Introduction

Groundwater is an essential natural resource that has a significant influence on biospheric, geospheric, and atmospheric processes. However, without water, no living being can stay alive for a long time, except few microbes and bacteria. About two-third of the earth's surface is covered by oceans, which contain a huge volume of water, but still there is a severe water resource scarcity because of the limited accessibility of fresh water. Glacier melts, subsurface water, and the rivers are the only source of fresh water, which can be consumed in domestic and economic activities (Das et al. 2017). Since the early twentieth century, the population growth is very high in a few developing countries such as China, India, Brazil, etc. Due to the over-population growth, the consumption rate of water is extremely high due to intensive agricultural and industrial practices. As a result, the reservoirs in these countries are depleting rapidly and regions are becoming more vulnerable to drought (Das 2019a). Therefore, these regions are required to have proper watershed management and groundwater resource planning to prevent water resource scarcity.

Management of watersheds, available natural resources, and assessment of groundwater can easily be done through morphometric analysis. The surface characteristics, geomorphic landforms, and underlying geological structures with their characteristics the region's topographic information can be described through drainage basin morphometry (Das et al. 2018). Groundwater occurrence, movement, and storage in an area are controlled by the geological and geomorphological setup of the area. There is a direct and indirect influence of geological and geomorphological factors on the hydrological settings of an area. At the same time, the amount of runoff, infiltration, and rate of percolation are governed through the physiographic elements such as relief and slope (Das 2019b).

Numerous researchers performed studies on the morphometric analysis of different basins through conventional methods (Horton 1945; Smith 1950; Strahler 1957). Morphometric analysis is comprehensively utilized for integrated land and water resources management, and identification of groundwater potential zones (Mishra et al. 2011; Das and Pardeshi 2018a). The morphometric analysis assisted for identification of suitable sites for the construction of artificial recharge dams (Zaidi 2011).

Until late twentieth century, drainage basin analysis used to be done using traditional techniques such as topographic map analysis and aerial photograph interpretation. However, innovation of GIS made the geo-environmental assessments easier than ever (Ghasemlpunia and Herfeh 2017; Das 2019c). Digital elevation model (DEM) coupled with GIS software is an easy, time and money-effective tool for various geological and geomorphological applications such as drainage basin analysis, extraction of lineaments, geo-environmental modelling (Farr and Kobrick 2000; Grohmann 2004; Sreedevi et al. 2009; Benffai et al. 2017). Many multi-criteria decision techniques have been used by previous workers through GIS platform for the assessment of groundwater, for example, analytical hierarchy process, fuzzy logic, weighted overlay analysis etc. (Hosseini and Saremi 2018).

The Central Groundwater Board of India (CGWB 2012) estimated that the annual replenishable water resource in the country is about 431 km3. States like Rajasthan, Gujrat, Maharashtra, Panjub, and Haryana are being depleted since last few decades (Rodell et al. 2009). Savitri and Vashisthi basins are located in the central part of coastal Maharashtra, where a severe water resource scarcity occurs during the summer. These two basins contribute a large area of agricultural land which require a constant supply of water throughout the entire year. Hence, water resource management and identification groundwater potential zones are critical requirements in these regions. Hence, an attempt has been made in this study to evaluate the morphometric analysis and groundwater potential zonation of Savitri and Vashisthi basin through geospatial analysis. Besides mapping of groundwater potential zones, the aim of our work was extended to compare two popular methods (frequency ratio and influencing factor) that are widely used for groundwater potential mapping.

Study area

Savitri and Vashisthi catchments are located at the base of the Western Ghat escarpment, encompassing an area of 2262 and 2163 km2, respectively, (Fig. 1). All the streams within Savitri and Vashisthi basins are originating from escarpment of the Western Ghat, flowing westward and falling into the Arabian Sea. A large part of both basins are covered by dispersed vegetation. Very steep slope characterises the upstream section of both catchments. Dendritic drainage pattern is the primary drainage arrangement followed by parallel and rectangular in the study area. Vashisthi shows numerous small water bodies such as lakes, dams, etc. The topographic slope is towards the west and many isolated hills can be observed in both basins.

