Environ. Process. DOI 10.1007/s40710-017-0257-4 O R I G I N A L A RT I C L E
GIS-Based Spatial Distribution of Groundwater Quality and Regional Suitability Evaluation for Drinking Water Kuldeep Tiwari 1 & Rohit Goyal 1 & Archana Sarkar 2
Received: 15 May 2017 / Accepted: 14 July 2017 # Springer International Publishing AG 2017
Abstract Hydrochemistry mapping of groundwater is one of the important aspects of water resources management of an area based on usages such as drinking water or irrigation, and their quality requirement. Such mapping could lead to optimal utilization and protection from quality deterioration. In this paper, a geospatial based water quality index is developed for preparing water quality class suitability map. The aim of this study is to provide an overview of the spatial variation of groundwater quality parameters, i.e., Fluoride (F_), Nitrate (NO3_), Chloride (Cl_), Total Dissolved Solids (TDS), pH and Total Hardness (TH) in the Khushkhera - Bhiwadi - Neemrana Investment Region (KBNIR). Groundwater samples were collected from 14 locations, tested in the laboratory and were analyzed using Geographical Information Systems (GIS) techniques. Geospatial analyst tools were used to generate various thematic maps, and interpolation techniques were applied to identify the spatial distribution of groundwater quality parameters. Groundwater quality was analyzed in detail and compared with BIS and WHO water quality standards. Multivariate statistical techniques, such as Correlation Matrix Analysis and Principal Component Analysis (PCA) were carried out. The correlation coefficient showed TDS highly correlated with chloride (r = 0.725), and a negative correlation (r = −0.656) between pH and nitrate. It was observed that 36.02% of the populated area depends on excellent quality groundwater and 23.56% on good quality groundwater, while poor and moderately poor quality groundwater is available for respectively 6.26% and 34.16% of the populated area. Also, only 65.88% of the total geographical area within the KBNIR region is hydrochemically suitable for drinking purposes. Attempts are also made to compare these results with PCA results. Keywords Hydrochemistry . Groundwater quality . GIS mapping . Kriging, strip plot
* Kuldeep Tiwari
[email protected]
1
Department of Civil Engineering, Malaviya National Institute of Technology, Jaipur, Rajasthan, India
2
National Institute of Hydrology, Roorkee, Uttarakhand, India
Tiwari K. et al.
1 Introduction Groundwater is the readily available source of fresh water for living organisms on the earth for their survival (Tiwari et al. 2015a). The suitability of groundwater for a particular use depends on its quality. The groundwater quality assessment is necessary to ensure its optimal and sustained safe use. There are different standards of water quality depending on the type of water usage (Babiker et al. 2007). Rajasthan is one of the deficient states of India with respect to groundwater quality and storage. Despite this, groundwater is an important source of drinking water in Rajasthan and fulfills 91% of the drinking water requirements of the state (Munoth et al. 2015). The groundwater quality depends on many natural processes and varies in time and space (Babiker et al. 2007; Singh et al. 2013). Sustainability of groundwater quality and quantity is essential for drinking, domestic and irrigation purposes. In view of water security of drinking water, the water quality assessment is of high importance. It is also important for ecological integrity and sustainability. In recent years, groundwater quality has become a major issue of concern in Rajasthan. According to the CGWB (2010), the concentration of Fluoride (F¯), Chloride (Cl¯), Nitrate (NO3_), TDS and Iron (Fe) were found to be higher than the other physiochemical parameters of groundwater in most of the districts of Rajasthan state. Many other studies have also reported in the literature, that the groundwater of Rajasthan is affected by these and several other parameters (e.g., Hussain et al. 2010; Jagtap et al. 2012; Machiwal and Jha 2015; Mitharwal et al. 2009; Mudgal et al. 2009). The developing cities of Rajasthan are facing the problem of deteriorating groundwater quality due to several reasons, such as inappropriate water management, anthropogenic activities, acidic precipitation, urbanization, massive use of fertilizers and pesticides in agriculture and various industrial activities (Bhaduri et al. 2009; Boateng et al. 2016; Singhal and Goyal 2014). Assessment of groundwater quality is a complicated process that depends on use of appropriate statistical tools for its assessment. Groundwater geochemistry has been assessed by various authors using different methodologies, such as multivariate statistical analysis for determining physiochemical properties followed by Hierarchical Cluster Analysis (HCA) (Banoeng-Yakubo et al. 2009; Boateng et al. 2016; Dehghanzadeh et al. 2015) and Principal Component Analysis (PCA) to classify groundwater samples into clusters for further analysis (Amiri and Nakane 2009; Bodrud-Doza et al. 2016; Gulgundi and Shetty 2016; Rajput and Goyal 2017; Ravikumar and Somashekar 2017). These techniques are applied on large number of samples for pattern recognition and to explain the variance of a large set of intercorrelated parameters. They reduce the dimensionality of the dataset by indicating association between different parameters. In this, each separate cluster represents specific water quality concerns (Ravikumar and Somashekar 2017). Another method, such as correlation matrix (CM)(Bodrud-Doza et al. 2016; Kumar et al. 2015) is performed on the physio-chemical parameters to asses the type and strength of relations between different quality parameters. Few others have developed water quality indices (WQIs) (Bodrud-Doza et al. 2016; Gorai et al. 2014; Tiwari et al. 2015b; Nag and Ghosh 2013; Kumar et al. 2015) for evaluation of hydrochemistry of groundwater. WQI is an effective method for evaluating drinking water quality and suitability in any area (Banoeng-Yakubo et al. 2009). Multiple Linear Regression (MLR) models were also applied by Amiri and Nakane (2009) to predict the level of water quality variables using compositional and spatial attributes of land cover. Gorai et al. (2014) developed a fuzzy hierarchical model for the prediction of water quality index based on fuzzy reasoning. According to Babiker et al. (2007), the sensitivity analysis model indicated that
GIS-Based Spatial Distribution of Groundwater Quality and Regional...
