Application of GIS technique to address the uranium contamination in groundwater of a hard rock aquifer, South India

Abstract The present research aims to determine the sources and concentration of uranium in the groundwater of the hard rock aquifer. Two hundred and fifty-four groundwater samples were collected, and the analysis shows the uranium concentrations range from 0.5 to 113 ppb (pre-monsoon) and 0.5 to 120 ppb (post-monsoon). Factor analysis illustrates that the lower concentrations of uranium in the study region may be due to the poor dissolution of minerals in the subsurface. Saturation Index values confirm that the Uraninite mineral is undersaturated, and the U4O9 has dominant distribution during both seasons. During both seasons, four samples have a uranium concentration of more than 60 ppb, which might be due to the fertilizer usage in the agricultural activities in the study area. The GIS technique highlights the uranium-contaminated location and polluted water quality parameters using the Inverse Distance Weighted spatial interpolation plots.


Introduction
Uranium is a naturally occurring radioactive element present in a minor quantity globally. It causes serious health effects to human beings when it exceeds the permissible drinking water limit. Many factors, such as groundwater depletion, uranium mining, usage of phosphate fertilizers, etc., are responsible for the elevated uranium levels in groundwater aquifers (Sharma et al. 2018). Uranium concentrations in groundwater ranged from 0.2 to 453 lg/L with a mean concentration of 51.4 lg/L due to the concealed sandstone-type uranium deposits in northeast china (Zhang et al. 2020). In India, groundwater is the primary source of drinking water in rural areas. Previous studies reported that groundwater in shallow aquifers in India has a uranium concentration ranging from 0.0 to 2876 lg/L (14,377 samples) (Central Ground Water Board (CGWB) 2020). The baseline study of the uranium concentration in India shows that a higher uranium concentration is present in India's groundwater aquifers (Nepolian et al. 2022). Especially in Tamilnadu, 1208 samples were analysed, and the maximum concentration of 302 lg/L in Krishnagiri District (Central Ground Water Board (CGWB) 2020). The observed uranium concentration well exceeds the permissible limit (60 lg/L) recommended by Atomic Energy Regulatory Board (AERB), India (AERB 2007). The higher concentration of uranium present in the groundwater of India leads to significant health effects such as kidney stones, congenital disabilities, cancer, etc. . Hence the detailed measurements and sources identification of uranium in the groundwater aquifers of India is essential for drinking water quality management. In recent years many researchers in India have focused on this field and reported many studies about the uranium sources and elevated levels of uranium in groundwater aquifers (Sar et al. 2018;Sharma et al. 2018;2020;Pandey et al. 2021;Saikia et al. 2021).
In Tamilnadu, many researchers have reported the sources and higher uranium concentrations in groundwater aquifers in recent years (Adithya et al. 2019;Ganesh et al. 2020;). The previous study shows that central Tamilnadu has a uranium concentration ranging from 0.79 to 71.93 ppb due to the lithological contact of the groundwater in the subsurface (Adithya et al. 2019;Devaraj et al. 2021). In the Tiruvannamalai mountain region, Tamilnadu has reported the uranium concentration to vary from 0.2 to 25.8 ppb, which may lead to cancer risk due to the ingestion of this uranium-contaminated groundwater (Ganesh et al. 2020). In higher background radiation places of Tamilnadu (Kanniyakumari district), the groundwater uranium concentration ranges from 0.1 to 10.5 ppb and the lower uranium concentration due to the reducing nature of the groundwater . The groundwater in the Salem district of Tamilnadu has uranium levels ranging from 0.01 to 385.4 ppb inferred the interaction of groundwater with the metasediments, intrusive granite, syenite, and carbonatite minerals (Muthamilselvan 2020). In the Cuddalore district of Tamilnadu, the uranium ranges from 0.1 to 24.67 ppb occurs by the leaching of minerals and anthropogenic impacts (Anandhan et al. 2020). The literature review reflects the lack of study about the sources and uranium concentration in the groundwater of Virudhunagar district, a hard rock aquifer region. Hence this region has been selected for the present study to find the sources and the states of saturation.
The study area has reported many groundwater quality issues in recent years. The Industrial Township of the study region has heavy metal pollution due to the industrial effluents and municipal dumpsites ). The study area has also reported health risks due to the elevated fluoride and nitrate concentration present in the groundwater by the impacts of agricultural and industrial effluents . Hotspots of the unsuitable quality of groundwater for drinking and irrigation purposes were identified (Lakshmi et al. 2021b). Pandaldudi region has been determined as the region with the most unsuitable groundwater quality for drinking and irrigation (Udayanapillai and Kaliammal 2016). Previous studies of this district demonstrate the lack of analysis of the uranium and its sources in groundwater. Moreover, in this district, people depend on groundwater for drinking during summer. Hence the present study focuses on determining the process and the sources of the higher uranium concentrations.