Fig. 1
figure 1

Location map of the study area referenced to India and Maharashtra state. Savitri and Vashisthi are classified as sixth-order drainage basin

Full size image

The major lithology is basalt in both catchments. Laterites can be found at a different elevation (0–1400 m) with a flat to undulating topography (Widdowson and Cox 1996). The thickness of the laterite is highly variable from the coastal region to the escarpment. Savitri and Vashisthi basins show numerous dykes having an orientation with NNE–SSW direction (Das and Pardeshi 2018b). The entire study area can be categorised into three prominent geomorphic divisions: (1) very steep escarpment in the eastern part from where all the streams are originating; (2) the gentle pediment with scattered hills in the central part; and (3) the lowland coastal region.

The study area shows monsoon climatic condition where the rainfall intensity is very high in the monsoon season (June–Sept). The average annual rainfall in the study area is about 3000 mm (Doke et al. 2018). Though substantial rainfall occurs during monsoon, the summer period experiences drought like conditions in both catchments.

Methodology

In this study, we performed drainage morphometry and groundwater potential mapping using geospatial techniques. We obtained Shuttle Radar Topographic Mission (SRTM) DEM from https://srtm.csi.cgiar.org/. Initially, we processed the DEM in ArcGIS 10.1 software and projected to WGS 1984 UTM 43N coordinate system. After that, we performed flow direction, flow accumulation, and basin tools to generate the basin raster. Subsequently, basin rasters were converted into vector format, and Savitri and Vashisthi basin DEMs were extracted using those basin vector files. After this procedure, the methodology was divided into four sections for morphometric analysis, extraction of lineaments, degitization of different thematic layers, and demarcation of groundwater potential zones.

Morphometric analysis

We used filled DEM to create flow direction, flow accumulation, and stream order maps in ArcGIS. Afterward, the data were exported in the Microsoft Excel software and all the essential morphometric parameters were calculated. Table 1 is providing the list of morphometric parameters that we selected for this study with corresponding formulae.

Table 1 Morphometric parameters and corresponding formulae
Full size table

Extraction of lineament

We extracted lineaments from the shaded relief map. To prepare shaded relief map, SRTM DEM was used in ArcGIS. Das et al. (2018) demonstrated that the accuracy of extraction of the lineaments from shaded relief map depends on the sun elevation angle. Few certain angles are excellent for accurate extraction whereas few creates false topographic perception phenomena (FTPP) (Das et al. 2018). Therefore, the extraction was done on the basis of four different azimuth angles at 90° interval. Digitization of all the geology and geomorphic surface features (for example cliffs, faults, dykes, straight valleys, etc.) was done manually. Consequently, all these surface features created from different azimuth were merged into a single shapefile. Landsat images and Google Earth were utilized to remove the redundant feature and the man-made surface features such as rail track, roads, canals etc. from the final lineament map.

Preparation of thematic layers

We utilized eight conditioning parameters such as geology, lineament, slope, drainage density, rainfall, soil texture and depth, and well density to prepare groundwater potential maps Lithology maps were digitized from the geological quadrangle map, acquired from the Geological Survey of India (GSI). The slope maps were derived directly from SRTM DEM using surface tools in ArcGIS. The stream network vectors were used to prepare the drainage density map using line density tool. Same approach was taken to prepare lineament density raster.

We obtained grid based daily rainfall data (1950–2013) from India Meteorological Department (IMD). Mean rainfall and rainfall intensity for each pixel were calculated and maps were constructed. The average annual rainfall was divided by 365 days to obtain rainfall intensity. We acquired soil texture and depth data from National Bereau of Soil Survey and Land Use Planning (NBSS-LUP) regional office in Nagpur. These datasets were digitized in ArcGIS to prepare soil texture and soil depth map. Topographical maps of 1:50,000 scale were used to dizitize all the wells available in the study area. 70% of the wells available in each basin were selected randomly for training the FR model and rest 30% wells were used to validate both FR and IF models. Using this well inventory (the training data) the well density maps was produced.

