Groundwater status in and around Study area
The domestic usage of the studied bore wells precludes to measure the groundwater levels in those wells. Hence, the data from nearest observation well at Malkajgiri village (~ 3 km south of the study area) monitored by State Groundwater Board, Telangana, has been used to get the best assessment of hydrodynamic conditions in the study area. Monthly groundwater levels measured in this piezometer of about 30 m deep are presented in figure 3 from 2005 to 2018 along with the monthly rainfall data. Seasonal groundwater level fluctuations in the observation well shows, in general between 10 to 15 m in various normal/above normal rainfall years, whereas < 5 m during the drought (2011 to 2015) period, with an overall drop in groundwater level of about 5 m between 2005 and 2018. 2011 being below normal rainfall year the raise in groundwater level between pre and post monsoon is only about 3 m, and further, the drought effect on groundwater system during pre-monsoon of 2012 is clearly seen, where the groundwater levels dropped down to 25 m (Fig. 3). The groundwater levels remained at about 25 m bgl till pre-monsoon 2016 with little monsoonal fluctuations. The drought effect on groundwater system is observed not only in the study area, but also entire Telangana state (CGWB, 2016). The drop-in groundwater level is highest in the Medak district in which the study area is located. During this period, most of the dug wells and the shallow wells had gone dry, indicating the desaturated phreatic aquifers. To meet the water demand, several deep bore wells were drilled in the study area, as well as in the Telangana State.
In the study area, the D1, D2 and D3 wells drilled during the severe drought period (2014-15) and started extraction from the deep aquifer. As these wells are being used for domestic purpose and quite deep, it is not possible to measure the groundwater levels.
Hydrochemistry
Based on the physico-chemical analysis of 57 (56 groundwater and one surface water) samples collected from the study area, statistical distribution of different ions is given in Table 2 while Table -3 shows the total data. A large variation is observed in almost all the ionic concentrations of shallow and deep well waters (Table 2) as well as a large variation within the deep well water chemistry.
For the four shallow well waters, the pH is less than 7, and the TDS ranges from 502 to 840 mg/L. The Ca and HCO3 concentrations are more than twice of the Na and Cl concentrations respectively. High NO3 concentrations ranging from 32 to 125 mg/l are found in all the 4 shallow well samples. In comparison to shallow groundwater samples, the single surface water sample from the granite quarry shows very high TDS (3108 mg/L), could be due to enrichment of ions owing to evaporation as the sample collected when only little water left in the quarry.
For the deep bore wells, the average pH value (~ 7.5) tends towards alkaline while the TDS ranges from 959 to 1862 mg/L. To study the temporal variations of different ions in deep wells, D1, D2 and D3 are sampled many times, whereas D4 and D5 were sampled only once along with other three deep wells in April 2015. Hydrochemistry of all these 5 wells is almost the same, as these wells encountered the same aquifer and are situated in close proximity. However, a large temporal variation is found in the hydrochemistry of deep well waters (Fig 4 of D1, D3 and D4 shown in Supplementary Information as SI-1 and 2). For these three deep wells, the dominating Cl ion from 2015 to first half of 2016 decreased sharply during the second half of 2016, whereas, the HCO3 increased. In D1, Na and Ca concentrations are almost same (~300 mg/l) during 2015 and first half of 2016 (Fig. 4), and later, both the ions decreased to less than half. However, water chemistry of D2 and D3 is observed to be slightly different from that of D1. During the first half of 2015, D2 and D3 show relatively lower Na and Ca (SI-1 and 2) than D1, while from second half of 2015 to first half of 2016, their Na and Ca concentrations are almost similar to that of D1. Subsequently, concentrations of all chemical parameters in waters of there 3 wells drastically decreased, whereas HCO3 and NO3 increased. Before September 2016, there were only traces of NO3 (except in April 2015) in all the 3 well waters, its concentration increased considerably afterwards.
