Introduction

Solar’s availability across the globe is leading to increased demand in solar energy systems worldwide [1, 27, 91]. The need for solar power plants is global, and many regions of the world are now developing this new renewable technology [5, 95]. Every renewable energy technology is at a distinct stage of development, research, and commercialization [87, 93]. They differ as well in terms of expected costs both now and in the future, resource availability, the existing industrial base, and any potential impact on greenhouse gas emissions [7]. Solar Energy is identified as one of the most environmentally friendly energy sources that does not contribute to or threaten global warming is solar energy [10, 62, 84]. It is unique in nature due to its free, clean, and profuse source with the ability to meet the rising energy demands of earth [14, 68, 92]. Also assures that social and economic success will occur without environmental degradation or climate change effects, which will result in growth and sustainable development [14].

India is a tropical nation, and because of its location, it is well suited to harnessing solar energy through the use of solar PV systems and large-scale solar plants Mishra et al. [53]; [27, 70]. The daily needs of mankind can be fulfilled if the energy from the sun could be stored, eliminating the need for dependence on fossil fuels [11, 32]. India is undergoing extensive urbanization, and an uninterrupted energy supply is the essential factor driving such demographic expansion and rising GDP in these metropolitan areas [70]. By 2030, there will be a predicted increase in primary energy demand of more than 60, or 1.7% annually [84]. Therefore, the energy produced by non-conventional energy sources causes resource depletion and environmental damage [70]. Fossil fuels, which are currently the widely used sources of energy, have led to health concerns leading to respiratory diseases, asthma, lung infection, and cancer [36]. In nature, fossil fuels are limited and can get exhausted in the future at the current rate of consumption. The idea of carbon-negative or low-carbon sustainable energy systems has been a rising issue in society due to the impending concerns of unchecked climate change [57]. Hence it is all the more important to look for alternate and cleaner sources of energy for sustainable development.

India has committed significant resources to boost solar energy production to fulfill the country's rising energy needs and lessen its reliance on coal-based energy generation [38]. One such effort by the Ministry of New and Renewable Energy (MNRE) to fulfil the need for energy holistically through the Solar City Mission (SCM) programme. The SCM initiative targeted to reduce about 10% of the demand for conventional energy in urban areas by the use of RE sources, therefore encouraging to use solar energy technology. Approximately 5000 trillion kWh of solar energy are incident across India each year. Over 5 kWh/m2 of global radiation is received annually over more than 56% of India’s entire geographical area [52]. The Indian government's supportive policies give economic incentives, enticing tariff plans, and tax exemptions, all of which aid in the nation's sustainable development [38].

Solar energy is relatively affordable, clean, and has low operating expenses, which makes it more accessible due to its cheap contraction costs and high rate of power line distribution [51]. Optimal selection of location for solar photovoltaic panel installation is crucial and there have been several studies in the past, trying to address the problem. Mapping technique is one of the most widely used methods. These techniques, along with developing technology have been combined as a single system as GIS. One of the most beneficial GIS applications is the tool that is GIS based [65, 89]. Such techniques use an understanding of the geographical distribution of metropolitan areas, along with historical data and field surveys, and they are very time and money efficient [81]. The literature has included efforts to incorporate renewable energy sources across the globe into GIS bases [8, 94]. Remote sensing (RS) and GIS have been employed in other investigations of RE sources, such as wind, solar, and hybrid power plants, to identify places that meet ecological requirements [44]. The impact of land topography, slope, and surface characteristics on the dispersion of solar radiation on the earth's crust [27]. The factors that need to be considered includes distance to road, the availability of arid land, the distance from water bodies, the slope, the direction of the sun’s radiation, the latitudes and longitudes, the neighbourhoods and the existing built-up area, etc. [68]. The overview of GIS and machine learning-based suitability assessment of global solar energy has been outlined in a tabular form (Table 1).

Table 1 The overview of GIS-based suitability assessment of solar energy in national and global scale
Full size table

To maximize the resources of solar energy, a variety of strategies have been employed in research to determine the optimal site placement for the installation of solar PV panels. It is believed that the map-based overlay analysis technique is awkward. Both the technique and the evaluation parameters lack independent validation [81]. Additionally, the parameters used for evaluation lack suitable standardisations. MIF is a more straightforward, dependable, and adaptable overlay analysis technique for assessing the expansion of urban infrastructure than the more complex Analytic Hierarchy Process (AHP) [54]. To characterize intricate methods for estimating and decision-making, the AHP technique, which is a pair-wise comparison method of factors used in multi-criteria selection and is principally based on a matrix eigenvalue, has been utilized [72]. Artificial neural networks and cellular automata-based techniques have been used in this process [47]. MIF, on the other hand, is a non-sensitive technique that relies on a precise set of assumptions rather than input data. AHP also employs pair-wise comparison, which was based on consistency ratios. A comparison matrix’s consistency should be less than 0.1 [79]. If there are any variations, the comparison questionnaire should be filled out more accurately to get near to the necessary consistency.

