Introduction

Groundwater is the most precious natural resource for human survival all over the world (Wu et al. 2012). Groundwater quality, however, is deteriorating at an alarming rate due to the changing environment and intensive human activities (Li et al. 2012), which poses significant health risks to people who consume it and take baths with it. Numerous incidents of groundwater contamination have been reported in the USA (Hudak 2010), China (Han et al. 2013; Li et al. 2014a; Liu et al. 2007), India (Chidambaram et al. 2014), and many other regions of the world (e.g., Mexico, Esteller et al. 2015, and Korea, Venkatramanan et al. 2014). Groundwater contamination is especially serious in alluvial plain of northwest China because of high population density in the alluvial plains, intensive human activities, high water demand, and vulnerable environment (Ma et al. 2009). The Western Development Program (WDP), initiated in 2000 by the central government of China, introduces a number of industries to this area, bringing along serious contamination to groundwater while creating wealth and prosperity to local residents. To change this situation, Chinese government has recently released a series of national plans, guidelines, and laws to combat groundwater pollution: for example, the National Plan of Groundwater Pollution Prevention for the years 2011–2020 (Ministry of Environmental Protection of the PRC 2011a), the Technical Guidelines for Environmental Impact Assessment: Groundwater Environment (Ministry of Environmental Protection of the PRC 2011b), and the new Environmental Protection Law of the People’s Republic of China that has taken effect on January 1, 2015 (Standing Committee of the National People’s Congress of the PRC 2014).

Due to the increasing importance of water-quality assessment in water resources management, a number of scholars have performed this important work in various regions by a variety of methods (e.g., Al-Rawabdeh et al. 2014; Golekar et al. 2013; Rajesh et al. 2015; Vetrimurugan et al. 2013; Vetrimurugan and Elango 2014). Recently, Rao et al. (2013) applied electrical resistivity tomography (ERT) methods in the delineation of groundwater contamination and potential zones in India. Such geoelectrical techniques were also adopted in the research conducted by Metwaly et al. (2013) to characterize the lateral distribution of contamination plumes. Zhang et al. (2013) estimated the nitrogen loss and predicted the nitrate concentration in groundwater in an agriculture-impacted aquifer of the UK using the MT3DMS and GIS techniques. Li and Qian (2011) performed an assessment on drinking water quality and human health risk because of chemical pollutants in groundwater in an industrialized region of China. Similarly, health risk assessment was conducted by Wang et al. (2014) and Giri and Singh (2015) independently to examine the adverse impacts of groundwater contaminants on human health. These studies showed that ingestion of the contaminated water was a major pathway of human health risk. Barroso et al. (2014) proposed an integrated approach for the assessment of groundwater pollution in an agricultural area of Portugal. This integrated approach included geological and hydrogeological mapping, multivariate analyses, GIS techniques, and aquifer-vulnerability evaluation. Nunes et al. (2013) proposed a space-filling coverage method of optimizing groundwater-monitoring networks for plume detection and quantification. This method is advantageous in terms of simplicity and implementation speed. Multivariate analysis is a useful technique widely used in water-quality assessment and pollution source identification (Subba Rao 2014; Wu et al. 2014; Zumlot et al. 2013). In addition, neural network (Sirat 2013), water-quality index (Amiri et al. 2014; Li et al. 2010), and fuzzy theory (Li et al. 2013a, 2014a) among many others are also popular approaches in water-quality assessment and groundwater contamination identification. All these methods and techniques act as effective tools in groundwater-quality assessment and management.

Guanzhong Plain is located in the mid-south of Shaanxi province of China. It is a part of the Weihe River basin, and the Weihe River is the largest tributary of the Yellow River (Huang et al. 2015). Rapid industrialization and urbanization in this area have brought environmental problems and water crisis, which, however, trigger the fast development of water-related research. Most recently, Yang et al. (2015) assessed the status of heavy metal pollution in Weihe River as well as the associated health risk, and they found that the tributaries of the Weihe River showed more serious pollution than the main river, and Arsenic was the most important pollutant imposing adverse health risks to local residents. However, prior to Yang et al. (2015), the overall water quality of the Weihe River was assessed by Lu et al. (2010) using an entropy-weighted fuzzy synthetic evaluation, who found that the river was heavily polluted with Hg as the major pollutant. The health of the Weihe River is influenced by natural and anthropogenic factors (Guo et al. 2014). The Baojixia water diversion is a typical anthropogenic factor influencing the hydrological regime of the Weihe River (Guo et al. 2014). The research conducted by Li et al. (2014b) indicated that human activities and the mixing of different recharge water could have significant impacts on the fluoride concentration in shallow groundwater. In addition, the impacts of climate change on river runoff and river environment have also been widely investigated in this area (e.g., Huang et al. 2014, 2015; Zuo et al. 2014; Wang et al. 2015). However, most of these studies focus at river water contamination and hydrological regime impacted by natural variation and human activities. Research on groundwater contamination and associated human health risks is rather limited. Therefore, the objectives of the present study are (1) to identify the major pollutants and the pollutant load distribution in the shallow groundwater and (2) to assess the status of groundwater quality and its potential risks to human health. This study will be helpful to local decision makers for protecting the groundwater quality and utilizing the groundwater resource more effectively. International scholars may also find it useful in similar studies.