parameters which reflect comparatively lower water quality and significant spatial variability could be mapped. Groundwater quality parameters have also been analysed using various types of statistical distribution diagrams like Piper trilinear diagram (Dehghanzadeh et al. 2015; Shanmugasundharam et al. 2015; Singh et al. 2013; Tiwari et al. 2015b), Wilcox diagram (Dehghanzadeh et al. 2015; Nag and Ghosh 2013; Shanmugasundharam et al. 2015; Singh et al. 2013), Gibbs plot (Kumar et al. 2015; Shanmugasundharam et al. 2015; Singh et al. 2013), box plot (Boateng et al. 2016) and scatter plot (Dehghanzadeh et al. 2015; Singh et al. 2013) to understand the hydrochemical processes occuring in the groundwater system, which have resulted in the observed spatial and temporal variation in the quality of groundwater. In the present work, stripe plot diagram has been used to represent statistical distribution of different groundwater quality parameters. The same is also done with principal component analysis, a well known method for groundwater quality assessment. Attempts are made to compare the results using a GIS-based methodology. Remote sensing and geospatial tools can be used for water resources management strategies and planning (Singhal and Goyal 2014). Various researchers have applied geostatistical analysis using kriging and Inverse Distance Weighting (IDW) interpolation technique (Goyal et al. 2010; Jat et al. 2009; Shanmugasundharam et al. 2015) to identify pixel-wise value of the parameter and perform zonal statistics for assessment at zones levels. Zones could be based on administrative boundaries or geophysical boundaries. GIS is an important and effective tool for integrating spatial and temporal data with other information. It allows analyzing and handling problems about various quality parameters in their spatial extent. Some studies have reported on the use of GIS tools and remote sensing in groundwater quality mapping in different land uses (Ding et al. 2016; Nas and Berktay 2010). In the present work, not only groundwater quality assessment has been carried out by different statistical tools but groundwater quality is also mapped with spatial distribution of population to understand percentage of population falling under unacceptable groundwater quality area. This could be useful to planners for water resources planning of the area. The study area is the Khushkhera - Bhiwadi - Neemrana Investment Region (KBNIR), which is currently undergoing intense development. Rapid urbanization and industrialization will directly or indirectly pollute the groundwater. The main objective of the current study is to present the dependency of populated area on a particular water quality class as a new methodology, and also to analyse and interpret groundwater samples collected from various locations of KBNIR area to assess the groundwater quality. Followed by interpretation of analysed data using statistical tools, such as correlation matrix and strip plot, and geostatistical analysis, such as kriging, spatial analysis is carried out using GIS tools. Further, the analysed data were used for classifying groundwater quality in different zones and compute the population dependency map.
2 Materials and Methods 2.1 Study Area The KBNIR (Fig. 1) is the first investment region in Rajasthan under the proposed Delhi – Mumbai Industrial Corridor Project (DMIC). Groundwater uses, as per development plan of the study area, comprise the following components: domestic water, agricultural water and industrial water. KBNIR falls within the semi-arid region and is frequently facing water
Tiwari K. et al.
Fig. 1 Study area location
scarcity and quality problems. Due to the semi-arid climate, it is characterized by very hot summers and very cold winters. In May and June, the maximum temperature reaches up to 48 °C. The average annual rainfall is 588 mm with fairly good amount of rainfall occuring during the southwest monsoon period. The soils of the region can broadly be divided into two classes, viz., Older Alluvial Soils and Red Gravelly Soils. Older Alluvial Soils are found in major part of the KBNIR region. Red Gravelly Soils are found in some parts of Neemrana Tehsil. According to Khandelwal et al. (2014), the population forecasted for the year 2051 would be 94,689 capita for the KBNIR area.
2.2 Collection of Data and Analysis Groundwater samples were collected from 14 different locations in the study area during the month of June 2015. The locations of groundwater sampling sites are shown in Fig. 1. All collected samples were analysed using Standard Methods (APHA 2012) to determine various hydrochemical parameters, such as Fluoride (Ion-Selective Electrode Method), Nitrate (Ultraviolet Spectrophotometric Screening Method), Chloride (Argentometric Method), Total Dissolved Solids (Total Dissolved Solids dried at 180 °C), pH (Electrometric Method), and Total Hardness (EDTA Titrimetric Method) in groundwater and results are given in Table 1. After calibrating the instruments with known standards, the analytical measurement error obtained was less than 2% and anions/cations balance error was within ±5%. After evaluation, all quality parameters of collected samples have been compared to BIS (2012). BIS has two limits, i.e., acceptable limit and permissible limit in the absence of alternate source. If any parameter exceeds the permissible limit, that water is considered unfit for human consumption.