Modelling the geographical viability of solar energy resources employing GIS-based multicriteria decision analysis (MCDA) is one such approach [22]. The AHP is one among the MCDM methods which is been widely used in several research for installation of solar photovoltaic in suitable site [79]. The literature survey shows the lack of applications of MIF-based techniques that identify suitable site locations for the installation of solar PV, despite the fact that there have been numerous MIF-based studies conducted in the world in the areas of groundwater potential mapping [66], food evaluation [59], appropriateness of the site suitability for urban development [48], dam site suitability [63], ideal location for electric vehicle charging station [64], evaluation of the alluvial aquifer [86], estimating the possibility of land deterioration [48] and appropriate locations for water-saving technology [78]. The literature shows, there are limited studies with GIS-MIF integration in determining suitable site locations for PV solar plant installation [63]. It creates a nuanced layer to the decision-making process, ensuring a more precise and context-specific evaluation [47].

The major contribution of the proposed method lies in its innovative integration of GIS and the MIF technique for site selection of solar PV farms. The study addresses a critical aspect of renewable energy generation by systematically evaluating and delineating optimal locations for solar PV farms in Nashik, India. By considering a comprehensive set of influencing factors such as solar radiation, land surface temperature, relative humidity, wind speed, aspect, land use, distance from roads, elevation, and vegetation density, the proposed methodology goes beyond conventional approaches. The use of MIF for weight assignment adds a nuanced layer to the decision-making process, ensuring a more precise and context-specific evaluation. Furthermore, the study validates its findings with actual energy data generated from solar panels, demonstrating the practical utility and accuracy of the proposed approach. This comprehensive and validated methodology contributes significantly to the field of solar energy planning, offering a dynamic tool for decision-makers in both public and commercial sectors for identifying and delineating optimal sites for solar PV farms in diverse geographic areas.

Methodology

Study area

The study area is situated between latitudes 20°8′18.69"N to 19°48′50.72"N and longitude 73°36′17.10"E to 73°49′3.69"E. This study area (Fig. 1) has a total area of 890.35 km2. The elevation of study area varies from 530 to 1068 m above mean sea level and situated in Maharashtra's North-West. The research area experiences substantial variability in its yearly rainfall, which is roughly 713.50 mm on average [3]. In the study area, spatiotemporal average temperatures ranges from 20.02 to 36.92 degrees Celsius and yearly rainfall ranges from 500 to 3400 mm. The temperature gets quite chilly throughout the winter, with low as 2 degrees Celsius [37]. The majority of Nashik's urban population lives close to the Ghats [40]. Spatio-temporal mean wind speeds in study area were 2.40–2.61 m/s and mean relative humidity was 59.86–62.77%, with minimum values occurring in the warmer months. Based on aridity indices, the region's climatic characteristics are described as semiarid [31].

Fig. 1
figure 1

Location of study area with Land Surface Temperature (LST)

Full size image

Energy issues and rationale of the study area

The daily coal need for Nashik is between 12 and 13 thousand million tonnes. Western Coalfields and South-eastern Coalfields are connected to Nashik for the supply of coal. Due to Nashik’s remote location from the collieries, the freight component is significant and typically accounts for 80–100% of the cost of coal [77]. Since the hydro power plants are dependent on rain, the electricity situation in the summertime gets affected [9, 90]. In Western Nashik part, new construction is taking place in the commercial, residential, and institutional sectors. As a result, the western area can be seen as an energy-forward metropolis where solar energy resources have the potential to be used to lower energy consumption. By utilizing green building methods, new structures can lower their energy requirements [12, 43, 76]. By installing solar water heating systems and switching to energy-saving appliances, individuals can actively participate in city energy management. Building Integrated Photovoltaics (BIPV) systems enable commercial and institutional buildings to produce their own electricity. The commercial sector in western Nashik has the greatest daily energy needs, followed by institutional, at 4569879.6 and 1065240 kW, respectively. The average daily demand for the residential sector is 443139.84 kW. Street lighting requires 14919 kW, while other types of lighting, such as signal lighting, hoarding lighting, and park lighting, need 7959.5 kW [77]. The industrial Estates of the Maharashtra Industrial Development are where most of the industrial activity in the Nashik Municipal Corporation area is primarily concentrated namely, the Ambad Industrial Estate and the Satpur Industrial Estate. There is a chance that businesses producing solar technology will be included in this well-planned and established industrial area [34, 76]. Moreover, the importance of switching to alternative energy sources has increased due to the increasing depletion of fossil fuels in Maharashtra State [56, 79]. The demand for energy across the Maharashtra state has been increasing in recent years due to the rapid growth of the population [35, 43, 67]. PV is foremost effective method of producing electricity [15, 74]. In addition to preventing an energy catastrophe, this also assists to lessen carbon footprint and protect the environment [33].