Determination of groundwater potential zones

Frequency ratio

Frequency ratio is the prospect of occurrence of a certain factor (Bonham-Carter 1994). The frequency ratio is build upon the relationship between observational well availability and the factors influence the groundwater potential. Frequency ratio of each parameter can be calculated using the following formula:

$$\mathrm{FR}=\left[\frac{{P}_{\mathrm{gw}}/{T}_{\mathrm{gw}}}{{P}_{\mathrm{f}}/{T}_{\mathrm{f}}}\right],$$
(1)

where FR is frequency ratio, Pgw is the number of pixels with wells in each sub-category, Tgw is the total number of wells, Pf is the number of pixels in the classes of a sub-category, Tf is the total number of pixels in each factor (sum of all sub-category of a certain factor).

Table 2 depicts the frequency ratio for each factor considered in this study to prepare the groundwater potential maps. According to FR calculation, values were assigned to all factors. Later, all the parameters were integrated in ArcGIS and the groundwater potential maps were calculated using the following expression:

Table 2 Rank and weight assessment using FR and IF technique for Savitri basin
Full size table
$$\mathrm{GPZ}=\sum_{i=1}^{n}\mathrm{FR},$$
(2)

Equation (2) can be re-written in simplified form as

$$\mathrm{GPZ}={\mathrm{LG}}_{\mathrm{FR}}+{\mathrm{SL}}_{\mathrm{FR}}+{\mathrm{DD}}_{\mathrm{FR}}+{\mathrm{LD}}_{\mathrm{FR}}+{\mathrm{RI}}_{\mathrm{FR}}+{\mathrm{ST}}_{\mathrm{FR}}+{\mathrm{SD}}_{\mathrm{FR}}+{\mathrm{WD}}_{\mathrm{FR}},$$
(3)

where LG is lithology, SL is slope, DD is drainage density, LD is lineament density, RI is rainfall intensity, ST is soil texture, SD is soil depth, WD is well density, and FR is the frequency ratio value.

Influence factor technique

Influence factor technique is a simple overlay based model where weights and ranks are assigned based on the researcher’s knowledge on different parameters and their influence in groundwater potential (Das 2019a; Doke 2019). The ranks and weights were assigned to each parameter and their sub-categories based on previous literature review (see Table 3 for weights; and Table 4 for literatures). Consequently, the maps were constructed using the following expression in the ArcGIS environment:

$$\mathrm{GPZ}=\sum_{i=0}^{n}{(W}_{i}\times {R}_{i}),$$
(4)

where Wi is the weight of each factor, Ri is the rank of each sub-category.

Table 3 Rank and weight assessment using FR and IF technique for Vashisthi basin
Full size table
Table 4 Table of the parameters that were used in different studies for groundwater potential zonation
Full size table

Results and discussion

The results of this study are categorized into two major parts. The first section describes the morphometric analysis through drainage quantification and terrain measurement in the study area. The second section will give a comprehensive description of different influencing parameters and the groundwater potentiality of the study area. The significant results are discussed below.

Morphometric analysis

Stream order (u)

Strahler (1964) proposed that streams in a basin can be categorized on the hierarchical basis. At the upstream, first order is the streams having no tributaries; second order starts when two first-order streams join and the count go on. Savitri and Vashisthi basins are classified as sixth-order drainage basins. There is a significant negative correlation can be found between the stream order and stream number (Fig. 2). In all the natural river systems, as the stream order increases, stream number decreases.

Fig. 2
figure 2

Relationship between stream order and stream number. As the stream order increases in a drainage basin, number of streams decreases significantly

Full size image

Mean stream length (Lsm)

Mean stream length (Lsm) epitomizes the characteristic size of the components of a drainage network and its contributing basin surface area. Lsm can be computed by dividing total the length of streams in an order to the sum of all the streams of the same order. Table 5 represents the variation of Lsm for Savitri and Vashisthi basins (0.78–54.23 and 0.84–52.14). According to Horton’s (1945) law, Lsm against the stream order in a semi-logarithmic scale makes a straight trend line. This relation indicates the erosion laws which are acting on the geological material along with similar weathering-erosion characteristics of the drainage basin evolution. Anomalies in this relationship could be followed by structural control over an area.