GIS-Based Spatial Distribution of Groundwater Quality and Regional... Table 1 Analysed value of various hydrochemical parameters (mg/L) SNo
Location
1 Bawed 2 Ajaraka 3 Bhgola Jat 4 Nangal 5 Dadhiya 6 Bawadi 7 Gadli 8 Phauladpur 9 Isrisinghpur 10 Mirzapur 11 Shriyani 12 Jonaicha Khurd 13 Todarpur 14 Beejwad Chauhan Weight Relative weight
NO3¯
F¯
TDS
Cl¯
pH
TH
48.85 35.53 31.33 38.11 25.86 22.07 16.57 35.16 38.01 17.58 30.96 24.47 24.9 26.94 5 23.8
0.25 2.52 1.68 0.45 1.69 0.64 0.54 0.25 0.6 0.52 0.71 0.61 0.55 0.93 4 19.04
1610 2398 1785 1032 951 717 704 3585 918 931 1950 3551 1098 1274 4 19.04
329.89 579.82 428.69 182.44 164.95 114.96 97.47 224 194.9 189.94 546 1109.65 219.93 289 3 14.28
8.29 8.75 8.46 8.31 8.84 8.7 8.87 8.26 8.6 8.8 8.7 8.51 8.86 8.6 3 14.28
510 305 225 355 170 230 170 220 250 200 200 400 135 238 2 9.52
2.3 Groundwater Quality Parameters Figure 2 illustrates the Flowchart of the methodology that has been adopted for preparing the water quality dependency map. In the present study, GIS tools were used for the preparation of contours as well as classified maps. Geospatial distribution maps were prepared for all the parameters considered. Figure 3 shows the spatial distribution of different chemical parameters of groundwater quality based on the laboratory analysis of groundwater samples. Geospatial values of all analyzed water samples are classified into three classes based on BIS (2012). Weights are assigned to all quality parameters, according to their relative importance and overall water quality for drinking purpose (Table 2). The maximum weight of 5 has been assigned to the Collecon of Groundwater samples
Data download
Water quality analysis Satellite data (LISS-III) GIS Database generaon Area demarcaon Generaon ofinterpolated GIS maps Land use land cover classificaon Classified maps according to BIS
Populated area extracon Parameter goodness index P1c, P2c……PNc
Weighted analysis Wt1, Wt2 …… Wtn Water quality dependency map
Development of water quality suitability map
Fig. 2 Flowchart of GIS based methodology
Tiwari K. et al.
Fig. 3 Geospatial variation of groundwater quality parameters: a Nitrate; b Fluoride; c TDS; d Chloride; e pH; f Hardness
nitrate (Kumar et al. 2015) because it has major importance in water quality assessment. The hardness is given a minimum weight of 2 because it may not be as harmful as compared to other parameters. Overlay analysis of every thematic map was carried out to find the combined quality class considering all elements. Eqs. 1 and 2 are used to calculate suitable map class for every raster cell: Pi ¼ Pc * W t
ð1Þ
SMC ¼ ∑Pi
ð2Þ
where: Pi is the parameter goodness index; Pc is a parameter class; Wt is the weight assigned to each parameter; and SMC is the Suitability map class.
Table 2 Comparison of observed concentrations with standard specifications for groundwater as per BIS (2012) and WHO (2011) Quality Observed KBNIR
NO3¯ F¯ TDS Cl¯ pH TH
BIS (2012)
WHO (2011)
Min
Max
Mean
Geospatial mean Acceptable Permissible
Acceptable Permissible
16.57 0.246 704 97.47 8.26 135
48.85 2.52 3585 1109.65 8.87 510
29.74 0.85 1607.43 333.69 8.61 257.71
29.21 0.79 1660.85 340.21 8.6 267.68
50 1 500 200 7–8.5 100
45 1 500 250 6.5–8.5 200
(Where: NO3¯, F¯, Cl¯, TH, observed in mg/L and TDS in ppm)
No relaxation 1.5 2000 1000 No relaxation 600
No relaxation 1.5 1500 600 6.5–8.5 500
GIS-Based Spatial Distribution of Groundwater Quality and Regional...
2.4 Water Quality Dependency map The groundwater quality in a populated/settlement area is affected by anthropogenic activities, like seepage from un-sewered sanitation system such as soakpits and septic tanks (Jat et al. 2009). Land use/land cover classification has been carried out to find out populated/settlement area using IRS Resourcesat-1 LISS-III data at 23.5 m spatial resolution (Hu and Wang 2013; Tiwari and Khanduri 2011). Further zonal statistics were computed for each populated area using quality suitablity class map. Thus, the percentage of population with access to different quality of groundwater was determined.
3 Results and Discussion The minimum, maximum and mean values of groundwater physicochemical parameters compared to BIS (2012) and WHO (2011) are summarized in Table 2 for the study area.
3.1 Groundwater Quality Classified Maps 3.1.1 Nitrate The nitrate concentration in groundwater is generally low but can reach high levels as a result of the contamination of groundwater. Nitrate contamination is caused by the intensive use of nitrogen fertilizers, crop irrigation using domestic and industrial wastewater and by oxidation of ammonia from human or animal waste (Della et al. 2007). Some recent studies have shown that the increased concentration levels of nitrate in drinking water can cause various clinical manifestations (Kumar et al. 2002). The blue baby syndrome known as methaemoglobinaemia has been reported in new born babies due to high level of nitrate concentrations exceeding 50 mg/L in drinking water (Mudgal et al. 2009; WHO 2011). Fig. 4a shows the distribution of nitrate in groundwater in the KBNIR area. Low nitrate concentration (<45 mg/L) is shown in light blue colour which occupies approximately 99.40% (16,152 ha) of the KBNIR area, whereas 0.60% (97 ha) of the area has more than the desired level of nitrate concentration (> 45 mg/L).