Study Area

The study area is a part of the Guanzhong Plain. It is situated in the south of the plain, and is bounded by the Weihe River in the north, Qinling Mountains in the south, Yuxian River in the west, and Luowen River in the east (Fig. 1). It extends over about 15 km from west to east and 13 km from south to north, covering approximately 175 km2 in area. The Shidi River runs through the study area in the middle from south to north and finally flows into the Weihe River. Within the study area, two landforms can be classified according to the shaping forces: inclined alluvial-pluvial plain in the south and alluvial plain in the north. The alluvial plain where most of the villages are located is flat. There is a fertilizer plant (Shanhua) situated in the inclined alluvial-pluvial plain, which discharges industrial wastewater directly into the Shidi River, causing serious contamination to river water and groundwater around.

Fig. 1
figure 1

Location of the study area and sampling sites

The study area is situated within the warm semi-humid continental monsoon climate zone (Li et al. 2014b). According to the data from the weather station in Hua County, the mean temperature in this area is 13.4 °C, and the highest and lowest temperatures in record were, respectively, 43 and −16.5 °C (observed on June 19, 1966 and January 30, 1997, respectively). The average annual rainfall is 581.2 mm, with about 50 % being concentrated in July, August, and September (Zhang 2014). Evaporation in this area is medium, and the annual mean evaporation rate is 830.7 mm. The evaporation from April to August accounts for over 65 % of the total evaporation in a year (Zhang 2014).

Two types of aquifers can be identified within the Quaternary deposits in this area: phreatic aquifer and confined aquifer (Zhang 2014). The phreatic aquifer composed of alluvial sands and coarse sands is the main aquifer for water supply, and the confined aquifer consisting of sands, fine sands, and thin clayey layers provides only a small portion of water for various uses. Groundwater flows generally from south to north, but groundwater cones have been formed locally in and around Lijia (Fig. 1) because of heavy groundwater abstraction (Zhang 2014). According to the water balance calculation, lateral inflow, precipitation infiltration, irrigation infiltration, and river leakage are the main components of groundwater recharge, accounting for 43.51, 19.79, 19.14, and 17.56 % of the total groundwater recharge, respectively. Artificial abstraction is the most important pattern of groundwater discharge, accounting for over 90 % of the total discharge, while the discharges by evaporation and lateral outflow are minimal, accounting for only 7.68 and 1.29 %, respectively.

The study area is a traditional agricultural region, which supports farming activities such as vegetable production and greenhouse cultivation throughout the year. The intensive agricultural activities, however, bring serious nonpoint source pollution to soils and groundwater. The effluents casually discharged by the fertilizer plant worsen this situation further, posing great risks to human health of the local residents. Therefore, it is necessary and meaningful to carry out this study to provide more information for groundwater-quality protection and management.

Materials and Methods

Sample Collection and Analysis

For this study, 90 samples were collected from shallow pumping wells during October 2013. The wells were pumped for several minutes to eliminate the influence of static water. Samples were collected using 5-L white plastic bottles, and they were washed using the water to be sampled before sampling. The sampling sites were recorded by a potable GPS, and the locations are shown in Fig. 1. After collection, these bottles were labeled to avoid any possible error between collection and analysis. All the collected samples were transported to the laboratory immediately for physiochemical analyses. The procedures of sample collection, transportation, and conservation followed the standard methods of the APHA (Rice et al. 2012).

The groundwater samples were analyzed for major ions (Na+, K+, Ca2+, Mg2+, HCO3 , SO4 2−, and Cl), fluoride (F), nitrate (NO3 ), nitrite (NO2 ), chemical oxygen demand (CODMn), total hardness (TH), total dissolved solids (TDS), and pH. The pH was measured in situ using a potable pH meter; Na+ and K+ were analyzed by flame photometer; Ca2+, Mg2+, and TH were analyzed using the EDTA titration method; F was determined by fluoride-selective electrode; and CODMn was measured using KMnO4 as the oxidant. TDS was measured by drying at 180 °C and then weighing; and SO4 2−, Cl, NO3 , and NO2 were found by means of ion chromatography. During the analysis, duplicates were introduced for QA/QC.