Table 5 Results of drainage network properties in Vaitarna and Ulhas basins
Full size table

Bifurcation ratio (Rb)

The degree of structural and lithological control over an area can be determined using bifurcation ratio. This parameter is a fundamental study to understand the tectonic or structural anomalies of a drainage network. Schumm (1956) demonstrated bifurcation ratio as the ratio of the number of streams in an order to the number of streams in the higher order in the catchment. In ordinary cases, the bifurcation ratio varies between 3 and 5, except places where strong geological control predominates (Das 2018a). Mean bifurcation ratio of Savitri and Vashisthi basins are 4.35 and 4.25, respectively. Very low mean bifurcation ratio of these basins is indicating geological structures do not have significant dominance on the drainage network (Chow 1964; Sreedevi et al. 2009; Pandey and Das 2016). A strong relationship between drainage pattern and bifurcation ratio can be found. Regions having parallel or rectangular drainage pattern show higher bifurcation ratio compared to dendritic or trellised drainage pattern.

Drainage density (Dd)

Horton (1932) demonstrated drainage density as the ratio between the total length of streams within a basin and the basin area. Drainage density represents the closeness of spacing of streams. Factors such as geology, slope, and relief control the drainage density of an area significantly. Region having lower relief shows lower drainage density compared to the higher relief area (Strahler 1964). Drainage density of Savitri and Vashisthi basins are 0.97 and 0.93 km/km2, respectively. Lower drainage density on both basins represents homogeneous lithology, slope, and relief characteristics.

Stream frequency (Sf)

According to Horton (1932), stream frequency can be defined as total number of streams per unit area. Stream frequency values for Savitri and Vashisthi are 0.80 and 0.74, respectively, (Table 5). Stream frequency of a drainage basin mainly depends on the lithology and structural character of a drainage basin.

Drainage texture (T)

Smith (1950) demonstrated that lithology, climate, catchment relief, infiltration capacity, vegetation cover, and the stage of drainage development are the prime factors to regulate the drainage texture. Coarse drainage texture can be found where lithology is characterized by massive hard rock. Conversely, a region having soft or wear rock shows fine drainage texture. Drainage texture of a basin is merely the product of drainage density and stream frequency. Very coarse drainage texture is exhibited by Savitri and Vashisthi mainly because of the hard rock basaltic terrain.

Circularity ratio (Rc)

Strahler (1964) introduced circularity ratio as the ratio of the basin area to the area of a circle having same perimeter of the drainage basin. Rc values of 1 indicate circular basin shape and similar infiltration capacity over the entire basin. Lower values of Rc represents lower infiltration rate is very low and occurrence of high-velocity runoff can be observed. Rc values of Savitri and Vashisthi basin are 0.28 and 0.34. Moderate values of Rc indicate both basins are characterized by mature stage of drainage development.

Elongation ratio (Re)

Elongation ratio (Re) of a basin is the ratio between the diameter of a circle having the same area as the basin and total length of the basin (Schumm 1956). Re of a drainage basin indicates the shape of the basin where Re value of 1 indicates perfect circular basin. For elongated basins, Re value becomes close to 0. Table 6 shows the classification of required Re values for different shape of basins. Savitri and Vashisthi basins are showing high elongation ratio. It is indicating both of these basins are having moderately circular shape.

Table 6 Results showing variation of basin morphometric properties of Vaitarna and Ulhas basins
Full size table

Form factor (Ff)

In quantitative drainage morphometry, form factor is the dimensionless ratio between basin area and the square of basin length (Horton 1945). There is a direct relationship can be found between form factor to the stream flow and shape of the basin. Lower form factor indicates elongated nature of drainage basin, whereas a higher value indicates rectangular or elongation shape. Form factor values for Savitri and Vashisthi basins are 0.46 and 0.56, respectively, clearly indicate that these basins are having semi-circular shape.