Fig.4 a Nitrate Index map; b Fluoride Index map
Tiwari K. et al.
3.1.2 Fluoride Another contaminant which is commonly found in the groundwater of Rajasthan is Fluoride (Munoth et al. 2015). Fluoride is known to be beneficial to human health if taken in limited quantity (0.5 to 1.5 mg/L). When fluoride concentration in drinking water is beyond the prescribed limit (> 1.5 mg/L), then it is harmful to human health (Hussain et al. 2010). At higher levels, it causes molting of teeth, dental and skeletal fluorosis (Munoth et al. 2015). The Fluoride concentration map is represented in Fig. 4b. The areas within the acceptable limit (<1 mg/L) are shown in sky blue colour, occupying 80.26% of the study area which is suitable for the drinking purposes. The area of around 9.56%, as shown in light blue colour, has moderate concentration of Fluoride (1.0 to 1.5 mg/L). The remaining area (10.18%), mainly in the northern part of Mundawar block, has high Fluoride concentration (>1.5 mg/L) and is shown in dark blue colour.
3.1.3 Total Dissolved Solids (TDS) TDS is one of the most important characteristics of groundwater, which determines the quality of water for drinking and domestic purposes. Water containing high dissolved solids may cause health issues such as constipation. High level of TDS may be aesthetically unsatisfactory for domestic purpose. As per Water Quality Association (WQA 2017), there are no long-term adverse effects on health, due to the consumption of low TDS water, but Kozisek (2005) has indicated that low-mineral drinking water may be a risk factor for hypertension, coronary heart disease, gastric and duodenal ulcers, chronic gastritis, goitre, pregnancy complications and many other complications in newborn babies including jaundice, anemia, fractures and growth disorders. According to BIS (2012), TDS values of groundwater for drinking water should lie between acceptable to permissible limit (500 to 2000 mg/L). Light yellow colour in Fig. 5a represents TDS values between 500 to 2000 mg/L in 75.38% (12,248 ha) of the area and red colour represents TDS values above the maximum permissible limit, i.e., >2000 mg/L in 24.62% (4001 ha) of the area which is unsuitable for drinking purpose.
Fig. 5 a TDS index map; b Chloride index map
GIS-Based Spatial Distribution of Groundwater Quality and Regional...
3.1.4 Chloride Chloride is a negatively charged ion that is commonly present in groundwater. The presence of chloride in drinking water is classified into three categories namely good quality (< 250 mg/L), permissible limits (250 to 1000 mg/L) and unsuitable (>1000 mg/L). High chloride concentration (>1000 mg/L) in groundwater is unsuitable for drinking and other purposes. Around 0.80% (129.25 ha) of the total geographical area of the KBNIR has high chloride concentration (Table 3; Fig. 5b). Chloride concentration within the permissible limit (250 to 1000 mg/L) occupies 58.51% (9507 ha) of the study area. The groundwater with chloride content less than 250 mg/L occupies 40.70% (6613 ha) of the study area.
3.1.5 pH The pH value of drinking water is an important parameter that determines whether the water is acidic or alkaline in nature. A number of minerals and organic matter interact with each other resulting in increase or decrease of pH values. In the present study, groundwater with pH values within permissible limits (6.5 to 8.5) occupies 12.80% (2079.5 ha) of the study area and is shown by light green colour in Fig. 6a. The area with pH values above 8.5 is shown in green colour (Fig. 6a), which occupies nearly 87.20% (14,169.5 ha) of the study area.
3.1.6 Total Hardness The Total Hardness (TH) is another important groundwater quality parameter and its permissible limits depend upon its uses, such as domestic, agricultural or industrial. Depending on the environmental conditions, hardness of water may have varing impact on different communitites. Hardness in water is caused by a high concentration (Kumar et al. 2015) of certain salts (e.g., calcium, magnesium, carbonates chlorides and nitrates). Soft water may have an adverse effect on mineral balance in humans. Hard water does not have adverse effect (CWDR 2013) on human health but Sengupta (2013) has discussed about an inverse relationship between the hardness of drinking water and cardiovascular disease. Kozisek (2005) reported that concentrations beyond permissible limit of the hardness showed a higher risk of gallstones, urinary stones, kidney stones and arthropathies in population. The TH concentration map of the studied area is presented in Fig. 6b. The areas with moderate concentration (200 to 600 mg/L) occupy almost 95.08% of the study area which is shown in magenta colour. The remaining small part of the KBNIR with low concentration (<200 mg/L), which is approximately 4.92% of the area, is shown in light pink colour. In the literature, many studies have reported variation in groundwater pollutant levels. Jat et al. (2009) evaluated fluoride concentration in groundwater in Ajmer district of Rajasthan state before and after urban growth, and found that it has doubled from 0.96 to 1.90 mg/L between 1992 and 2003. Singh et al. (2013) observed that TDS, F¯, NO3¯ and TH are higher Table 3 Percentage distribution of water quality class based on of different water quality parameters Class
NO3¯
F¯
TDS
Cl¯
pH
TH
Good Moderate Bad
99.40 0.60 0.00
80.26 9.56 10.18
0.00 75.38 24.62
40.70 58.51 0.80
12.80 0.00 87.20
4.92 95.08 0.00
Tiwari K. et al.
