handled the slope layer to evaluate the water flow and conserved in the specific region. In the present study, the flow of river water is entirely using the slope gradient of the basin. The figure shows that more extensive slopes are identified in the eastern, southeast part and lower slopes are determined in the northern and northeast parts of the study areas. The flat slope region causes more infiltration than to steep and moderate steep region. It is reported that the range of slope varied between 0 and 3.61 (Fig. 3). About a higher percentage is covered by moderate and more minor slope in river areas.

Soil texture
Soil texture is a crucial controller of infiltration of land water using voids and infiltrates into the subsurface to reach the aquifer. The preparation of soil type's maps is based on the mean size of the solid particle in soil, namely sand, silt and clay. The weights are allocated for every soil texture in rendering to its penetration rate. A higher weightage has provided clay loam, clay, sandy clam and clay (Fig. 3).

Geomorphology
Geomorphology is a scientific study of creating, changing, configuring and relating landforms to underlying structures. Visual interpretation of satellite image geomorphic units/ landforms is categorized, which depend on visual image elements. The topographic information in SOI topo maps aids in interpreting satellite imagery (Moharir et al. 2020). Geomorphology is a clear way to define and examine the environments and their processes systemically. In the study area, geomorphology was divided into plateau, pediment and pediplain to calculate the potential groundwater zone (Fig. 3).

Land use and land cover
Land use changes are governed by anthropological intervention and general phenomena such as agronomic demand and skill, residential development, consumption patterns, urban expansion and economic development, science discipline and skill and added factors (Pande et al. 2018(Pande et al. , 2021. As a result, early and accurate information regarding LU/LC of earth's surface changes identification is essential for interpreting genetic and environmental phenomenon relationships and interactions to enhance decision-making. Supervised classification was done, and five classes have been found: forest land is good quality, forest land is poor quality, agriculture land and water body (Fig. 4).

Groundwater depth
The groundwater level is under the earth's surface. It is saturated water or the level to which groundwater would increase in the fine that is bored in a pressurized aquifer.
Groundwater level analysis is the most significant feature to use in the preparation of potential areas of groundwater. The groundwater depth information is observed during the monsoon period; it is valuable data to check suitable zones for sustainable groundwater planning structures in the basin area. The inverse distance weighted method (IDW) created the groundwater level map in the ArcGIS software (v10.5). Groundwater depth map recorded minimum, and maximum values are 4.5-14 m (Fig. 4).

Calculation of groundwater potential zone maps
Contributing the thematic maps by weighted overlay method (WOA), and consistency ratios are computed for groundwater potential zone maps. The value shows that the finding matrices were authentic (< 0.10) with rational reliabilities. Therefore, reclassified maps were combined based on their found weightiness (Table 5) by overlay analysis technique with groundwater potential map made by AHP technique (Fig. 5). The AHP maps of potential groundwater areas were delineated into five types (such as very poor, moderate, good, poor and very good possible zones of groundwater) founded on the pixel values; these map classes were divided by normal breaks technique in the Arc GIS system (Fig. 5).
To minimum the grid marking, the map was developed by the common filter of the GIS system. The existing study emphasizes that preparing a potential map of groundwater depends on significant teen parameters, LULC, soil texture, DEM, flow direction, geomorphology, hillshade, slope, NDVI, aspect and groundwater depth. The weightage was assigned to the whole ten thematic layers, and their features depend on the multi-influencing factor (MIF) (Magesh et al. 2012). The groundwater potential areas map was classified as very poor, moderate, good, poor and very good using the MIF technique (Fig. 5).

Validation of groundwater potential zone maps
Through cross-validation, the accuracy of the produced possible zones map of groundwater has been compared to groundwater level data of wells collected from the CGWB (2016-2017). The difference might be due to the long-years over-withdrawal of groundwater for farming works that have co-operated the groundwater harvest volume at the current environmental situation in the river basin. Access to groundwater for agricultural use is also flattering progressively week to the dangers of groundwater reduction, which has been informed by numerous studies (Chindarkar et al. 2019). To avoid groundwater depletion, modifying the conventional cropping types and implementing drips, sprinklers and micro-irrigation practices may be helpful for specific areas (Wang et al. 2018;Kaarakka et al. 2019). Total 36 observation water level data are covered on the described potential areas map of groundwater was used to validate (Fig. 6). Thirty-six observation wells locations, nearly 25 wells were found to contest the demarcated potential zones maps (Table 9). The full correctness of groundwater potential zones mapping by implemented AHP, MIF and GIS approach has been identified to be 86% and 80% correct, signifying the dependability of the techniques (Shekhar and Pandey 2014;Mohamed and Elmahdy 2017;Saranya and Saravanan 2020). The investigation added measurable 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0%