Hydrochemical facies variations
Hydrochemical facies variation between the shallow and deep groundwater and the temporal changes in the deep groundwater are shown in the piper diagram (Fig. 5). Four shallow groundwater samples (S1 to S4) are mixed type, wherein 3 (except S1) samples have Ca and HCO3 ions dominating the others indicating the freshly recharged groundwater (Ray and Mukherjee, 2008, Reddy et al., 2009). In S1 well, Ca and Na concentrations are almost equal but slightly differ from those in other 3 samples (S2 to S4) indicating a little ionic exchange. Deep waters from wells D1, D2 and D3 collected during January 2015 to June 2016 (12 samples from each well) and D4 and D5 (one-time collection) are either Ca-Na-Cl or Ca-Na-Cl-SO4 type (Table 3, except one sampling of D1) considered as paleo-waters (Fig. 5). The sample from D1 collected on 16-04-2015 show Ca-Na-Cl-HCO3 type (Table 3). Sudden drop in Na, Cl and SO4 and increase in HCO3 changing the water type from Ca-Na-Cl to Ca-Na-Cl-HCO3, indicates mixing of fresh water. Though sampling was made from wells D1 and D2 merely a day after summer rain storm (on 16-4-2015), only D1 indicate fresh water mixing. In September 2016, D1 and D2 water type changed to Ca-Na-SO4-C1 with increased SO4 concentration and decrease in that of Cl. Further samplings between December 2016 to August 2019, HCO3 concentration increased while that of Cl and SO4 decreased in all the 3 deep well waters. This change in water quality (Fig. 5) indicates mixing or replacement of fresh water with paleo-water. Single quarry sample indicated mixed water type (Na-Ca-K-Cl-SO4).
Temporal variation of 14C activity and stable isotopes in groundwater
14C activity in the deep groundwater
14C activity measured in twenty samples from three deep well waters (D1 to D3) for the period 2015 to 2019 varies from as low as 35.41±0.48 pMC (uncorrected age of 8510±110 y BP) to 100 pMC (Modern, defined as 95% of the 14C activity for AD 1950, Stuiver and Pollach, 1977) (Fig. 6a, Table 3). For the first sampling, long residence time (time spent in the aquifer or time with reference to recharge, lowest pMC) is observed for the D1 sample (35.41 pMC (8510±110 y BP), followed by D2 (66.62 pMC, 3360±83 y BP) and relatively short residence time for D3 (85.01±0.8 pMC, 1340±78 y BP). This difference in pMC may be due to mixing of some water from other sheet joints (Cook et al., 2005). Fairly large fluctuations are observed in the 14C activity of the waters sampled from 2015 to mid-2016, wherein D1 and D3 show different pattern than D2. Since December 2016, all the water samples yielded “100 pMC” indicating fresh water influx into the aquifer. Change in 14C ages in deeper aquifer water is the result of mixing with shallow water (Takahashi et al., 2013). A good correlation is found between the HCO3 concentration and the 14C activity in the paleo-waters (Fig. 6b). However, mixing/replacement of fresh water perturbed the correlation, indicating the fresh water ingression in December 2016 sampling.
Temporal variation of stable isotopes in groundwater
Out of the four shallow well waters, S4 yielded depleted stable isotope values (δ18O -3.29 ‰ and δD -18.8 ‰), while the S2 and S3 waters show relatively enriched values δ18O (-1.7 to -1.0 ‰) and δD (-17 to -10 ‰). The well S1, located in the east of the quarry, shows highly enriched δ18O (8.28 ‰) and δD (27.69 ‰) values. As expected, the surface water sample from the granite quarry (an evaporative body) measured positive δ18O (2.04 ‰) and δD (7.77 ‰) values. Hence, the well S1, located on the down gradient of granite quarry, has expectedly maximum influence of quarry water and the S4 has minimum influence.
δ18O and δD values of majority of water samples from D1, D2 and D3 are around -2±0.5 ‰ and -17±2 ‰ respectively (Fig. 7, Table 3). However, the anomalous values are observed in all the three wells.
The local meteoric water line (LMWL: δD = 7.8713 * δ18O + 9.8933) drawn based on the rainwater data set of year 2010, is almost similar to Indian Meteoric Water Line (IMWL, δD = 7.93 * δ18O + 9.94, Kumar et al., 2010) as well as the global meteoric water line (GMWL: δD = 8 * δ18O +10: Craig, 1961). The δ18O vs δD plot (Fig. 7) for the groundwater samples from the study area shows linear relationship with a characteristic slope of 4.4 (δD = 4.4057* δ18O - 6.4356, R² = 0.9484), excluding the two sets of data (collected during 26-09-16 and 20-01-18) for the deep wells. Overall, the low slope value is indicative of the evaporation (He et al., 2012, Bahir et al., 2019). Stable isotope values (~ -2±0.5 ‰ of δ18O and ~ -17±2 ‰ of δD) of deep well waters with longer residence time, indicate fairly stable climatic conditions at the time of recharge (Roy et al., 2021). However, the enriched stable isotope values in the mixed water even after mixing of the recent water with paleo-waters certainly indicates that the source of this recent water has undergone evoporation. The single surface water sample from the quarry pond also fall on the evaporation line, indicating that this may be one of the prominent sources of the water mixing with the paleo-groundwater. Highly depleted δ18O and δD values for the 26-09-16 collection from D1 and D2 (Table 2) may be due to direct rainfall recharge through preferred pathways, where, few rainwater samples (during the August) also measured similar values. Probably, during this period, paleo-water in the deep aquifer might have totally exhausted and allowed the maximum fresh water input to the aquifer, changing the 14C activity to 100 pMC or ‘Modern’ (as indicated in December 2016 collection). The set of 3 deep samples (D1, D2, D3) collected on 20-01-18 indicate enrichment of the stable isotope values in different proportions, probably due to major contribution from an evaporative body.