Sampling area
The study area falls in Virudhunagar district, located in southern Tamilnadu state, covering an area of 4288 km 2 (Latitude 9 11 0 15 00 -9 46 0 23 00 and Longitude -77 19 0 32 00 -78 23 0 58 00 ) ( Figure 1). This district is covered by Madurai and Sivagangai district (on the north), Tirunelveli and Tuticorin district (on the south), Kerala state (on the west), Ramanathapuram district (on the east), and Theni district (on the northwest). The total population of this district is 1751301 as per the census 2001 (Central Ground Water Board (CGWB) 2008). The lithological map of this district shows that the fissile hornblende biotite gneiss and charnockite are major rocks distributed in this district ( Figure  2). The deep red loam, black soil, and red sandy soil are major soils spread over this district. The main occupation is irrigation activities, and 1155 sq. km of the area has been irrigated, with paddy as the major crop. This district's major surface water sources are Vaippar, Gundar, and Arjuna nadhi. The average annual rainfall is 800 mm, and the northeast monsoon (October to December) contributes higher rainfall in this district (Central Ground Water Board (CGWB) 2008). The groundwater serves as the region's major source for domestic needs . The important aquifer systems in the district are unconsolidated and semi-consolidated formations and weathered, fissured, and fractured crystalline rocks. The depth of the water level in the district varied between 0.67 and 12.12 m bgl during pre-monsoon and varied from 0.49 to 8.78 m bgl during post-monsoon (Central Ground Water Board (CGWB) 2008). The seasonal fluctuation shows a rise in water level, which ranges from 0.35 to 2.8 m. Many small-scale and large-scale industrial activities also depend on groundwater reserves.

Sampling and analysis
Sample collections were performed during the pre-monsoon (127 samples, September 2016) and post-monsoon (127 samples, March 2018) in the pre-selected locations of 6 Â 6 grid in the study area ( Figure 1). One litre HDPE (High-Density Polyethylene) bottles (pre-rinsed with concentrated nitric acid) were used for the sample collection. The samples were collected from hand pumps and borewells with depths ranges from 120 to 200 m bgl. In situ water quality parameters, potential Hydrogen (pH), Electrical Conductivity (EC), Total Dissolved Solids (TDS), and Oxidation Reduction Potential (ORP) were noted in the field using a handheld water analysis kit (Portable digital water and soil analysis kit -EI 161). The samples were labelled, tightly packed, stored in an icebox (4 C), and safely transmitted into the laboratory. The laboratory water quality parameters were analyzed using high-grade chemicals, calibrated instruments, and standard procedures (APHA 1995). The titrimetric method was used to determine the concentration of calcium, magnesium, chloride, carbonate, and bicarbonate ). The calorimetric method was employed to estimate the concentration of nitrate, phosphate, and sulphate in the collected samples (Elico, SL-177) (Raja and Neelakantan 2021a). The flame photometer was used to determine the concentration of sodium and potassium present in the collected samples (EI-381E) (Lakshmi et al. 2021b). The fluoride concentration in the collected samples was measured by an ion-selective electrode (HI-4110) (Raja and Neelakantan 2021b). The instrument calibration and measurement of blank samples were conducted during the sample analyses to achieve the accuracy and precision of the water quality analysis. The accuracy and precision of the analysis were checked by the ion-balance check (IBC) by following Eq. (1) Where, T C is total cations and T A is total anions. The calculated ion balance check was observed within the acceptable limit of ±10 (Domenico andSchwartz 1990 Matthess 1982).

Uranium estimation
The uranium concentration in the collected samples was estimated using the LED Fluorimeter LF-2a. This instrument is specially made for measuring uranium in water samples and has a detection limit of 0.2 ppb (Rathore et al. 2001). The phosphate buffer (pH-7, 0.5 ml) is added to the collected water sample (5 ml), and the fluorescence of the formed uranium-phosphate complex is measured at the wavelength of 405 nm . The standard addition method is employed for the uranium analysis to avoid the matrix effect (Ganesh et al. 2020). The instrument was calibrated with a standard uranium solution and blank sample ).

Hydrochemistry and equilibrium modelling
Geochemical modelling of the analyzed data was computed using PHREEQC, Version 2.8 (Parkhust and Appelo 1999). The minerals' saturation indices (SI) for the input analytical data are extracted using the PHREEQC software. Saturation Indices represent the particular mineral that has under-saturated (SI < 0), saturated (SI ¼ 0), and supersaturated (SI > 0) in the groundwater samples . Saturation indices are computed by the following Eq. (2) in the PHREEQC program using ion activity product (IAP) and stability constant (k) (Parkhust and Appelo 1999).
The MDUSA program was used to compute the chemical equilibrium diagrams (uranium speciation and pH vs. Eh plot) for the analytical data (Puigdomenech 1999). These plots were used to visualize the predominant mineral distribution in the collected groundwater samples (Khawassek et al. 2018).