Data acquisition and preparation

Several influencing factors, such as solar radiation, wind speed, land surface temperature, relative humidity, vegetation, elevation, land use, aspect, etc., control the ideal site location for solar PV power plants. The elements that affect the selection of the best location for solar photovoltaic farms were taken into account by taking into account the effects of the parameters as well as the field observed and geospatial datasets availability. In the present study, nine geo-environmental, and geo-physical factors were selected to evaluate the optimal site for solar PV farms, and for each influencing factor, the thematic layers were prepared. The land use layer was prepared using support vector machine supervised classification approach. The primary reason for using supervised classification rather than the CORINE dataset was to use up-to-date datasets [73]. The vegetation density was prepared using Normalized Difference Vegetation Index (NDVI) from the Landsat-8/OLI sensor datasets and sum to prepare the vegetation thematic layer. The ArcGIS spatial analyst tool was used to develop aspect and elevation layers using the 30 m Shuttle Radar Topography Mission (SRTM) DEM data. The gathered toposheets from Survey of India (SOI) were employed to digitize the distance from the road network in the study area and checked utilizing Landsat-8/OLI imagery. The LST was prepared using the thermal datasets (band-10) of Landsat-8/OLI acquired from the U.S. Geological Survey. The energy data of eleven solar PV power plant sites were collected from the solar energy facility owner and supplier. In addition to above geospatial data, solar radiation, wind speed, and relative humidity data were extracted from the China meteorological assimilation driving datasets and merged to prepare the thematic layers.

Demarcation of optimal sites for solar PV farms

The integration of GIS-MCDM in site selection studies for solar PV farms has been accomplished in a number of research in the literature. The weight or importance of the factor/criteria is typically determined from the literature in MCDM methods. The MCDM methods that are often utilized in the literature are the analytical hierarchy process (AHP), fuzzy logic index models, ranking based, frequency ratio, best–worst, certainty factor, weights-of-evidence, and multi-influencing factors. The AHP introduced by Saaty [72] is a superb and well-known technique that is frequently employed to produce an adequacy map. The AHP already has been regarded as among the most significant methods for determining weights. The AHP involves the computation of the eigenvectors and eigenvalues of a square preference matrix of order “K” that includes the complete preference information of all potential constellations of pairings of the k criteria. For members of interest groups and decision-makers, it could be challenging to explain how the principal eigenvector is used to derive criteria weights [55]. The MIF technique is particularly useful for estimating the site of solar PV power plants since it considers the interrelationship among the influencing factors and offers an economical, quick, and accurate evaluation. The MIF technique is well-fitted for incorporating multiple factors that influence solar energy potential, such as land availability, slope, distance to infrastructure, and environmental constraints. It allows for a comprehensive assessment of site suitability [48]. The MIF technique is flexible in its application, and can be adapted to different geographical contexts and data availability. It can also be easily updated as new data becomes available. The MIF technique is a transparent and reproducible method that means the decision-making process is clearly documented and can be replicated by others. Nevertheless, its application in GIS-based integrated site selection applications for polar PV power plants is quite restricted. However, a decision-making process based on an interrelationship between influencing factors can be made mathematically more simpler using the MIF technique. The spatial correlations between both the independent and dependent variables are determined by the MIF techniques. It is developed based on the score assigned for the important and minor relevant factors that influencing the suitable solar PV power plant sites. The MIF approach basically follows the procedures which is illustrated in Fig. 2. First, each factor is allocated a rank based on its relative importance to a suitable sites for solar PV farm location. In present study, the major and minor interrelationship between the influencing factors found out using extensive literature survey [11, 17, 23, 24, 29, 58, 71, 75, 85, 91]. The more significant the factor, the higher its weight. The major factors were given a weight of 1, while the minor factors were given a weight of 0.5. The sum of all weights from each element determines the relative score (\({W}_{{M}_{i}}\)) of a factor impacting solar PV power plant potential [64]. The significance of each factor (\({W}_{i}\)) was then determined using Eq. 1 based on the relative of the scores.