Sources of recharge to shallow and deep aquifers
Broadly, the shallow and deep-water chemistry shows that, shallow waters are acidic type (pH <7) and deep waters are slightly alkaline type (pH ~7.5). Acidic type of water in shallow aquifer indicate direct rainfall recharge (generally, rainwater is acidic) either through direct percolation or through preferred pathways. Average TDS value of the shallow groundwater (656 mg/L) is less than half of the deep aquifer water (1381 mg/L), indicates that these two waters are different. The major ionic dominance in shallow water is HCO3 > Cl >SO2 and Ca > Na > Mg, whereas in the deep waters, it is Cl >SO2 >HCO3 and Na >Ca > Mg. Very high NO3 concentrations are observed in the shallow waters, whereas in most of the deep-water collections it is low. F concentrations are almost the same in the shallow and deep waters.
The maximum permissible limit of NO3 concentration in drinking water is 50 mg/l under WHO (2008) and 45 mg/L (BIS 2012) as per Indian guidelines and its concentration in groundwater is an indicator of anthropogenic pollution (Reddy et al., 2010, 2015, Wakode et al., 2014). The absence of drainage system for sewage in the study area compels individual house/apartments to have their own soak pits, which, generally are the sources of high NO3 concentration in phreatic aquifers (Reddy et al., 2015, Sridevi et al., 2017). In the present case, shallow well waters have NO3 concentrations measured between 30 to 125 mg/L. Deep groundwater (under confined like condition) from D1 to D5 wells yielded only <10 mg/L of NO3 until mid-2016, except one sample from D1 in the beginning of 2015 rainy season (Fig. 4). However, NO3 concentration increased many folds in the D1 to D3 wells water from September 2016 onwards (Figs. 4, and SI-1 and 2). The relation between Cl and NO3 concentrations clearly shows 2 groups of water (Fig. 8), i.e, paleo-waters and the recent recharge. While there is no relation between Cl and NO3 for paleo-waters, a good relation exists for recent recharge. Samples collected from shallow wells also falls in this group.
Sudden raise in groundwater and depletion of stable isotopes due to snow melt recharge has been attributed to extremely rapid and localized recharge to fractured rock aquifers (Gleeson et al., 2009). Sudden change in water chemistry (presence of high NO3, K, HCO3, Fig. 4) and enriched stable isotope values are observed for the collection of 16-04-2015 for D1 (a day after the summer rainstorm event). This data shows fast migration of fresh water from a evaporative surface water body through desaturated weathered/fractured zone and mixing with deep aquifer water. However, the hydrochemistry and stable isotope values regained to the previous values (prior to the rainstorm event) within a week (Fig. 4). It clearly indicates that the mixing of fresh water has influenced only a limited extent around the bore well and not the entire aquifer. This could be possible when fresh water entered through the well hole (like artificial recharge) connecting the shallow and deep aquifers. When the shallow aquifer thus got desaturated, the sheet joints in the quarry facilitated fast migration of the rainstorm water from the quarry to the shallow aquifer initially. Subsequently, the well holes (connecting the shallow aquifer and partially depressurized deep aquifer) provided the direct path for this fresh water (rainstorm water) to migrate from shallow to deep aquifer. Thus, the fresh water entered in to the deep aquifer through well hole, mixed with deep aquifer water mainly around the well and thereby, diluted the deep aquifer water. Further, subsequent pumping yielded less mineralized water for a short time, as indicated by the measurements of the water samples collected on 16-04-2015 (Fig. 4). Hence, the chemistry of pumped water changed back to the original concentration within a week (Fig. 4). April 2015 summer storm effect is minimal on D2 well, which is 50 m east of D1. Though sample could not be collected from the D3 well on 16-04-2015, lowering of certain ion concentrations observed in the sample collected on 22-05-15 (Fig. 5), indicates the influence of the fresh water mixing on this well.