Fig. 6 a pH index map; b Hardness index map
than the desirable limits in many water sample locations in Bundelkhand region, e.g., TDS varied from 152 to 2667 mg/L and F¯ varied from 0.10 to 3.43 mg/L. Kumar et al. (2002) reported higher concentration of NO3¯ in groundwater in most parts of Rajasthan. Similarly, Hussain et al. (2010) has reported on the effect of high fluoride concentration on the local population. Table 3 shows the spatialy distributed area of groundwater quality parameters computed by water quality class maps.
3.2 Statistical Distribution of Water Quality Parameters 3.2.1 Strip Plot The hydrochemistry of the groundwater samples is shown in the strip plot (Fig. 7). Strip plots can be used to compare groundwater quality data of the same parameter for different collected samples. The plots are constructed using the median value as a red bar, the mean value as a plus sign, the range of 50% of the samples in circle with center at the mean value. Figure 7 shows that the differences between mean and median in nitrate, pH and TH are low and comparatively higher in fluoride, TDS and chloride. Further, it can be seen that, for all parameters, the mean value is higher than the median value except for pH. Another finding is that the range of central 50% value of fluoride, chloride and TDS is narrow in strip plots, and on the other side, nitrate, pH and TDS are showing a high range. TDS ranged from 704 to 3585 mg/L, with an average of 1607.43 mg/L. Importantly, fluoride and TH plots indicate that the mean value is outside the center circle, indicating higher occurrence of low values. Also, in case of TDS, although the mean value is within the circle, it is inclined towards lower values. Such indications are not easily available in other statistical methods, showing the importance of strip plot charts.
3.2.2 Correlation Matrix Analysis In this study, the multivariate statistical method, Pearson’s correlation coefficient (r), was used to identify the relationship between observed groundwater quality parameters for drinking purpose. Pearson correlation coefficient value ranges from −1 to 1. If the r value is near zero, it indicates that there is no relationship between the two variables (Mudgal et al. 2009). Interparameter correlation matrix of each water quality parameter of the groundwater samples are
GIS-Based Spatial Distribution of Groundwater Quality and Regional...
Strip plot (F-)
Strip plot (TDS )
50
3
4000
45
2.5
3500
40
2 1.5
3000
TDS
35
F-
NO3-
Strip plot (NO3-)
30
2500 2000
1 1500
25 0.5
20
1000
0
15
500
Strip plot (pH)
Strip plot (Cl-)
Strip plot (TH)
1200
8.9
550
1000
8.8
500 450
8.7
800
400
TH
600
pH
Cl-
8.6 8.5 400 8.4
350 300 250 200
200
8.3
0
8.2
150 100
Fig. 7 Strip plot distribution of groundwater parameter
given in Fig. 8 where red color shows a positive corelation and blue color shows a negative correlation. The pH value shows significant negative correlation with NO3¯ (r = −0.656) and TH (r = −0.621) with 95% confidence level during the study period 2015. TDS had a strong positive correlation with Cl¯ (r = 0.725) and weak negative correlation with pH (r = −0.394). There is a weak positive correlation between Cl¯ and TH (r = 0.450). No significant correlations were obtained in some of groundwater quality parameters in KBNIR. Similar type CM analysis by other authors (e.g., Boateng et al. 2016; Mudgal et al. 2009; Tiwari et al. 2015b) have indicated that Cl¯ and TDS are significantly positively correlated (r = 0.57). A strong positive correlation (r = 0.71) was also observed between F¯ and pH. The concentrations of TDS, Cl¯ and pH are similar in populated as well as non-populated area, therefore, the source of contamination may be geogenic.
3.2.3 Principal Component Analysis Principal component analysis has been reported as a powerful technique for pattern recognition that attempts to clarify the variation of the intercorrelated variables of a dataset (Rajput and Goyal 2017; Ravikumar and Somashekar 2017). In the present study, PCA has been carried out on the groundwater quality dataset using varimax rotation in xlstat-2017. Criteria selected
Tiwari K. et al.