Fig. 2
figure 2

Relation between the MIF of optimal solar PV farms

Full size image
$${W}_{i}=\frac{{W}_{{M}_{i}}}{{\sum }_{i=1}^{9}{W}_{{M}_{i}}}\times 100,{W}_{{M}_{i}}=minor\,effect+major\,effect$$
(1)

Figure 2 shows the relation among the multi-influence factors of suitable solar PV farms site. The results obtained with MIF are presented in Table 1.

Development of thematic factor layers

It is essential to make thematic numerical maps of the factors before performing PV farm site selection with GIS-MCDM. Nine thematic layers that considers the relative effects of the influencing factors were constructed to assess the best locations for solar PV farm in the study area. Solar radiation, wind speed, land surface temperature, relative humidity, vegetation, elevation, aspect, and land use were chosen as affecting factors by considering the effects of the components and the availability of field observation with spatial variations. The most prevalent in the literature and freely available for the study area were selected while choosing the relevant factors (Table 2).

Table 2 influence factors with interrelated effects and weights
Full size table

Solar radiation: The efficiency of higher solar radiation regions is better suited for solar PV power plants. The efficiency is impacted by high solar radiation, and it is crucial to use the data provided for the area because solar radiation, such as the duration of sunshine, differs by state. The amount of solar radiation that is absorbed by the environment is called solar irradiance, and it is measured in units of area by a particular surface (MJ/m2) [24, 30]. The reclassified Solar radiation map is presented in Fig. 3.

Fig. 3
figure 3

Distribution of solar radiation in study area

Full size image

Land surface temperature (LST): An important indicator that reflects the impact of solar PV farm on the local environment and the impact of temperature on the farm is land surface temperature. During the day, each solar panel in the power plants converts energy into electricity, however, the electrical converter and convection equipment operate as heaters for the panels. A rise in land temperature from the outside will raise the optimum working temperature of the panels, resulting in a loss of performance. Also, the solar panels can warm up from behind thanks to the surface's quick heat radiation [28, 88]. Figure 1 shows the LST map that has been reclassified. The steps listed below are mainly used to prepare the LST map.

  1. 1)

    B10 from Landsat-8 OLI was used to calculate top of atmospheric spectral radiance.

    $$L\lambda ={M}_{L}\times {Q}_{CAL}+{A}_{L}$$
    (2)

where, \(L\lambda\) is top of atmospheric spectral radiance, \({M}_{L}\) multiband radiance, \({Q}_{CAL}\) is the band 10 image, and \({A}_{L}\) is the band-specific additive rescaling factor.

  1. 2)

    To estimate the brightness temperature, radiance is transformed to at-sensor temperature through the thermal constants given in Landsat-8 metadata.where, \(BT\) refers to brightness temperature, the band-specific thermal conversion constants from metadata are denoted by the \({K}_{1}\) and \({K}_{2}\). To obtain the data in °C, the radiant temperature is adjusted by adding absolute zero.

    $$BT=\frac{{K}_{2}}{\mathit{ln}\left(\frac{{K}_{1}}{L\lambda }+1\right)}-273.15$$
    (3)
  1. 3)

    Vegetation Proportion Calculation.where, where the \(NDV{I}_{s}\) and \(NDV{I}_{\upsilon }\) values are the soil and vegetation pixel thresholds.

    $$p\upsilon ={\left(\frac{NDVI-NDV{I}_{s}}{NDV{I}_{\upsilon }-NDV{I}_{s}}\right)}^{2}$$
    (4)
  1. 4)

    Using an NDVI-based relationship to predict surface emissivity.where, \(\varepsilon\) is land surface emissivity based on NDVI.

    $$\varepsilon =0.004\times p\upsilon +0.986$$
    (5)
  1. 5)

    Computation of land surface temperature.where \(LST\) is the land surface temperature in °C, λ is the wavelength of radiance emitted, is the computed emissivity based on the proportion of vegetation (\(p\upsilon\)) [48].

    $$LST=\frac{BT}{\left(1+\frac{\lambda \times BT}{14380}\right)\mathit{ln}\varepsilon }$$
    (6)

Relative humidity: The performance of the project's yearly energy yield and the solar PV power plants' long-term deterioration rate are both somewhat impacted by the ambient air humidity. The term "air humidity" refers to the water vapor conveyed by the air. Relative humidity is a measurement of humidity factor, including the quantity of water vapor in the atmosphere. The amount of moisture that can be carried increases as temperature rises. High relative humidity areas are not suitable for solar energy generation. It increases the likelihood of rain because the water's attempt to mix with the surrounding air will lead the vapor to condense and turn into liquid [26, 88]. The reclassified relative humidity map is presented in Fig. 4.