Groundwater-Quality Assessment

The Quality Standard for Groundwater of China (Bureau of Quality and Technical Supervision of China 1993) classifies groundwater quality into five grades using a comprehensive water-quality index (CWQI): excellent (grade I), good (grade II), fair (grade III), poor (grade IV), and very poor (grade V). According to the standard, CWQI can be computed as follows (Li et al. 2014c):

$$ F = \sqrt {\frac{{(F_{i} )_{\text{mean}}^{2} + (F_{i} )_{\hbox{max} }^{2} }}{2}} \quad \left( {i = 1,{ 2},{ 3}, \, \ldots ,n} \right) $$
(1)
$$ (F_{i} )_{\text{mean}} = \frac{1}{n}\sum\limits_{i = 1}^{n} {F_{i} }\quad \left( {i = 1,{ 2},{ 3}, \, \ldots ,n} \right) $$
(2)

where n denotes the number of indices selected for the assessment, F is the value of CWQI for a given sample, and F i represents the assigned value for the ith index by the Quality Standard for Groundwater (Bureau of Quality and Technical Supervision of China 1993). (F i )mean is the mean of F i , and (F i )max is the maximum value of F i . Once F is determined, groundwater-quality classification can be obtained as per the following criteria: F < 0.8 indicates excellent water, 0.8 < F ≤ 2.5, good water, 2.5 < F ≤ 4.25, medium water, 4.25 < F ≤ 7.2, poor water, and F > 7.2 indicates very poor water (Li et al. 2014c)

As groundwater is also used for irrigation in the area, groundwater quality for irrigation purpose was also assessed in this study in terms of SAR, Na%, and RSC which are expressed as follows (Davraz and Özdemir 2014; Li et al. 2013b):

$$ {\text{SAR }}=\frac{{{\text{Na}}^{ + } }}{{\sqrt {\frac{{{\text{Ca}}^{{ 2 { + }}} {\text{ + Mg}}^{{ 2 { + }}} }}{ 2}} }} $$
(3)
$$ {\text{Na}}\% =\frac{{{\text{Na}}^{ + } {\text{ + K}}^{ + } }}{{{\text{Na}}^{ + } {\text{ + K}}^{ + } {\text{ + Ca}}^{{ 2 { + }}} {\text{ + Mg}}^{{ 2 { + }}} }} \times 1 0 0 $$
(4)
$$ {\text{RSC }}=\left( {{\text{HCO}}_{3}^{ - } {\text{ + CO}}_{3}^{{ 2 { - }}} } \right){ - }\left( {{\text{Ca}}^{{ 2 { + }}} {\text{ + Mg}}^{{ 2 { + }}} } \right) $$
(5)

where all ions are expressed in meq/L. SAR is the sodium adsorption ratio, a measure of the sodicity of soil (Davraz and Özdemir 2014). Na% represents the sodium percentage, and RSC is the residual sodium carbonate.

Human Health Risk Assessment

Contaminated water can pose risks to human health through several pathways of exposure such as water intake, breath, and direct dermal contact. The Unites States Environmental Protection Agency (USEPA) has established models for human health risk assessment (Kim et al. 2004; Li and Qian 2011; Momot and Synzynys 2005; Wu et al. 2012). In 2014, the Ministry of Environmental Protection of the PRC released the new Technical guidelines for risk assessment of contaminated sites (Ministry of Environmental Protection of the PRC 2014). The human health risk assessment models recommended in the new technical guidelines are on the basis of the USEPA models, but are assigned unique parameters to reflect the actual contamination situations in China. Similar to the USEPA models, four steps are involved during the assessment: hazard identification, dose–response assessment, exposure assessment, and risk characterization (Li and Qian 2011; Momot and Synzynys 2005). Based on the data available, NO3 , NO2 , and F were selected as the parameters for risk assessment in the present study. These parameters are non-carcinogenic pollutants according to the International Agency for Research on Cancer (IARC) and the USEPA, and therefore, only the non-carcinogenic risks were assessed in this study. There are mainly two exposure pathways: oral intake (drinking) and dermal intake (showering). The risks through the two pathways can be calculated as follows:

Non-carcinogenic risk through oral intake is calculated as follows:

$$ {\text{Intake}}_{\text{oral}} = \frac{{{\text{C}} \times {\text{IR}} \times {\text{EF}} \times {\text{ED}}}}{{{\text{BW}} \times {\text{AT}}}} $$
(6)
$$ {\text{HQ}}_{\text{oral}} = \frac{{{\text{Intake}}_{\text{oral}} }}{{{\text{RfD}}_{\text{oral}} }} $$
(7)

where Intakeoral represents the daily average exposure dosage through oral pathway per unit weight (mg/(kg·d)), and C is the concentration of the pollutant in water (mg/L). IR indicates the ingestion rate of water through drinking (L/d). In this study, the ingestion rate of water is 1.5 L for adults and 0.7 L for children under 12 years old according to statistical investigation. EF and ED are exposure frequency (d/a) and exposure duration (a), respectively. The EF is 365 days per year for both adults and children, but ED is assigned 30 years for adults and 12 years for children. BW is the body weight of a person (kg) with 60 kg for adults and 15 kg for children in the present study. AT is the average time for non-carcinogenic effect (d), and in this study, the average times for adults and children are, respectively, 10,950 days and 4380 days. HQoral denotes the hazard quotient of non-carcinogenic risk by oral intake pathway, and RfDoral means reference dosage for non-carcinogenic pollutant through oral intake pathway. The values of RfDoral for NO3 , NO2 , and F are 1.6, 0.1, and 0.06 mg/(kg·d), respectively.