ROC Curve
100% 41% 20% 10% 6% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% cross-validation maps of groundwater prospect areas by receiver operating characteristic (ROC) graphs (Fig. 7). For various cut-off themes of variable, the ROC charts a true positive rate against an untruthful positive rate and every idea on the arc denotes the sensitivity couple referring to a threshold. On the other side, the field in the Curve (AUC) calculates how well a factor can differentiate both analytical sets from each other. AUC value inside the 0.5-0.6 range recommends low accurateness of forecast. At the same time, the 0.6-0.7, 0.7-0.8, 0.8-0.9 and 0.9-1 levels recommend that very good, average, excellent and good accurateness of forecast, correspondingly (Andualem et al. 2019;Kumar et al. 2020). The validation of outcomes exposed for good prediction by the AHP method as the AUC of the map of the potential areas was 0.86 (Fig. 7). In this context, validation of results correlated reasonable prediction using MIF technique as the AUC was 0.80 of potential groundwater zones. Therefore, the current work has accomplished a suggestively higher correctness level with the highest quantity of appropriate thematic maps by using AHP and MIF procedures (Ahmed and Mansor (2018) The observed potential zones can give primary strategies in choosing new wells locations for the removal of energetic water resources management in the basin areas. This outcome could be valuable for the implementation of a groundwater harvesting strategy. It is exciting to note the limitations and room of the work and the originality of the work. Ten thematic layers have been selected in this research, but the addition of some data such as hydro-geological and geomorphology maps (geomorphology and water level) could have provided more accurate results. Sustainable planning primarily emphasizes the use of groundwater, water management, restricted abstraction, advanced water structure, water awareness, alternative cropping methods, domestic with industrial water conserved. The non-polluted groundwater preparation would be valuable for water improvement in the long term.
Likely plans are suggested to entire basin area with poor aquifer zone (Fig. 8). There are theoretical limitations to the analysis due to being of the adopted approach. MIF and AHP is a used that can also suppress persuaded errors. Factors of impact are the rate of water flow for agriculture, and the use of drinking water often plays an important part in regulatory possible in groundwater zones which have not been considered in this investigation. However, given these variables in the examination technique and estimation, the results are accurate and scientific. The AHP results can help enhance the managing of groundwater in the river basin.

Recommendation
Reworking of managed aquifer recharge (MAR) techniques with sustainable groundwater harvesting activities is suggested to rise the potentiality areas of groundwater for the drought patches in the basin. As application of new groundwater recharge structures are cost-intensive, planning and progress of new cement nala bund, CCT and farm pond, development and tunneling of the drainage line with presenting pools might be supportive to suitable sites of groundwater recharge and storing capacity in the study area. The current investigation can be helpful in the soil and water management operations to make community publicity. In order to obtain more specific guidance on the rate of groundwater which is obtained from various groundwater potential areas, this study is necessary for the technically prediction of groundwater in various potential regions in the Mula river basin.

Limitations
While this investigation follows a cost-efficient scientific way of analysis of various natural resources parameters, some limitations cannot be unheeded. The main limitations are defined below: (i) The various satellite and field data of this study are deepened on satellites, supplied by various agencies and departments. However, maximum of this map is an extremely widespread in countryside. A maximum difference in whole these parameters may be identified at the local stage or small scale. Therefore, by reason of this issue, the produced groundwater potential maps could not be very precise at a same local scale. (ii) AHP is an experiential technique where the ranking of criteria is completed based on proficiency. Although this research displays very good correctness, an addi-tional detailed map may be forecasted by using the one way by rearranging the ranks. (iii) MIF and AHP results are validated with water level data, but the groundwater level is changed in region to region, although ROC plot has been given more precise results at local, international and national scales. (iv) The sustainable groundwater planning and development have been prepared by AHP results but the soil and water conservation structures planning can apply on field, it is a big problem. (v) This investigation is totally intensive on AHP, MIF and RS that give a general estimate of groundwater potential zones and planning maps. However, for risky detailed valuation, ground survey-based geophysical approaches are very beneficial. However, groundwater potential zonations using ground-based geophysical survey and field survey are not only expensive but also tremendously time-consuming.

Conclusion
The present study investigated the identification of potential zone in the basin using MIF and AHP. In this paper, two types of methods have been applied for the delineation of potential groundwater areas. AHP and MIF have found 595.82-302.65 and 868.86-302.65 km 2 areas as extremely possible for finding groundwater. MIF technique has made more remarkable outcomes than the AHP method than set theory based on the equation. This study also recommends the enclosure of other geological, groundwater and metrological factors to develop the result. This paper discussed and correlated the two techniques with each other which techniques best for the potential zones of groundwater. Two groundwater potential area maps were cross-validated with the groundwater depth data in the ROC and AUC models. The cross-validation outcomes recognized AHP as a more effective technique (accuracy = 0.86) to demarcate potential groundwater zones for the basin area. The cross-validation results found MIF is a moderately effective method (accuracy = 0.80) to extract the groundwater potential mapping in the area. This result of the study can be useful to apply sustainable groundwater plans for rainwater harvesting, soil erosion and cropping types. In the area, four to three big dams and water bodies are found to be better conditioning parameters for potentiality areas of groundwater, so speedily recovery of water frame is required to be stopped directly. Therefore, adequate management of groundwater is significant for the long sustainable social, financial and environmental development for the Mula river area of Maharashtra in India. It developed sustainable groundwater.