Relatively lower 14C activity (higher ages) in all the 3 deep wells for June 2016 collection indicate presence of the paleo-water (Fig. 6a) till that time when the shallow aquifer was dry and hence, no freshwater contribution to the deeper aquifer. Change in hydrochemistry for September and December 2016 collection and 100 pMC 14C activity for December 2016 collection shows fresh water ingression into the deep aquifer. It may be inferred that excessive extraction from the limited potential deep aquifer during the drought conditions in 2015, desaturated it to a great extent and thus creating favorable conditions to receive more fresh water during the excess rainfall year 2016. Due to the heavy rainfall, initially it recharged the shallow aquifer and subsequently, the same migrated to deep aquifer through the 3 studied deep wells and probably through many other such deep wells existing in the surrounding area. The steep increase in HCO3 and NO3 and simultaneous drop of SO4 and Cl provides credible evidence that young/fresh groundwater has refilled the sheet joints emptied by excess pumping. It is further confirmed by the hydrogeochemical measurements and 14C activity of the 3 deep wells waters in August 2019 and stable isotope data (Table 2).
Conceptual hydrogeological model of the study area.
Occurrence of sheet joints in the granitic terrain is a common phenomenon (Maheshwari et al., 2013). Quite often, they intersect with vertical joints at shallow level (Devandal et al., 2006, Gleeson et al., 2009, Collins et al., 2020). Most of the previously reported conceptual models for granitic/hard rock regions related to groundwater, mainly focused on the top weathered and fracture zones (Devandal et al., 2006, Lachassagne et al., 2011, Guihéneuf et al., 2014) and barely discuss the role of blind/open sheet joints as potential aquifers at greater depths. Based on the drilling information, well construction details of studied wells, field conditions and observations, a conceptual model (Fig. 9) is proposed to understand the groundwater dynamics in the present study area.
The large variations observed in hydrochemistry and stable isotope values for the shallow (60 to 100 m) and deep (~400 m) wells waters clearly indicate that the respective aquifers were not communicating in the absence of connecting vertical joints/fractures and hence, the deep aquifer was under confined like condition. Drilling information of 5 deep wells shows that, for all the wells, an aquifer was encountered almost at the same depth (~ >360 m) after crossing the hard and compact granite, thereby implying the existence of deep aquifer in this area. Low 14C activity in the deep groundwater (long residence time) specifies stagnation/long flow paths of groundwater in the discrete/blind sheet joints. Sudden change in hydrochemistry and stable isotope values only in D1 well water (~ 300 m east of granite quarry), merely a day after the summer rainstorm event, subsequent to a prolonged drought, indicates fast migration of fresh water from the nearby surface water body to the phreatic aquifer and further to the deep aquifer. Fast migration of freshwater to deep aquifer is possible only by transfer through the bore well holes which connect the phreatic and the deeper aquifer. Field observations in the granite quarry show a few sheet joints dipping towards east, which might be connected to the phreatic aquifer. When there was no water in the phreatic aquifer due to severe drought during 2014, the deep groundwater (paleo-water) was being exploited. This probably led to drastic reduction in quantum of the paleo-groundwater in the deep aquifer, which got replenished with fresh water during the high rainfall year (2016) and continued up to 2019, as evidenced from the hydrochemical and isotopic data of this study. This data complemented by the longer residence time for the three deep well waters (D1, D2 and D3) from January 2015 to June 2016 are interpreted in terms of entrapment/long flow paths to the deep fracture system, where the aquifer is under confined like condition. Such a confinement within the solid granite could be possible, provided there are no or limited vertical joints crosscutting the deep and shallow aquifers. As long as the entrapped groundwater is under confined like condition, probability of the recent recharge to the deep aquifer stands minimal. Initiation of groundwater extraction from this deep aquifer in 2014 turned into over-exploitation by 2015 through drilling of additional deep wells due to which, the aquifer with limited potential was de-saturated during the drought conditions in 2014 and 2015. During the high rainfall conditions (2016) initially, the shallow aquifer replenished with (i) fresh groundwater through natural rainfall recharge (ii) induced recharge through fracture network from granite quarry as well as (iii) from surrounding surface reservoirs. Once the deep aquifer got connected with the shallow one through several deep well holes, initially a part of the recharged water to shallow aquifer migrated to deep aquifer. Upon saturation of the deep aquifer, water levels in the shallow aquifer would rise. In such a scenario, the deep aquifer sustains owing to their communications and receiving induced recharge from the shallow aquifer.