Relief (R)

Geomorphic surface processes such as erosional mechanism and mass-movement intensity of an area exclusively depend on the relief of the drainage basin. Basin relief is basically the dissimilarity between the maximum and the minimum elevation within the catchment area. Very high relief basins are generally governed by steep gradients with dominating hillslope processes and slope failures (Das 2020). Comperatively, low-lying flat regions with deep chemical weathering character is the main geomorphic processes in low relief catchments. Also, the materials eroded from high relief zones by fluvial processes generally tend to deposit on the low relief sections as the stream power becomes lower. Relief of Savitri and Vashisthi basins is 1393 and 1193 m, respectively. Very high relief of Savitri and Vashisthi indicates governing hillslope mechanisms with high-velocity runoff and less infiltration potential of the study area.

Relief ratio (Rh)

According to Schumm (1956), relief ratio is a dimensionless metric that is the height–length ratio of horizontal distance to the longest extent of the basin to the principle drainage line. This factor indicates overall steepness and the erosional physiognomies of the drainage basin. Savitri (19.82) and Vashisthi (19.11) basins show very high relief ratio. High relief ratio of these basins indicates very steep slope and higher erosional vulnerability.

Hypsometric integral (Hi)

Hypsometric integral of a watershed can be used to determine the geological stage of drainage development (Wood and Snell 1960; Pike and Wilson 1971). Hypsometric integral of a region is the relative proportion of the highland and lowland areas. Savitri and Vashisthi basins show very low hypsometric integral (Table 6). Considering to the Devis’s cycle of erosion, both of these basins are falling to the old stage of drainage development.

Influence of different parameters for groundwater potential

Lithology

The groundwater potential of a region change place to place depending on lithological variation. Regions having massive plutonic rocks such as granite and diorite generally have very low groundwater potential due to lack of pores. However, volcanic rocks such as basalts can sometimes have pores because of pyroclastic material such as ashes, angular element structures, and vesicular textures. Therefore, basalts are generally having low to medium groundwater potential. Very high groundwater potentials can be found in regions lying on sedimentary rocks such as partially lithified shells, well-sorted fine-grained sandstone, etc. Regions over deep unconsolidated alluvium are having very high groundwater potential as well.

In Savitri and Vashisthi basins, the major rock type is flood basalt, erupted during Cretaceous–Tertiary boundary in the geological time scale about 65 million years ago. Near the coastal regions, some patches of lateritic soil can be observed (Figs. 3 and 4).

Fig. 3
figure 3

Geo-environmental parameters considered in this study to map different groundwater potential zones in Savitri basin. a Lithology, b slope, c drainage density, d lineament density, e soil texture, f soil depth, g rainfall, h well density

Full size image
Fig. 4
figure 4

Geo-environmental parameters considered in this study to map different groundwater potential zones in Vashisthi basin. a Lithology, b slope, c drainage density, d lineament density, e soil texture, f soil depth, g rainfall, h well density

Full size image

Slope

Slope is the degree of inclination of a region concerning to the horizontal surface. Understanding the slope of a region is necessary for many purposes such as agriculture, planning, construction of engineering structures, groundwater potential assessment, flood mapping, landslides etc. (Pham et al. 2020). The slope of a region is controlled by variation in lithology with different rate of resistance, different geomorphic processes, and climate. Regions having a very low slope are generally having higher groundwater potential. The slope variation in Savitri and Vashisthi basins are ranging from 0° to 64° and 0° to 62°, respectively, (Figs. 3 and 4).