Fig. 8 Correlation matrix of the groundwater quality parameters during June 2015 at significance level alpha ≥ 0.05
for the minimum acceptable limit of loadings for factor extraction was 1 according to Dehghanzadeh et al. (2015). PCA results, including the eigenvalues, variability percentage and cumulative percentage by each principal component (PCs) of the water quality parameters are shown in Table 4, which includes the factor score of each PC. It is seen that the first 5 PCs explain 98.5% of the total variance. Figure 9 shows a biplot of the water quality parameters and water quality observation locations within the study area. Figure 9 also presents variation of each groundwater quality parameter with respect to PC factors 1 (47.68%) and 2 (24.93%). The first PC explained 47.67% of the total variance and was best represented by pH with a strong positive loading, Cl¯ with moderate negative loading and NO3¯, TDS and TH with strong negative loadings. PC2 was dominated by Cl¯ and F¯ with strong positive loadings, moderate positive
GIS-Based Spatial Distribution of Groundwater Quality and Regional... Table 4 Varimax rotated principal component analysis and factor score for groundwater samples Parameters
F1
F2
F3
F4
F5
NO3¯ F¯ TDS Cl¯ pH TH Eigenvalue Variability (%) Cumulative % Observation location Bawed Ajaraka Bhgola Jat Nangal Dadhiya Bawadi Gadli Phauladpur Isrisinghpur Mirzapur Shriyani Jonaicha Khurd Todarpur Beejwad Chauhan
−0.754 0.133 −0.699 −0.574 0.855 −0.851 2.861 47.677 47.677
−0.341 0.658 0.511 0.729 0.367 −0.136 1.496 24.926 72.603
0.478 0.722 −0.308 −0.210 0.038 0.121 0.905 15.076 87.679
−0.113 −0.097 −0.364 0.252 0.246 0.462 0.492 8.200 95.879
0.269 −0.134 0.034 0.084 0.235 −0.107 0.165 2.753 98.632
−3.136 −0.668 −0.384 −1.262 1.825 1.414 2.437 −1.755 0.038 1.837 0.000 −2.606 1.823 0.436
−1.895 2.306 0.772 −1.769 0.591 −0.674 −0.312 −0.524 −1.122 −0.183 0.674 2.316 −0.092 −0.088
0.662 1.955 0.877 0.402 1.084 −0.231 −0.700 −1.118 0.527 −0.797 −0.457 −1.773 −0.486 0.055
0.906 −0.025 −0.653 0.182 −0.177 0.351 0.347 −2.068 0.022 0.416 −0.122 0.924 −0.112 0.010
0.219 0.107 −0.550 −0.414 −0.056 −0.413 −0.167 −0.056 0.589 −0.259 0.773 −0.146 0.671 −0.299
loading of TDS and moderate negative loading for NO3¯, and accounted for 24.92% of the total variance. PC3 explained 15.07% of the total variance with 0.905 of the total loading.
3.3 Classification of Drinking Water and Suitability Evaluation Presently, valuable groundwater resources have come under enormous risk due to the drastic increases in population, modern agricultural practices (use of pesticide and fertilizer) and industrialization due to waste disposal practices (Babiker et al. 2007; Jat et al. 2009). Contaminated groundwater is directly affecting the human health. So, groundwater quality mapping provides a way to understand overall water quality conditions. Water Quality Class (WQC) maps developed for the groundwater quality parameters indicate that there is a large variation in quality from station to station. Figure 10a presents populated/settlement area which has been used to identify the percentage of population exposed to groundwater with different suitability class. Figure 10b demarcates the study area into different groundwater quality/ suitability classes for drinking purpose. From the results of the spatial distribution of poor, moderately poor, good and excellent water quality maps, the percentages of the area for each quality class are summarized in Table 5. The spatial distributed map of WQC indicates that poor class is observed in the eastern side of the study area, while moderately poor class is found in the western side of study area (Fig. 10b). Poor quality water in these regions could be due to over exploitation of groundwater and agriculture practices. The excellent water quality area is found in the central part of the KBNIR. The spatial maps of WQC present a complex distribution pattern. Figure 10b shows the final water quality map that was produced by overlaying of the thematic maps of different parameters. The spatially distributed groundwater quality map of KBNIR was obtained using a parameter
Tiwari K. et al.
Fig. 9 Biplot of the water quality parameters (represented by red points) and water quality observation points (represented by blue points) in the study area according to PC factor 1 (47.68%) and factor 2 (24.93%)
goodness index analysis. The area for excellent water quality was found to be 43.22% of the total geographic area (TGA) at the south east and centre part of the KBNIR. The areas with good water quality cover 22.66% of the total study area. The rest of the study area lies in poor quality and moderately poor quality covering about 10.46% and 23.66% of the total area, respectively. Figure 11 shows a population depedency map of the study area. Based on spatial distribution of populated/settlement area, only 36.02% population area has access to excellent water quality, covering villages Isrisinghpur, Mirzapur, Bawedi, Behrorjat Molawas, Nagal, Todarpur, Palawa,
Fig. 10 a Populated area map; b Quality class area
GIS-Based Spatial Distribution of Groundwater Quality and Regional... Table 5 Percentage area falling under different water quality class for populated and total study area Quality
Populated area %
Total area %
Poor Moderately poor Good Excellent
6.26 34.16 23.56 36.02
10.46 23.66 22.66 43.22
Gadli, Manethi and Phauladpur etc. Out of the total populated area, 23.56% and 34.16% are dependent on good and moderately poor water quality, respectively. Only 6.26% of the total populated area lies in poor category and so the population of Ahir Bhagora, Ajaraka, Darwarpur and Sundarwari etc. villages are dependent on poor quality water due to high concentration of TDS, fluoride and pH. This indicates that source of contamination may be geogenic. Some other studies have also indicated that most of the part of Rajasthan is hydrochemically unsuitable for drinking purpose (Bhaduri et al. 2009; Jagtap et al. 2012; Kumar et al. 2002; Mitharwal et al. 2009). Singh et al. (2013) also reported poor water quality in some sample sites of Bundelkhand region. Tiwari et al. (2014) observed similar results for WQC by computation of WQI and found the following: 21% of the area falling in excellent, 58% in good and 21% in poor categories. In order to compare the results of PCA and population groundwater quality dependency map derived using the goodness index, zonal statistics tools were used to compute mean values
Fig. 11 Map of populated/settlement area with dependency on type of groundwater quality
Tiwari K. et al. Table 6 Mean values of PCs in various groundwater quality categories Quality Category
PC-F1
PC-F2
PC-F3
PC-F4
Poor Moderately Poor Good Excellent
−0.28 −0.97 −0.38 −0.02
1.48 0.66 −0.19 −0.65
1.24 −0.79 −0.31 −0.16
−0.13 0.15 −0.08 0.06
of PCs 1 to 4 (Table 6) for different grounwater quality zones. As can be seen from Table 6, in the area with excellent quality groundwater, all the PC factors are relatively low, indicating low values of all parameters. Similarly, in very poor groundwater quality areas, PC factors 2 and 3 are highly positive, indicating high F¯, TDS and Cl¯. In poor groundwater quality areas, PC factor 3 is highly negative and PC factor 2 is positive; therefore, it can be concluded that both analyses indicate similarities.