Fig. 4
figure 4

Distribution of relative humidity (RH) in percentage

Full size image

Wind speed: Wind direction and speed are two of the most important factors for reducing solar PV farms' temperatures. To provide safer solar PV farms designing and operating conditions, wind speed is a significant factor. It does not have a strong wind cooling function since the wind direction in the study area is generally the wind that conveys hot air [21, 88]. The reclassified wind speed map is presented in Fig. 5.

Fig. 5
figure 5

Wind speed variation in study area

Full size image

Aspect: Land aspect indicates the slope direction. In other words, it impacts a location's angle with respect to the sun, which in turn impacts how long it is sunny. The best solar PV power plants are chosen to have the south direction in the aspect maps made using DEM because, depending on the season, the south-facing sites in northern hemisphere countries receive lots of sunlight. The aspect map generated from the DEM depicts aspects in 10 distinct directions [18, 25]. The reclassified aspect mapping is presented in Fig. 6.

Fig. 6
figure 6

Aspect shows the slope direction in study area

Full size image

Land use: The solar PV farms construction requires the determination of suitable and unsuitable land use. Solar PV farms are not appropriate for all types of land. In some countries, for example, it is forbidden to cut down trees such as olives, tea, and hazelnuts, and any construction in these regions is prohibited. Furthermore, soil types are a consideration that has a direct impact on the cost. While hard floors raise the cost of power plants, slippery floors inflict more damage to power facilities during natural catastrophes such as earthquakes and landslides. Land use maps were classified into five different land-use classes in the study using the support vector machine classification approach. The main reason for using supervised classification because of its worldwide acceptability. The reclassified land use land cover map is presented in Fig. 7.

Fig. 7
figure 7

Land use/land cover scenario in the study area

Full size image

Euclidean distance from road: Roads and railroads, as well as other forms of transportation, are vital factors for the development of every region. It is expected that the road network should be established to lower costs since the installation of solar PV power plants requires a significant quantity of building, supplies, and operation to the region [23, 88]. The reclassified road network map is presented in Fig. 8.

Fig. 8
figure 8

Euclidean distance from the major roads

Full size image

Elevation: Precipitation and temperature have a relationship with elevation. Solar radiation increases as altitude increases due to a decrease in the thickness of the atmospheric layer, and the efficiency performance of solar panels increases due to a decrease in temperature at high altitudes. However, installing photovoltaic farms at high elevation is not suggested because the rise in altitude impedes transportation and raises project costs [26, 83]. The reclassified elevation map is presented in Fig. 9.

Fig. 9
figure 9

Elevation map of the study area

Full size image

Vegetation health: The significance of land use has already been stated. Building a solar PV farm in areas with poor vegetation health, on the other hand, is a poor decision. As a result, detecting and investing in areas with low plant health is critical for sustainable agriculture and the protection of natural life. As a result, it was calculated using Eq. 7 and classified into five classes based on NDVI Landsat 8 OLI data describing vegetation health [18, 38]. The reclassified vegetation map is presented in Fig. 10.

Fig. 10
figure 10

Vegetation density map showing the vegetation cover in the study area

Full size image
$$NDVI=\frac{(NIR-\mathit{Re}d)}{(NIR+\mathit{Re}d)}$$
(7)

Sensitivity analysis

Sensitivity analysis is a valuable tool in identifying the significance of important factors. Sensitivity analysis is the process of examining how changes in the input variables affect the analysis's results. Sensitivity analysis can help increase the model's accuracy in situations where the ambiguity of the inputs prevents accurate decision-making. In this research, the sensitivity analysis method was utilized to verify and assess the effectiveness of the MIF results and rank the dominancy of the factor. Map removal sensitivity is a measure of the sensitivity of eliminating one or many variables from the suitability analysis in the resulting suitability mapping. By eliminating any factors and generating a new suitability map each time, map removal is a crucial sensitivity analysis for assessing the possible suitability of a solar PV power plant site. Thus, the relative importance of the factors is determined [42]. The mathematical notation of map removal is given in Eq. 8.

$$SM=\frac{\left|\left(\frac{PSA}{N}\right)-\left(\frac{PS{A}{\prime}}{N{\prime}}\right)\right|}{PSA}\times 100$$
(8)

where the total number of factor layers utilized to create Photovoltaic Sensitivity Analysis (PSA) is N, while the number of factor layers considered to create PSA is N’, SM refers factor related sensitivity index, the analysis for suitability of the solar PV farms by all factor layers is represented by PSA. Finally, Solar PV power plant site selection through MIF technique has been shown in a flowchart (Fig. 11).