Non-carcinogenic risk through dermal intake is calcuated as follows

$$ {\text{Intake}}_{\text{dermal}} = \frac{{{\text{DA}} \times {\text{EV}} \times {\text{SA}} \times {\text{EF}} \times {\text{ED}}}}{{{\text{BW}} \times {\text{AT}}}} $$
(8)

where Intakedermal represents the daily average exposure dosage through dermal intake per unit weight (mg/(kg·d)), and EV is the daily exposure frequency of dermal contact event (1/d). In the present study, EV is assigned 1 for adults and children, which means that residents in this area take a shower every day. DA indicates the exposure dosage of every single event (mg/cm2), and it can be estimated using Eq. (9), where K is the coefficient of permeability of the skin (cm/h), C is the concentration of the pollutant in water (mg/L), t is the contact time for a single shower (h/d), and it is approximately 0.4 h per day for adults and children according to statistical investigation. CF is a conversion factor which equals to 0.001. SA in Eq. (8) represents the skin surface area and can be estimated using the empirical Eq. (10), where H denotes the height of a person. The other parameters have the same meaning as described earlier.

$$ {\text{DA}} = K \times C \times t \times {\text{CF}} $$
(9)
$$ {\text{SA}} = 239 \times H^{0.416} \times {\text{BW}}^{0.517} $$
(10)

The hazard quotient (HQ) of non-carcinogenic risk by dermal pathway can be determined by Eq. (11), where RfDdermal represents the reference dosage for non-carcinogenic pollutant through dermal pathway. The values of RfDdermal for NO3 , NO2 , and F are 0.08, 0.05, and 0.06 mg/(kg·d), respectively.

$$ {\text{HQ}}_{\text{dermal}} = \frac{{{\text{Intake}}_{\text{dermal}} }}{{{\text{RfD}}_{\text{dermal}} }} $$
(11)

Total Non-carcinogenic Risk

The total non-carcinogenic risk is represented by hazard index (HI). HI < 1 means the non-carcinogenic risk is acceptable, while HI > 1 indicates the risk is beyond the acceptable level. The HI of a given pollutant through multiple pathways can be calculated by summing the hazard quotients by Eq. (12). The total risk posed by multiple contaminants can be expressed as Eq. (13):

$$ {\text{HI}}_{i} = {\text{HQ}}_{\text{oral}} + {\text{HQ}}_{\text{dermal}} $$
(12)
$$ {\text{HI}}_{\text{total}} = \sum\limits_{i = 1}^{n} {{\text{HI}}_{i} } $$
(13)

where HI i is the hazard index of non-carcinogenic contaminant i, and HItotal is the total hazard index of all non-carcinogenic contaminants concerned.

Results and Discussion

Levels and Spatial Distribution of Major Parameters

Table 1 shows the minimum, maximum, median, mean, and standard derivation of the main physiochemical parameters. The pH ranges from 7.72 to 8.31 with an average of 8.13, which indicates that the groundwater in the study area is slightly alkaline. The dominance of cations in groundwater is Ca2+ > Na+ > Mg2+ > K+ according to their mean values. Sodium may originate from dissolution of halite and cation exchange, as halite is a common mineral in the Guanzhong Plain and cation exchange is also a widely reported process controlling the groundwater chemistry in this alluvial plain (Li et al. 2014b). The observed range of sodium has been placed between 9.02 and 229.00 mg/L with an average of 53.56 mg/L. Calcium is a common element in the environment. The calcium concentration in the study area ranges between 44.10 and 408.80 mg/L with an average of 119.10 mg/L. The dissolution of such mineral as calcite and feldspars is the main source of calcium, because calcite and feldspars are of geogenic origin and are abundant in the aquifer media of Guanzhong Basin, and their dissolutions have been reported (Li et al. 2014b). Magnesium is also a commonly found element in the environment. Its concentration has been observed to be in a range of 4.30–144.60 mg/L with a mean of 35.01 mg/L. The concentration of K+ in groundwater is usually much lower than Na+. It ranges from 1.84 to 117.60 mg/L in this study, and 11 samples exceed the limit prescribed by the WHO drinking standard (WHO 1997). In case of anions, the observed range of HCO3 is 213.60–738.30 mg/L with a mean of 368.32 mg/L, and two samples exceed the standard limit for HCO3 for drinking purpose (WHO 1997). The range of SO4 2− is observed between 12.00 and 689.20 mg/L, indicating that some locations exceed the permissible limit for drinking purpose (250 mg/L). The highest concentration of Cl is 257.00 mg/L, a little higher than the permissible limit (250 mg/L). The minimum of Cl is only 7.10 mg/L, and the average value is 58.37 mg/L. The distributions of SO4 2− and Cl are shown in Fig. 2. In general, SO4 2− and Cl concentrations show an increasing tend from south to north, but some high concentration spots can be observed, especially around Xiluo, Xiamiao, Nanzhai, and Shanfu. The spatial viability of the ions may be attributed to human activities such as fertilizer application, sewage, and animal waste disposals.