4 Conclusions This paper presents integrated approaches to characterizing the spatial distributed hydrochemistry and hydro-chemical suitability of groundwater for drinking properties in the KBNIR, Rajasthan. On the basis of analytical findings, the following conclusions can be drawn. In 55.88% of the total geographical area of KBNIR, the groundwater is suitable for drinking purposes, while in the remaining 10.46% of the area of the KBNIR, it is unsuitable for drinking purposes, and in 23.66% of the area, it is partially suitable for drinking. The results of spatial analysis indicate that 75.38% of the area is moderately affected by TDS and 10.18% of the area is severally affected by F¯. Also, the groundwater quality goodness maps were combined with spatial distribution of the populated area to determine the percentages of population affected by different classes of groundwater quality, thus making it easier to identify the area facing bad quality water. It can be concluded that all the parameters are more or less correlated with other parameters. Wide statistical variation exists in the TDS and Cl¯ values of the groundwater but NO3¯ is statistically distributed within permissible limit in KBNIR. Two different statistical tools, namely strip plot and PCA were used for statistical analysis of groundwater quality in the study area. The strip plot provides very useful indication of low mean value compared with median value and on the spread of the values, which are helpful in overall water quality mapping. Poor quality groundwater should not be used for drinking as it may lead to increased human health related problems. The waterborne diseases may occur due to the poor quality of water in infants, young children and the elderly people, who live especially in poor water quality areas. Hence, regular monitoring is important for keeping good water quality and a healthy environment. Acknowledgements The authors would like to acknowledge the PHE lab, Department of Civil Engineering, MNIT Jaipur for providing technical support during groundwater quality data analysis. Compliance with Ethical Standards Conflict of Interest This is to certify that the data, figures, graphs, tables used in the paper are based on doctoral study of the first author and there is no conflict of interest in publication of this paper.
GIS-Based Spatial Distribution of Groundwater Quality and Regional...
References Amiri BJ, Nakane K (2009) Modeling the linkage between river water quality and landscape metrics in the Chugoku district of Japan. Water Resour Manag 23(5):931–956 APHA (2012) Standard Methods for the Examination of Water and Wastewater, 22nd Edition, American Public Health Association (APHA) American Water Works Association (AWWA) & Water Environment Federation (WEF) Babiker IS, Mohamed MAA, Hiyama T (2007) Assessing groundwater quality using GIS. Water Resour Manag 21(4):699–715 Banoeng-Yakubo B, Yidana SM, Emmanuel N, Akabzaa T, Asiedu D (2009) Analysis of groundwater quality using water quality index and conventional graphical methods: the Volta region, Ghana. Environ Earth Sci 59(4):867–879 Bhaduri M, Goyal R, Gupta AB (2009) Groundwater quality in Sanganer area of Rajasthan. ISH Journal of Hydraulic Engineering 15(3):65–75 BIS (2012) Indian standard drinking water specifications BIS-10500. Bureau of Indian Standards, New Delhi Boateng TK, Opoku F, Acquaah SO, Akoto O (2016) Groundwater quality assessment using statistical approach and water quality index in Ejisu-Juaben Municipality, Ghana. Environ Earth Sci 75(6): 489–502 Bodrud-Doza M, Islam ARMT, Ahmed F, Das S, Saha N, Rahman MS (2016) Characterization of groundwater quality using water evaluation indices, multivariate statistics and geostatistics in central Bangladesh. Water Sci. doi:10.1016/j.wsj.2016.05.001 CGWB (2010) Ground water quality in shallow aquifers of India. Central Ground Water Board, Faridabad CWDR (2013) Cambridge Water Department Report −-2013 www.cambridgema.gov Dehghanzadeh R, Safavy Hir N, Shamsy SJ, Taghipour H (2015) Integrated assessment of spatial and temporal variations of groundwater quality in the eastern area of Urmia salt Lake basin using multivariate statistical analysis. Water Resour Manag 29(4):1351–1364 Della RC, Belgiorno V, Meriç S (2007) Overview of in-situ applicable nitrate removal processes. Desalination 204:46–62 Ding J, Jiang Y, Liua Q, Hou Z, Liao J, Fud L, Peng Q (2016) Influences of the land use pattern on water quality in low-order streams of the Dongjiang River basin. China: A multi-scale analysis, Science of the Total Environment 551–552:205–216 Gorai AK, Hasni SA, Iqbal J (2014) Prediction of ground water quality index to assess suitability for drinking purposes using fuzzy rule-based approach. Appl Water Sci. doi:10.1007/s13201-0140241-3 Goyal SK, Chaudhary BS, Singh O, Sethi GK, Thakur PK (2010) GIS based spatial distribution