Fig. 11
figure 11

Flowchart of solar PV farms site selection using the MIF

Full size image

Results

Characteristics of spatial variations of the influence factor

The evaluation of a suitable site for optimal site for solar PV farms is a critical component of renewable energy generation. Following is the geo-environmental, and geo-physical factors that influence optimal site for solar photovoltaic power plant. The integration of these influencing factors determines the degree of suitability and useful for identifying and delineating the unsuitable zone. The different thematic layers are designed to achieve the objective. GIS was used in this study to create nine different factor maps, the weights of which were calculated by MIF, to aid select highest and lowest suitable site. Five classes—all but aspect—have been used to classify the created maps. The class dynamic ranges, geographic areas they cover, and the percentage distributions of the factors are shown in Table 3. Since Nashik, India, is primarily flat, there aren't many areas with significant elevation. There are few places where transportation is poor because the road network exists, even if it is only one lane. Between 59 and 63% is the range for relative humidity, which is still a high level. In contrast, wind speed is not a significant impact in urban areas. Looking at the city as a whole, many regions are covered with vegetation, and many of these areas have high NDVI values, indicating that they are in generally good health. The Kriging method, which is deterministic spatial interpolation, was used to create maps of solar radiation, relative humidity, and wind speed. Multiple buffer analysis was performed while preparing the road network map.

Table 3 Thematic layers with their classes and areas occupied
Full size table

Delineation of optimal site for solar PV farms in study area

The most crucial factor that directly influences the outcomes of site selection procedures using GIS-MCDM investigations is weight assignment. The primary argument for MIF's use in this study is the lack of widespread MIF utilization in site selection studies for solar PV farms. The literature and the expertise of the paper's authors were utilized to determine the degree of importance of the factors. Weight assignment results performed with MIF are presented in Table 4. The weight of solar radiation is largest, whereas the weight of land usage and land cover is lowest. There is equal weighting given to relative humidity, aspect, and elevation. The suitability map, whose weighted sum is the final outcome, is affected differently by them because of the differences in their total weights.

Table 4 Factors relative weights with normalized and total weights for each features
Full size table

The weights of the factors determined by the MIF approach were used to construct a weighted layer map as part of the process of creating the reclassification suitability map (Fig. 12). For the relative humidity map, for instance, the features of the thematic layer range shown in Tables 3 and 4 were reclassified after digital maps using the Kriging method based on measurement points were constructed. The method of reclassification differs depending on the kind of data type that was utilized to create each factor map. In other words, making an elevation map does not require interpolation. The weighted sum of the factors using the weights generated with MIF was then used to create suitability maps shown in Fig. 11. The suitability maps in this study were quartile-divided into five grades. Final suitability maps were reclassified as very good, good, moderate, low, and very low. Spatial distributions of these areas are presented in Table 5. The study region is appropriate for the solar PV farms installation, according to the general abundance of good and very good areas. In comparison to other suitability fields, there are comparatively few very low fields. This is another indication of the Nashik’s suitability. The unsuitable regions of Nashik are in the south and west. Despite having high elevations, these areas are humid, have transportation issues, and have agricultural areas. The northeast, in general, is the strategic site due to the region's high levels of solar radiation, strong wind speeds, low levels of humidity, efficient transportation, and abundance of low-plant unhealthy areas.

Fig. 12
figure 12

Delineated optimal sites for in solar PV farms in study area

Full size image
Table 5 Optimum site selection statistics through MIF model
Full size table

The optimal sites of solar PV power plant delineated using MIF technique revealed that ‘very low’ suitability of site for solar PV power plant covering 4.866% (43.370 sq.km.) of the study area, ‘low’ suitability of site for solar PV power plant 13.190% (117.553 sq.km.), ‘moderate’ suitability of site for solar PV farm 31.640% (281.988 sq.km.), ‘good’ suitability of site for solar PV farm 32.347% (288.292 sq.km.), and ‘very good’ suitability of site for solar PV farm encompassing 17.957% (160.036 sq.km.) of the study area.