Drainage density

Earlier, many researchers described that region having a lower drainage density is generally having higher groundwater potentiality due to the impermeable lithology (Bagyaraj et al. 2013; Jenifer and Jha 2017; Thomas and Duraisamy 2017). However, it is evident that irrespective or lithological variation, water generally follow the slope and accumulate on the lower surface elevation of an area. Therefore, when lithology is considered, if the streams are flowing through the porous rocky surface, there will be a substantial amount of recharge from the stream itself, if the groundwater level is lower than the river bed. This phenomenon is more evident and effective in arid and semi-arid regions, where groundwater levels during most of the season stays below the river bed.

Lineament density

Lineaments are the linear or curvy linear feature on the earth’s surface which can be seen and mapped using satellite images (Marghany 2012). There is a significant importance of availability of lineaments with the groundwater potential of an area. Lineaments are weaknesses in the rock, play a crucial role in the infiltration process in the hard rock region (Das et al. 2018). Savitri and Vashisthi basins show numerous lineaments which are generally mafic basaltic dykes with an orientation in NNE–SSW and N–S direction (Figs. 3 and 4) (Das and Pardeshi 2018b).

Rainfall intensity

Occurrence of rainfall can be observed anywhere on the earth except the polar regions (Das 2018b). However, in some region, the intensity is very high while some places characterize very low intensity rainfall. Savitri and Vashisthi basins show a semi-arid climatic condition where rainfall occurs during monsoon months. In this region, rainfall is the only source of groundwater. Rainfall intensity is very high in the escarpment section and decreases gradually in the coastal region (Figs. 3 and 4).

Soil texture

There is a direct relationship between groundwater potential and soil texture of a region can be found. Soils which are moderately cohesive such as sandy loam and loam generally have higher groundwater potential. By contrast, soils which are highly compacted such as clay and silt generally have low groundwater potential due to the insufficient availability of soil pores.

Soil depth

Regions having very deep soil generally have higher groundwater potentiality as the water can infiltrate to the depth. Conversely, regions having shallow soil have lower potentiality of groundwater due to the massive bed rock, which partially works as an aquifuge layer, beneath the soil which prevent the water to infiltrate.

Well density

Before the technological advancement, the wells used to be built based on the availability of water in nearby wells which were built previously. Therefore, it can be observed that more number of wells are located only in the regions having higher yield of groundwater. Therefore, by including well density of a region, the extraction of different groundwater potential zones can be more accurate.

Groundwater potential analysis

The frequency ratio model is based on the relationship between the occurrence and their contributing area. The maps as a result of frequency ratio is being categorized into four groundwater potential zones—very high, high, moderate, and low (Fig. 5). Using this model, Savitri and Vashisthi basin show 9% and 17% of the total area are having very high, 30% and 32% are having high, 38% and 26% are having moderate, 23% and 25% are having low groundwater potential, respectively, (Fig. 6). The overall analysis shows that Vashisthi is contributing more areas with high groundwater potential compared to the Savitri basin.

Fig. 5
figure 5

The groundwater potential map of Savitri and Vashisthi basins. Notice the inset red and black dots which are well inventories used for training and testing the FR models. S1. Savitri (FR), S2. Savitri (IF), V1. Vashisthi (FR), V2. Vashisthi (IF)

Full size image
Fig. 6
figure 6

Bar graphs are showing the percentage of areas under different groundwater potential classes in Savitri and Vashisthi basins

Full size image

While the maps are prepared through influencing factor technique for Savitri and Vashisthi basin, they show that about 21% and 13% areas are covering very high, 33 and 34% are having high, 29% and 34% are having moderate, 18% and 19% area are covering low groundwater prospect, respectively, (Fig. 6).

Both catchments show a higher groundwater potential towards low-lying flat downstream section near the coast. The highlands in west and central hilly regions are characterized by very low groundwater potential because of the very steep gradient, the presence of massive basaltic hard rock, low drainage density, and shallow soil depth. Our study makes a notable observation that the regions with high-density lineaments are not entirely having very high groundwater potentials. This is probably because of the other factors such as lithology, soil cover, and gradient. However, these regions may influence the subsurface water recharge.