mapping and suitability evaluation of groundwater quality for domestic and agricultural purpose in Kaithal district, Haryana state, India. Environ Earth Sci 61(8):1587–1597 Gulgundi MS, Shetty A (2016) Identification and apportionment of pollution sources to groundwater quality. Environ Proc 3(2):451-461. doi: 10.1007/s40710-016-0160-4 Hu S, Wang L (2013) Automated urban land-use classification with remote sensing. Int J Remote Sens 34(3): 790–803 Hussain J, Hussain I, Sharma KC (2010) Fluoride and health hazards: community perception in a fluorotic area of Central Rajasthan (India), an arid environment. Environ Monit Assess 162:1–14 Jagtap S, Yenkie MK, Labhsetwar N, Rayalu S (2012) Fluoride in drinking water and defluoridation of water. Chem Rev 112:2454–2466 Jat MK, Khare D, Garg PK (2009) Urbanization and its impact on groundwater: a remote sensing and GIS-based assessment approach. Environmentalist 29:17–32 Khandelwal P, Tiwari K, Goyal R (2014) Rooftop rain water harvesting as part of IWRM plan of Khuskera-Bhiwari Neemrana Investment Region, ETWQQM -2014 Conf. Proceedings 145–149 Kozisek F (2005) Health risks from drinking demineralised water. World Health Organization, Geneva, pp 148– 163 http://www.who.int/water_sanitation_health/dwq/nutrientschap12.pdf Kumar S, Gupta AB, Gupta S (2002) Need for revision of nitrates standards for drinking water: a case study of Rajasthan. Indian J Environ Health 44(2):168–172 Kumar SK, Logeshkumaran A, Magesh NS, Godson PS, Chandrasekar N (2015) Hydro-geochemistry and application of water quality index (WQI) for groundwater quality assessment, Anna Nagar, part of Chennai City, Tamil Nadu, India. Appl Water Sci 5:335–343 Machiwal D, Jha MK (2015) Identifying sources of groundwater contamination in a hard-rock aquifer system using multivariate statistical analyses and GIS-based geostatistical modeling techniques. Journal of Hydrology: Regional Studies 4:80–110
Tiwari K. et al. Mitharwal S, Yadav RD, Angasaria RC (2009) Water quality analysis in Pilani of Jhunjhunu District (Rajasthan) –The place of Birla’s origin. Rasayan Journal of Chemistry 2(4):920–923 Mudgal KD, Kumari M, Sharma DK (2009) Hydrochemical analysis of drinking water quality of Alwar district, Rajasthan. Nat Sci 7(2):30–39 Munoth P, Tiwari K, Goyal R (2015) Fluoride and nitrate groundwater contamination of Rajasthan, India: a review. 20th International Conference on Hydraulics. Water Resources and River Engineering. IIT Roorkee. 17-19 December 2015 Nag SK, Ghosh P (2013) Variation in groundwater levels and water quality in Chhatna block, Bankura District, West Bengal – A GIS approach. J Geol Soc India 81:261–280 Nas B, Berktay A (2010) Groundwater quality mapping in urban groundwater using GIS. Environ Monit Assess 160:215–227 Rajput H, Goyal R (2017) Groundwater quality and contamination source assessment of Jaipur District using multivariate statistical techniques. International Conference on Agriculture, Environmental and Bio Sciences (ICAEBS), 27-29 April 2017, Chandigarh Ravikumar P, Somashekar RK (2017) Principal component analysis and hydrochemical facies characterization to evaluate groundwater quality in Varahi river basin, Karnataka state, India. Appl Water Sci 7(2):745–755 Sengupta P (2013) Potential health impacts of hard water. Int J Prev Med 4(8):866–875 Shanmugasundharam A, Kalpana G, Mahapatra SR, Sudharson ER, Jayaprakash M (2015) Assessment of groundwater quality in Krishnagiri and Vellore districts in Tamil Nadu, India. Appl Water Sci. doi:10.1007/s13201-015-0361-4 Singh AK, Raj B, Tiwari AK, Mahato MK (2013) Evaluation of hydrogeochemical processes and groundwater quality in the Jhansi district of Bundelkhand region, India. Environ Earth Sci 70(3):1225–1247 Singhal V, Goyal R (2014) Groundwater model to predict the impact due to textile units at Pali. Arab J Geosci 7(12):5185–5192 Tiwari K, Khanduri K (2011) Land use / land cover change detection in Doon valley (Dehradun tehsil), Uttarakhand: using GIS & Remote Sensing Technique. International Journal of Geomatics and Geosciences 2(1):34–41 Tiwari AK, Singh PK, Mahato MK (2014) GIS-based evaluation of water quality index of groundwater resources in west Bokaro coalfield, India. Current World Environment 9(3):843–850 Tiwari K, Goyal R, Sarkar A, Munoth P (2015a) Integrated water resources management with special reference to water security in Rajasthan, India. Discovery 41(188):93–101 Tiwari AK, Singh AK, Singh AK, Singh MP (2015b) Hydrogeochemical analysis and evaluation of surface water quality of Pratapgarh district, Uttar Pradesh. India Appl Water Sci doi. doi:10.1007/s13201-015-0313-z WHO (2011) Guidelines for Drinking-water Quality, 4th ed. ISBN 978 924 1548151 WQA (2017) Water Quality Association, Consumption of Low TDS Water https://www.wqa. org/Portals/0/Technical/Technical%20Fact%20Sheets/2015_TDS.PDF. Accessed 23 June 2017