Sensitivity analysis

Because the thematic layer removal evaluates the sensitivity of processes between factor layers and demonstrates the value of each layer for delineating the potential site for solar PV farms in Nashik, map removal technique is chosen and used in this study. Sensitivity test results are presented in Table 6. According to the results of the sensitivity analysis, the most precise estimate is obtained when all nine factors are considered. It is true that solar radiation and relative humidity have a significant role in the decision to construct a solar PV farm. MIF gave the least weight to factors including land use, and distance to road, and sensitivity analysis confirmed that this was the case. Additionally, it has been found that choosing the best solar PV power plant site is not greatly influenced by wind speed, although having a relatively large weight compared to other criteria.

Table 6 Sensitivity analysis statistics with variation index
Full size table

Validation of the predicted solar PV farms sites

The validation of the delineated solar energy potential zones is conducted using the actual energy data generated through the solar panels system. The efficiency of the solar farms is compared with the delineated potential zones of solar PV farms. The solar energy capacity of the plant in the considered sites ranged from 3 to 15 KW. The components of the solar plant in the study area encompass solar PV panels, an inverter, Alternating Current Distribution Box (ACDB) and Direct Current Distribution Box (DCDB), earthing electrode, structure, and a solar PV power meter. The solar PV panels consist of polycrystalline modules. Each solar PV module, with a power range of 270 to 310 Watts, is comprised of 72 cells in series, featuring an X-Pitch of 938 mm and a Y-Pitch of 1155 mm. The dimensions of the solar PV module measure 1955 × 989 × 40 mm, with a weight of 28 kgs. The module's efficiency varies from 14.0 to 16.0. The maximum output power (PMP) spans from 270 to 310 Watt, open-circuit voltage (Voc) ranges from 43.2 to 44.0 Volts, short-circuit current (ISC) ranges from 8.21 to 9.11 Amperes, maximum power voltage (VMP) spans from 36.0 to 36.5 Volts, and maximum power current (IMP) varies from 7.52 to 8.50 Amperes. The solar cells within the module consist of 156 × 156 mm high-efficiency polycrystalline 3 bus bar cells. The junction box is equipped with 3 bypass diodes. The frames feature 15 microns silver-anodized aluminum frames, while the front glass is constructed from 3.2 mm low-iron tempered glass. The energy data of 11 solar PV power plant show that 1 ‘Very high’ (97–100% efficiency) solar PV power plant out of 1 energy site falls under ‘very good’ potential zone for solar PV farms and 3 ‘high’ (93–96% efficiency) solar PV power plant out of 3 energy site falling under ‘good’ suitable site for solar farms. Moreover, 2 out of 2 ‘medium’ (89–92% efficiency) solar PV power plant falling in ‘moderate’ suitable site for solar power plant, 3 out of 4 ‘poor’ (85–88% efficiency) solar PV power plant falling under ‘low’ suitable site for solar power plant and 0 out of 1 ‘very poor’ (80–84% efficiency) solar PV power plant falling under ‘very low’ suitable site for solar power plant. The number of solar PV power plant that agreed with actual energy data are 9, and the number of solar PV power plant that disagree with the actual energy data are 2. As a result, the MIF technique's accuracy for solar PV power plant delineation in the research area is 81.80%. To validate the delineated solar PV farms more comprehensively, the Receiver Operating Characteristic (ROC) analysis has been conducted [48, 49]. The Area Under Curve (AUC) of ROC for the predicted site mapping is shown in Fig. 12. The value of the AUC for the delineated site was 0.839 shows the acceptable performance of the MIF technique applied in study area (Fig. 13).

Fig. 13
figure 13

ROC curve shows that the AUC value has successfully checked the error, and validated the MIF model

Full size image

Discussions

The significance of this study can be described that the interrelationship among the multi-influence factors by integrating GIS and MIF as well as validation using actual solar energy data with ROC analysis. This comprehensive study is utilized to present alternative locations for solar PV farms in Nashik using the most updated geographic information for the design and development of solar PV power plants in India. The planning and design method used in this research study could be used to install solar PV power plants in a number of study regions. Recently, some of the studies related to solar energy farm development across the Globe [2, 16] as well in India [80, 82] have executed for renewable energy mission. Nowadays, solar energy development and suitable site assessment is very much required to enhance the capacity of the renewable resources especially for the highly population-based countries like India, Bangladesh, and so others. Doljak, & Stanojevi´c [14] implied a GIS-based natural process for the selection of solar PV power plant. Similarly, this study also tried to delineate the solar PV farms in Nashik using different geo-physical and environmental parameters. In western part of Nashik, different commercial, residential, and institutional activities impacted on the high solar energy resources consumption that needs to be reduced into lower energy consumption. Otherwise, this problem can be shorted out by utilising green building methods, new structures can lower their energy requirements [12, 43, 76]. Even, by installing solar water heating systems and switching to energy-saving appliances, individuals can actively participate in city energy management [76]. Building Integrated Photovoltaics (BIPV) systems enable commercial and institutional buildings to produce their own electricity [77, 80]. Table 7 is used for better understanding of the GIS-based integrated results of MIF techniques.