Water quality index and factor analysis
The suitability of water for drinking is assessed based on the water quality index. The water quality index of the analyzed is calculated by following Eq. (3) using the relative weight of the parameters (W i ) (Table S1) and quality rating of the parameter (Eq. (4)) (Brown et al. 1970;. Where C i is the observed concentration of the parameters and S i is the standard permissible limit prescribed by Bureau of Indian Standards (BIS) (2012) ( Table S1).
The anthropogenic and natural polluting sources are evaluated using factor analysis (Ouyang 2005). The PCA (Principal Component Analysis) of the present analytical data was performed using the Varimax rotation method in the SPSS software package v.23.
The current analytical data extracted the factor component values and rotated space diagram from the factor analysis. The factors with Eigen values of more than one is chosen for this analysis ).

GIS technique
The water quality parameters that exceed BIS (2012) permissible limit and uranium-contaminated locations were determined using the spatial distribution plots drawn by GIS software (Arc Map, v.10.3). The spatial distribution of values was achieved by using the inverse distance weighted (IDW) interpolation method. IDW is a method that assumes the point of unknown data from the nearest observed data points (Ke et al. 2011). The spatial variation of water quality depends on the closest observation wells. Therefore, the inverse distance weighted average is used to interpret the present analytical data. The method is proved to be efficient and best for the spatial distribution of water quality data (Sapna et al. 2018;.

Geochemical and equilibrium modelling
The minimum, maximum, and mean values for the analyzed water quality parameters for the collected samples in the Virudhunagar district are tabulated (Table 1). The spatial variation of water quality parameters between the pre-and post-monsoon are given in spatial distribution plots (Figures S1-S8). The variation of uranium in pre-monsoon and post-monsoon is caused by the monsoon failure, leaching of minerals, and increased anthropogenic activities. The collected samples in both seasons are slightly alkaline depending on mean pH values (pre-monsoon À7.71) and reducing nature based on the mean ORP values (À41 mVpre-monsoon, The above statement is supported by the hydrochemistry study carried out in unconfined aquifers in the Indo-Gangetic plain, Punjab. The study shows a higher concentration of calcium and magnesium in the groundwater samples due to the mineral dissolution of calcite and dolomite (Gautam et al. 2022).
The results of the PHREEQC for uranium represent the under saturation condition of uraninite mineral during both seasons [mean SI values À26.2 (pre-monsoon), À25.8 (post-monsoon)] ( Table 2). The equilibrium modelling of uranium illustrates that U 4 O 9 is the dominant species in both seasons (Figures 3 and 4). The uranium species fraction diagram is plotted between the pH from 1 to 14 to understand the variation of the species depending on the pH. Under the reducing and slight alkaline nature, U 4 O 9 is reported to be the dominant species of uranium (Kumar et al. 2011;Devaraj et al. 2021). Hence, the undersaturation of uraninite and U 4 O 9 leads to the lower concentration of uranium [mean À15 (pre-monsoon), 16 (post-monsoon)] in the collected water samples during both seasons. The reducing nature of the groundwater leads to this uranium dissolution and forms the insoluble U 4þ leading to precipitation. The uraninite and U 4 O 9 dissolution reaction can be stated as follows:

Hydrogeochemistry of groundwater
Hydrogeochemistry of the collected groundwater samples was determined using the Piper, Durov, and Giggenbach triangle (Piper 1944;Durov 1948;Giggenbach 1988). The Piper plot of the present study reveals that the collected water samples have the dominant water types are mixed Ca 2þ -Mg 2þ -Cl À and Ca 2þ -Cl À ( Figure S9). Samples that have higher uranium also show the water types are calcium chloride, magnesium chloride, and mixed Ca 2þ -Mg 2þ -Cl À type ( Figure S10). Durov plot of the present investigation illustrates that the water samples have the reverse ion exchange and the mixing of ions is predominant during both seasons ( Figure S11). The samples with higher uranium also show the mixing and reverse ion exchange conditions in both seasons ( Figure S12). Giggenbach triangle of the present analytical data shows that the collected water samples are in a disequilibrium state during both seasons ( Figure S13). The reverse ion exchange condition in the collected samples during both seasons is responsible for the Na þ ions having reverse ion exchange with the Ca 2þ and Mg 2þ ions. Hydrogeochemical studies in an arid environment in northwestern Saudi Arabia confirm that the sodium ions in groundwater have replaced the calcium and magnesium ions (Zaidi et al. 2015). This type of reverse reaction occurs in the groundwater, leading to water being in disequilibrium during both seasons ( Figure S13).