Table 1 Statistics of major physiochemical parameters of collected samples
Fig. 2
figure 2

Distribution of major contaminants in shallow groundwater. a SO4 2−, b Cl, c TDS, d TH, e NO3 , and f F

TDS represents the dissolved salts in groundwater and is usually used to determine the suitability of groundwater for drinking purpose (Rajesh et al. 2015). TDS in the study area ranges from 248.00 to 2142.00 mg/L with an average of 685.64 mg/L, which indicates that some samples are brackish water that is unsuitable for drinking. TH represents approximately the sum of Ca2+ and Mg2+. The TH is in the range of 200.20–1283.70 mg/L with an average of 441.58 mg/L, indicating that some locations contain extremely hard water that cannot be used for domestic uses. The spatial distributions of TDS and TH look similar in Fig. 2. Very high concentrations of TDS and TH are observed in the regions downstream Shanhua and around Xiluo, which indicates that the effluents discharged by Shanhua can have significantly negative impacts on shallow groundwater quality around the fertilizer plant. High concentrations of TDS and TH near Xiluo may be explained by the agricultural activities from local residents. This can be further approved by the distribution of NO3 . It is widely accepted that NO3 is an effective indicator of agricultural contamination (Alexakis et al. 2012; Dar et al. 2010; Suthar et al. 2009). The concentration of NO3 (as N) is in the range of 0.55–186.70 mg/L with a mean of 20.21 mg/L. Several high nitrate zones have been found in Fig. 2e. The one discovered around Shanhua is attributed to industrial pollution, while the others such as the one around Xiluo and the one near Nanzhai are caused by agricultural activities. F is an essential element for human body at low concentration, but can be harmful when excessive. The WHO standards set 1.5 mg/L as the permissible limit for F in drinking water, while the national standards set this permissible limit to 1.0 mg/L. The F concentration has been observed to be in the range of 0.27–1.26 mg/L with an average of 0.53 mg/L. Although three samples have F slightly higher than the national limit, all are, however, within the WHO limit. The mean of F is lower than 1.0 mg/L, which means mixing of different water samples can reduce the F concentration, making it meet the national standard limit. High F concentration is mainly found in the middle, the southeast, and the northeast of the study area (Fig. 2f). The high level of F in groundwater is mainly attributed to lithological reasons in the study area. The aquifer media in this area are rich in fluorine-containing minerals such as fluorite (Li et al. 2014b; Zhang 2014).

Evaluation of Groundwater Quality

Drinking Water Quality

Comparison with national standards in Table 1 reveals that TH, NO3–N, NO2–N, TDS, and SO4 2− are the main contaminants in shallow groundwater of the study area. In addition, F, Cl, Na+, and CODMn also exceed slightly the permissible limits. The standards for drinking water quality (Ministry of Health of the PRC and Standardization Administration of the PRC 2006) and the quality standard for ground water (Bureau of Quality and Technical Supervision of China 1993) have set 450 mg/L as the permissible limit for TH for drinking purpose. In the study area, one third of the collected samples have TH concentration higher than the standard limit. The permissible limits for NO3 (as N) and NO2 (as N) are 20 and 0.02 mg/L, respectively, for drinking purpose. It is observed that 28 samples have NO3 concentration exceeding the standard limit and 27 higher than the permissible limit for NO2 in this study, accounting for 31.11 and 30 % of the total samples, respectively. Especially, the highest NO3−N and NO2−N concentrations are, respectively, 9 and 38 times higher than the standard limits, making it one of the most serious nitrate-polluted regions in the world. According to our field investigation and resident interview, there have already been many adults who have died of gastric carcinoma in adjacent regions, which is considered to be caused possibly by groundwater nitrogen pollution. Angelopoulos et al. (2009) reported serious nitrate pollution in Diakopto municipality, Greece, and found that nitrate concentrations were higher than 50 mg/L. Similarly, Aksoy and Scheytt (2007) found serious groundwater nitrate pollution in an agricultural region of Turkey with NO3 higher than 100 mg/L, and Cheong et al. (2012) reported that groundwater in the Gimpo agricultural area, South Korea was seriously contaminated by nitrate and displayed an average NO3 concentration of 79.4 mg/L. All these reports have indicated that agricultural activity is an important factor causing serious groundwater nitrate pollution. However, in the present study, the highest nitrate concentration was found near the fertilizer plant, indicating industrial point source pollution in addition to agricultural non-point source pollution in this area.