Table 7 Details study of sensitivity analysis statistics with variation index
Full size table

Although, GIS-MCDM methods offer convenience and flexibility, it should not be ignored that there are some limitations that must be taken into consideration when carrying out the process [13, 19, 20]. The primary drawback of GIS-based site selection is the cost involved when high spatial and temporal resolution data are required [46]. GIS data collection and analysis may call for specialized tools and software. For this reason, when collecting data, only reliable data suppliers should be used. GIS data often tends to become outdated (i.e., land use data, LST, and vegetation density, etc.). This means that the data used for site selection become obsolete and need to be updated over time.

Conclusions and policy making approach

This research attempted to provide a practical methodology for identifying the solar PV farms’ suitability index by integrating the interrelationship-based MIF technique with RS and GIS. Applying the MIF technique with nine influencing factors; solar radiation, wind speed, LST, relative humidity, vegetation, elevation, land use, Euclidean distance from road, and aspect, were assigned to identify optimal locations of solar power farms. Using the MIF process, weights were allocated to the thematic layers, and rankings were given to the criteria of each layer. Based on the site suitability index, the optimal location of the site where solar PV farms could be installed was delineated. The major findings of this study are given below:

  • The optimal sites of solar PV power plant delineated using MIF technique revealed that ‘Very low’ suitable site for solar PV farms covering 4.86% and ‘low’ suitable site 13.19% in the southern section of the study area. Followed by ‘moderate’ suitable site 31.64%, ‘good’ suitable site 32.34%, and ‘very good’ suitability of site for solar PV farms encompassing 17.95% in the north and north-east section of the study area.

  • Sensitivity analysis was employed in this study to assess how the dominant factor influence the solar PV farms site selection, according to this study, solar radiation, relative humidity, and elevation are more sensitive as well as the dominant factors for optimal site selection in the study area.

  • The validation of the delineated solar energy potential zones is conducted using the actual energy data generated through the solar panels system. The prediction accuracy of solar PV power plant prediction using MIF technique in the study area found to be 81.80%. Moreover, the ROC analysis results showed that the MIF approach achieved AUC of 0.839 was acceptable in delineating an accurate mapping of Solar PV farms for the Nashik City.

However, the integrated GIS-based planning method adopted for the study area has the potential to offer inclusive and practical advances for different geographic areas. The proposed methodology presents several notable advantages. Firstly, the integration of GIS and the MIF technique provides a systematic and comprehensive approach to site selection for solar PV farms. The consideration of multiple influencing factors ensures a nuanced evaluation, capturing the complexity of the environmental and geographical context. Additionally, the validation of the delineated solar energy potential zones with actual energy data underscores the practical utility and effectiveness of the proposed approach.

In the policy perspectives, the major contribution of the proposed method lies in its innovative integration of GIS and the MIF technique for site selection of solar PV farms in Maharashtra, India. The study addresses a critical aspect of renewable energy generation by systematically evaluating and delineating optimal locations for solar PV farms in Nashik, Maharashtra, India. By considering a comprehensive set of influencing factors such as solar radiation, land surface temperature, relative humidity, wind speed, aspect, land use, distance from roads, elevation, and vegetation density, the proposed methodology goes beyond conventional approaches. The use of MIF for weight assignment adds a nuanced layer to the decision-making process, ensuring a more precise and context-specific evaluation. Furthermore, the study validates its findings with actual energy data generated from solar panels, demonstrating the practical utility and accuracy of the proposed approach. This comprehensive and validated methodology contributes significantly to the field of solar energy planning, offering a dynamic tool for decision-makers in both public and commercial sectors for identifying and delineating optimal sites for solar PV farms in diverse geographic areas in Maharashtra. Additionally, the use of MIF for weight assignment enhances the accuracy and relevance of the decision-making process, contributing to a more reliable identification of optimal sites in a large-scale. Finally, it can be explicated that in the renewable energy contexts, promoting a broader application of the GIS-MIF integration for sustainable energy planning in the field of solar PV farms energy-based research to achieve the UN’s SDG-7 about clean and affordable energy.