Factor analysis and water quality index
The analytical data of the collected samples extracts the five factors in the factor analysis for both seasons (Table 3 and Figure 5). Factor 1 has the positive factor loading for TDS, Na þ , K þ , Cl À , and U represents the geogenic factor (i.e. mineral dissolution). The secondary salt dissolute from the rocks is responsible for this factor 1 (Nepolian et al. 2022). Factor 2 has positive nitrate, calcium, and magnesium loadings, indicating the anthropogenic factor. The dissolution of calcium and magnesium minerals has been influenced by the anthropogenic impacts (nitrate fertilizers used in agriculture, industrial effluents, and sewage disposal) of the study area. This type of anthropogenic impact that influences the leaching of minerals is confirmed by the reported study of groundwater in the adjacent regions ). This factor explains the mineral dissolution that occurred by the influence of anthropogenic sources. The spatial distribution of uranium in the present study reveals that the central part of the study area shows the uranium concentration exceeds the permissible limit of AERB (60 ppb) ( Figure 6). This higher uranium concentration may be due to the leaching of uranium minerals by the influence of the usage of phosphate & nitrogen-containing fertilizers in the agricultural area of this study region. The anthropogenic activities that induced uranium in the study region are confirmed by the previous studies in the semi-arid region of south India (Mathivanan et al. 2022). Factor 3 has the fluoride and bicarbonate parameters with positive values, illustrating that the groundwater alkaline condition favours the higher fluoride dissolution in the collected samples during both seasons. The earlier studies indicate that the alkaline state of the groundwater favours the higher dissolution of the fluoride minerals Neelakantan 2021a, Subba Rao et al. 2020). Under the alkaline condition, the following fluoride dissolution reaction may occur in the groundwater (Eq. (9)) Phosphate is present in factor 4, with the positive value representing the leaching of phosphate fertilizers that may be responsible for this factor. Factor 5 has positive values for pH and CO 3 2À indicating the following equilibrium reactions (Eqs. (10)-(12)) may occur in the subsurface of the study region. The adjacent regions address this carbonic acid equilibrium in the groundwater .
The quality of the collected water samples were classified as per the water quality index (WQI) and tabulated (Table 4). From the classification of water quality index, 40% of the pre-monsoon and 41% of the post-monsoon samples have the water quality from excellent to good water category. The remaining 60% (pre-monsoon) and 59% (post-monsoon) of the samples have the water categories between poor and unsuitable for drinking. The parameters, fluoride, chloride, nitrate, calcium, and magnesium, exceed the permissible limit of BIS (2012), indicating that these water samples is unsuitable for drinking.

Application of GIS
GIS technique is mainly used to highlight the higher uranium places and polluted water quality parameters. Figure 6 represents the spatial variations of the uranium. It is classified as green zone (<10 ppb), yellow zone (10-30 ppb), blue zone (30-60 ppb), and red zone (> 60 ppb). This uranium distribution plot of the study region shows the central part of the study area is more polluted with uranium (> 60) than the other places, and this may be due to the usage of phosphate fertilizers in agricultural activities. Most of the study regions are in the green and yellow regions, indicating that the locations are in the safer zones. The spatial distributions of other water quality parameters are shown in Figures S1-S8.

Conclusion
The groundwater in the study area (Virudhunagar district) has been polluted with calcium and magnesium due to mineral dissolution induced by the anthropogenic impact, especially nitrogen fertilizer used in agricultural activities. The lower uranium concentration found in the study region is mainly due to the alkaline and reducing nature of the groundwater. PHREEQC has confirmed the lower dissolution of uranium minerals, and U 4 O 9 is the predominant mineral species in both seasons. In the central part of the study area, the uranium levels have exceeded the permissible limit because the mineral dissolution of these places is higher by the impact of anthropogenic sources. Factor analysis confirms that the natural and anthropogenic effects are the primary source of groundwater pollution in the study area. The uranium distribution plot of the study region using the GIS identifies that the central part of the study area is more polluted with uranium (> 60) than the other places due to the usage of phosphate fertilizers in agricultural activities. This study suggests that the groundwater in this district has been polluted with fluoride, nitrate, chloride, calcium, and magnesium. Hence, water treatment must be needed before human beings consume the groundwater.

Ethical approval
This article does not contain any studies with human participants or animals performed by any of the authors.

Author statement
V. Raja: sample collection and analysis, Writing original draft; M. A. Neelakantan: Methodology, Project administration, Supervision.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This work is supported by financial support from the Board of Research in Nuclear Sciences, BARC, Mumbai, India, Grant no. 36(4)/14/72/2014-BRNS.

Data availability statement
The data that support the findings of this study are available from the corresponding author, Dr. M.A.N upon reasonable request.