F is the most electronegative and reactive element on earth and is beneficial to human health in trace amount (Li et al. 2014b; Wu et al. 2012). Groundwater fluoride pollution is a universal problem, and it is reported that there are 70 million people from 27 countries threatened and affected by fluorosis (Ghosh et al. 2013). There are three samples possessing F concentration higher than the permissible limit in the study area, accounting for 3.33 % of the total samples. This is not a serious situation with regard to fluoride contamination in the present study, as its concentration can be lowered by mixing different water samples. Therefore, it is highly recommended that local residents mix groundwater pumped from different wells before using the groundwater for drinking purpose so that the F concentration can be lowered and it will not have adverse effect on their health. This simple measure can also be applied to lower the concentrations of Na+ and CODMn, making them suitable for drinking.

The CWQI was applied to provide a simple measure of the overall groundwater quality. The results are shown in Table 2. A map showing the spatial distribution of the overall groundwater quality is also generated in Fig. 3 to provide better understanding of the groundwater pollution status. The assessment reveals that 34 samples are excellent and good quality water (grades I and II) that can be readily used for drinking. One sample is of medium quality (grade III) that can be used for drinking with caution and pretreatment, 23 samples are of grade IV, and 32 belong to grade V. The last two grades represent poor and very poor quality water that is unsuitable for drinking. The poor and very poor quality water accounts for 25.56 and 35.56 % of the total, respectively, suggesting that more than half of the samples have been seriously contaminated and are thus unfit for human consumption. The severity of the pollution is also evidenced by Fig. 3. The entire study area is widely covered by poor and very poor groundwater, especially in the downstream of the fertilizer plant and regions along the Weihe River and Shidi River, which is indicative of the influence from the effluents discharged by the plant and leakage of polluted river water. In the upstream of the fertilizer plant, groundwater is usually of good quality, as it is far from the residential and industrial regions and has not been seriously influenced. Overall, groundwater quality for drinking purpose is not satisfactory in the study area, and the fact that groundwater is the dominant source of water for human consumption makes the situation worse. Long-term consumption of the contaminated groundwater will pose significant risks to human health. Therefore, measures for groundwater contamination remediation and safe water supply should be implemented as soon as possible.

Table 2 Groundwater quality classification based on CWQI
Fig. 3
figure 3

Map of overall groundwater quality based on CWQI

Irrigation Water Quality

As groundwater is also used for irrigation in the study area, it is necessary to assess its suitability for this purpose. The suitability of groundwater quality for agricultural purposes is determined by salinity and alkalinity of groundwater (Thilagavathi et al. 2012). Higher salinity will reduce the osmotic activity of plants and prevent water from reaching the branches and leaves of plants, resulting in inferior production (Marghade et al. 2011; Srinivasamoorthy et al. 2014). Electricity conductivity (EC) is a typical measure of salinity. In general, EC should be less than 2250 μS/cm in irrigation water. Higher concentration of sodium in irrigation water will affect the soil permeability and make the texture of soil hard to plough (Vetrimurugan and Elango 2014). SAR and  %Na are good measures of alkali/sodium hazard to crops. Na% should be lower than 60 % in irrigation water (Bouderbala 2015), and SAR lower than 18 is favorable for irrigation water but may be acceptable if it is lower than 26 (Rao et al. 2012). Groundwater with RSC < 1.25 meq/L is safe for irrigation, and if the RSC value varies in the range of 1.25–2.25 meq/L, it is of acceptable quality, and if more than 2.25 meq/L, then it becomes unsuitable for irrigation (Jeelani et al. 2014).

The results listed in Table 1 show that the SAR values of all samples are less than 10 and Na% less than 60, indicating suitability of groundwater for irrigation. Using such groundwater for irrigation will induce no alkalinity risk. The range of RSC has been observed between −16.90 and 1.92 with an average of −2.83, suggesting good and acceptable quality for irrigation. In the present study, the analytic data were also plotted on the US salinity diagram (USSL 1954) and the Wilcox diagram (Wilcox 1948), both of which consider alkalinity and salinity of irrigation water simultaneously. The US salinity diagram (Fig. 4) shows that all samples were plotted in C2S1 and C3S1 except only a few in C4S1, indicating a good-to-acceptable quality for irrigation. The samples plotted in C4S1 contain high salinity, and are not suitable for irrigation. The Wilcox diagram (Fig. 5) also indicates all samples are suitable for irrigation except those plotted in zones IV and V. Samples plotted in zones IV and V are unsuitable for irrigation as they contain high salinity.