Validation

The area under the curve is a simple and convenient technique for validation of multi-criteria-based GIS models (Pradhan and Lee 2010; Bui et al. 2011). Validation of both models is done based on the area under the curve (AUC) method. To run frequency ratio, 70% of the total number of wells is used as training data and 30% is used for testing or validating the models. The same well inventory (30%) is used as a proxy of groundwater yield to validate the IF model. We have divided the output results (groundwater potential scores) into 20 equal classes. Then, we have plotted the cumulative area under different groundwater potential in “X” axis, while the cumulative number of wells in each class has been plotted in “Y” axis. Then, the AUC has been calculated using the following formula:

$$\mathrm{AUC}= \sum_{i=1}^{n=20}\frac{({X}_{1}+{X}_{2})}{2({Y}_{2}-{Y}_{1})},$$
(5)

where AUC denotes area under the curve, X indicates the cumulative percentage of different groundwater potential area, Y indicates cumulative percentage of wells, 1 and 2 designate two sequential points of data and n is the number of groundwater potential classes.

The AUC curve shows that Savitri and Vashisthi basins are having same efficiency (75%) in case of FR model (Fig. 7). By contrast, the accuracy of the IF model for Savitri and Vashisthi is 69% and 65%. Therefore, it can be clearly understood that the efficiency of FR model is better that IF model. Additionally, the accuracy of the IF model lies on the researcher’s knowledge of different parameters and assignment of scores which can lead to inaccurate results (Das 2019a).

Fig. 7
figure 7

Cumulative number of wells available in the study area are plotted against cumulative area under different groundwater potential zones for the assessment of accuracy of FR and IF models. The AUC for FR and IF is calculated for both the basins. The overall accuracy is higher in case of FR compared to IF technique in both basins

Full size image

Earlier, several studies were carried out to map groundwater potential zones utilizing the frequency ratio model (Manap et al. 2014; Al-Abadi et al. 2016, 2017; Das and Pardeshi 2018c). Almost all these studies indicated found fairly accurate results. Manap et al. (2014) applied frequency ratio to map groundwater potentials in Malaysia and found a high accuracy (85%). Groundwater prediction by Al-Abadi et al. (2016) showed the prediction rate of frequency ratio as high as 80%. The work of Balamurugan et al. (2017), which was applied in semi-arid India, showed a considerable groundwater map accuracy (77%). Das and Pardeshi (2018a, b, c) presented a comparison between influencing factor and frequency ratio and demonstrated higher reliability of the model while done through frequency ratio.

Besides AUC, we have compared the output results with published reports of CGWB of Raigarh and Ratnagiri district. It is found that the groundwater yield in the hilly regions is very low (< 0.5 lps), while in the central and coastal region, the yield is very high (3–10 lps). Therefore, this study can be considered as highly accurate for groundwater potential mapping of Savitri and Vashisthi basins.

Conclusion

The major aim of this study was to perform morphometric analysis and groundwater potential mapping of two semi-arid catchments in Maharashtra, India, utilizing frequency ratio and influencing factor techniques. Eight factors that influence the groundwater potential were assessed in GIS to built groundwater potential maps. The morphometric analyses suggested that Savitri and Vashisthi basins belong to mature or near-old stage of drainage development. The groundwater potential maps derived from the frequency ratio for Savitri and Vashisthi show that 9% and 17% are very high, 30% and 32% are high, 38% and 26% are moderate, 23% and 25% are encompassing low groundwater potential, respectively. On the other hand, the groundwater map predicted using influencing factor technique indicated about 21% and 13% very high, 33 and 34% high, 29% and 34% moderate, 18% and 19% low groundwater prospect for Savitri and Vashisthi basins, respectively. The comparison between FR and IF technique suggested that FR technique is more reliable (AUC = 75% for both basins) than the IF technique (AUC = 69% for Savitri and 65% for Vashisthi). This, among these two techniques, the FR has exceptionally high prediction accuracy for groundwater mapping. The outcomes of the present study might be helpful as fast-hand information to the local governing agencies and administrative bodies for the assessment, management, administration, planning, and sustainable utilization of groundwater and artificial recharge in the near future.