Fig. 4
figure 4

USSL diagram for irrigation water quality classification

Fig. 5
figure 5

Wilcox diagram for irrigation water quality classification

Overall, groundwater in the study area is generally suitable for irrigation except only a few samples that contain high salinity. It is therefore suggested that some measures need to be taken to reduce the salinity of groundwater. The most convenient and cheapest method is probably to mix different types of water. The mixing can effectively reduce the salinity content in irrigation water.

Health Risks

Groundwater is seriously contaminated by TH, NO3 , NO2 , TDS, SO4 2−, and F, making it unsuitable for drinking and posing high risks to human health. NO3 , NO2 , and F are considered non-carcinogenic pollutants to human body, and therefore they are selected to perform the non-carcinogenic health risk assessment in this study. The assessment results of health risks for adults are shown in Table 3, and the results for children are shown in Table 4. Two maps illustrating the risk distribution for adults and children generated on the basis of HItotal are shown in Fig. 6.

Table 3 Calculated adults hazard quotient and hazard index from collected samples
Table 4 Calculated children hazard quotient and hazard index from collected samples
Fig. 6
figure 6

Health risks distribution maps of adults and children generated based on HItotal. a For adults and b for children

Tables 3 and 4 show that the risks caused by ingestion of contaminated groundwater are much higher than that induced by dermal contact for both adults and children. The greatest risks because of dermal contact (HQdermal) for adults and children are 0.1172 and 0.1896, respectively, while the highest risks caused by oral ingestion (HQoral) for adults and children are 13.61 and 25.41, respectively, two orders of magnitude higher than the risks caused by dermal contact. Similar conclusion can be reached by comparing the mean values of HQdermal and HQoral. This indicates that oral ingestion is the most important pathway of human health risks, while the risks due to dermal contact are relatively negligible.

With respect to risks induced through drinking pathway (HQoral), the risk order is NO3  > F > NO2 , and similar order is obtained for risks posed by dermal contact according to mean risk values. According to mean values of HI, NO3 contributes most to the total risk for adults and children, and then F. NO2 contributes least to the total health risk. This suggests that NO3 is the most dangerous element responsible for high human health risk and then F. With respect to the total risk from multiple contaminants (HItotal), it ranges from 0.17 to 13.73 with a mean of 1.72 for adults, and for children the total risk ranges between 0.32 and 25.60 with an average of 3.20. There are 37 samples with HItotal > 1 for adults and for children the number is 53, which indicates that children are more vulnerable to contaminants than adults. In general, residents living in the study area are at high health risk, and more people may become prone to disease in the future.

The health-risk distribution maps for both adults and children show some similarities to nitrate distribution map (Fig. 2e), as the total risk is contributed mainly by NO3 . The region with the highest risks is observed near the fertilizer plant, indicating that the fertilizer plant is the biggest threat to human health. The risky region (HQtotal > 1) for children is bigger than that for adults, indicating again that children are more vulnerable to contaminants than adults,and thus should be given more care.

Because groundwater contamination is becoming even more serious than ever before due to fast industrial development and extensive application of fertilizers in agriculture, causing cancers in human body (Zhang et al. 2014a), similar health-risk assessment studies have been carried out all over the world (e.g., Cai et al. 2015; Giri and Singh 2015; López-Roldán et al. 2015; Navoni et al. 2014; Wongsasuluk et al. 2013; Yang et al. 2012). In the present study, a comparison regarding health risk due to groundwater contamination has been made among similar regions in China and the world to provide a better understanding of the health risk in the study area.

Ni et al. (2010) performed an assessment on health risk caused by ingestion of groundwater in Ya’an City, a city in the middle of China. They found the non-carcinogenic risks of nitrate were between 0.00208 and 1.51250, and the non-carcinogenic risks of fluoride were less than 0.7611. The non-carcinogenic risks caused by nitrate and fluoride are both lower in Ya’an than in the present study area, indicating that the pollution to the groundwater in the study area is much more serious. Su et al. (2013) performed a similar assessment in an agricultural region of Shenyang, a city in northeast China. They found the risks caused by nitrate were less than 1.124 and 1.495 for adults and children, respectively, indicating again that the risks caused by groundwater pollution were much greater in the study area. Similar to the present study, they also found children are at greater risk than adults. Zhang et al. (2014b) assessed the groundwater quality from the perspective of human health risk in Hetao Plain of mid-north China, and they found that 87.9 % of the collected samples were not suitable for drinking from the perspective of human health risk. It is easy to find from the comparison that the study area is among the most risky areas in China regarding groundwater contamination, and urgent action should therefore be taken to reduce the risk and guarantee the safety of drinking water for residents.

There has already been a lot of literatures reporting drinking water nitrate pollution and its adverse effects on human health, but the causal relationship between exposure to nitrates in drinking water and adverse reproductive effects is still not clear (Manassaram et al. 2006). However, the health risk assessment studies can provide useful information for water-quality management to ensure the general safety of drinking water for people. Baba and Tayfur (2011) suggested that too much aluminum in drinking water was one of the contributory causes of Alzheimer, a common illness diagnosed in people over 65 years of age. This information is useful and will force local decision makers to treat the water before it is supplied to residents. Wongsanit et al. (2015) found the risks caused by groundwater nitrate ranged between 0.04 and 4.58 for children in the lower Mae Klong basin, Thailand and between 0.02 and 2.29 for adults. The risks are lower than the present study. Cheong et al. (2012) found that all the groundwater samples in their study belonged to the domain of HI < 1 in which no health hazard can take place by nitrate in groundwater. It is embarrassing that the risks due to groundwater contamination are higher in China than Thailand and South Korea, the two neighboring countries of China. It is mandatory that central and local Chinese governments have to take action to control and eliminate groundwater contamination to guarantee the basic human right.

Strategies Coping with Groundwater Contamination and Health Risk

Shallow groundwater is seriously polluted by TH, NO3 , NO2 , TDS, SO4 2−, and F in the study area and the contaminated groundwater may pose great health risks to adults and children who live here and use groundwater for drinking. To cope with groundwater contamination and to reduce the health risks induced by ingestion of the contaminated groundwater, the following strategies are recommended.

  • As groundwater in the study area has already been contaminated, it is wise to pursue additional water for domestic uses. The field investigation has shown that stream water and spring water originating from the Qinling Mountains are good in quality. It is, therefore, highly recommended that local government establish new water collection and transition networks to supply the stream water and spring water to local residents for drinking purpose. Although the quantities of the stream water and spring water are still not clear, this recommendation can at least ease the water contamination problem in a short term.

  • Industrial effluents from the fertilizer plant should be banned from being discharged to the Shidi River. The industrial effluents contain high concentrations of NO3 , NO2 , and TDS. Without stopping the discharge of effluents from the fertilizer plant, groundwater pollution in this area cannot be effectively controlled.

  • After the industrial effluents are stopped from being discharged into the river, the next necessary step is to clean and recover the river water. The distribution maps of ions have indicated that leakage from the contaminated river is an important responsible factor for groundwater pollution. Therefore, to cure groundwater pollution, cleaning the contaminated river is a necessary step.

  • The Weihe River is also a seriously contaminated river that has negative impacts on groundwater quality nearby. Therefore, cleaning the Weihe River is also recommended. However, cleaning the Weihe River is difficult, because the Weihe River runs across several cities such as Baoji, Xi’an and Weinan and it is receiving pollutants from many outfalls most of which are located out of the study area. Cleaning of the Weihe River needs collaboration from the above mentioned cities. Without collaboration, the cleaning will fail in the end.

  • Human activities such as the disposal of sewage and animal wastes should also be regulated. The wastes should be treated properly in a waste-treatment plant to reduce the adverse impacts on the environment and groundwater.

  • Groundwater-monitoring networks should be established as soon as possible. It is shocking that there still is no official groundwater-monitoring networks in this area, although groundwater has been contaminated to such an alarming degree. The lack of public awareness on groundwater protection and ignorance of groundwater research make this situation even worse.

Conclusions

Groundwater is the main source of water for drinking in arid and semiarid alluvial plains. In the present study, groundwater quality in an alluvial plain impacted intensively by agricultural and industrial activities was assessed for drinking and irrigation purposes. The health risks due to oral ingestion and dermal contact of contaminated groundwater were estimated. The following conclusions were summarized.

  • Shallow groundwater is seriously contaminated by TH, NO3 −, NO2 , TDS, SO4 2−, and F. The distribution of contaminants indicates that the industrial and agricultural activities are significant factors impacting the groundwater quality. Over 60 % of the water samples fall within poor and very poor quality grades, indicating unsuitability for drinking. However, groundwater in the study area is generally suitable for irrigation except some locations with high salinity.

  • Residents in the study area are at high health risk. The total risk from multiple contaminants ranges from 0.17 to 13.73 with a mean of 1.72 for adults, and for children the total risk ranges between 0.32 and 25.60 with an average of 3.20. Children are at higher health risk than adults. Risks posed through oral ingestion contribute greater proportion to the total risk than the dermal exposure pathway, and HQoral is two orders of magnitude higher than HQdermal. Besides, NO3 is the greatest contributor to the health risk, requiring additional attention from policy makers and research scientists.

  • To cope with groundwater contamination and reduce health risks to human beings, finding alternative clean water from the Qinling Mountains is a good choice in short term. However, cleaning contaminated river water, prohibiting industrial effluents, enhancing collaboration between cities, and establishing groundwater-monitoring networks are the only way of combating groundwater contamination in the long run.