Spatial correlation of groundwater pollutants with the vulnerability index in Kano Metropolis, North-western Nigeria

Natural and anthropogenic factors such as drought, desertification, rapid population growth, and agricultural development coupled with pollution from industrial activities and domestic sewage were identified as effective factors leading to an imbalance between water supply and demand as well as the general quality of both surface and groundwater resources in arid and semi-arid regions (Tukur et al. 2020). In the dryland region of West Africa, the interplay of climatic and geological characteristics makes surface water virtually inadequate, and as such groundwater appears to be the only reliable source of fresh water for domestic and agricultural use (Tukur et al. 2018a). As such, the demand for freshwater has led to drastic depletion and deterioration of groundwater in several parts of dryland regions in recent decades such as Israel, Iran and Iraq (Sethy et al. 2016). Although groundwater accounts for only 28.90% of the total freshwater in Nigeria, about 128 million people (85% of the total population) depended on this resource as of 2001 mainly due to the deterioration in the quality and quantity of surface water, insufficient water supply by water authorities, the effect of climate change and above all rapid population growth (Akujieze et al. 2003).

The fact that the Kano region lies in the semi-arid agro-ecological zone where rainfall is often erratic and inadequate in amount and distribution for many uses (Ajon et al. 2014; Umar et al. 2019), exploitation of groundwater especially in the metropolitan area has increased greatly. The exploitation of groundwater, therefore, remains the only option to supplement the ever-increasing demand for domestic water. However, the groundwater is highly vulnerable to pollution from various sources, principally from anthropogenic sources including improper sewage systems, oil spillages, and open dumping of waste, among others. As such, the risk of the degradation of the aquifer may be developed as the competition for the utilization of its resources increases.

Groundwater vulnerability assessment is an empirical method of determining the risk that groundwater in a particular area will become contaminated based on several physical parameters. Aller et al. (1987) were one of the first to use the term groundwater vulnerability to characterize how susceptible aquifers are to contamination. This concept was conceived to raise awareness of the dangers of groundwater contamination and to propose possible measures for the identification and protection of pollution-susceptible areas. Groundwater vulnerability assessments were developed to represent the natural groundwater aquifer characteristics governing the ease with which groundwater may be contaminated (Vrba & Zaporozec 1994). Groundwater vulnerability to pollution is often characterized as a dimensionless parameter with no direct physical interpretation. Numerous numerical models based on a host of related hydrogeological and geographical parameters have been established for estimating groundwater vulnerability, including DRASTIC (Aller et al. 1987), GOD (Foster 1987), EPIK (Doerfliger et al. 1999) and COP (Vías et al. 2005). The DRASTIC vulnerability model characterizes properties of the groundwater-saturated zone to attenuate the extent of surface contaminants to transport within aquifers. In most cases, these models integrate simple qualitative indices that bring together key factors capable of influencing the mobility and migration of key solutes across the groundwater system. Specific to each model, these indices (i.e., parameters) include depth of water table, net recharge, aquifer media, soil media, topography, the impact of the vadose zone, hydraulic conductivity (for DRASTIC), while the DRASTIC model framework is based on a more comprehensive quantification of the hydrological and geological parameters. Risk assessment considers the interaction between the subsurface contaminant load and the aquifer vulnerability at the concerned location (Mimi & Assi 2009). Groundwater pollution and vulnerability studies are the key issues in water resources management, especially in areas with a high level of risk of groundwater pollution such as metropolitan areas (Kano Metropolis inclusive) where high rising population, inadequate surface water availability and the dry climate coupled with pollution from various sources are identified as the influential factors leading to groundwater quality deterioration. Although there are studies on groundwater pollution in the Kano region, they are limited to the only aspect of vulnerability without integrating the quality index approach. For this reason, this paper integrated the groundwater quality index approach with a view to developing a spatial variability of groundwater vulnerability using organic and inorganic pollutants as water pollution indices. Specifically, the study aims at:

• Analysing inorganic pollutants using the water quality index (WQI) method.

• Spatial presentation of organic and inorganic pollutants.

• Correlation analysis between organic and inorganic pollutants with water quality parameters indices using the area under the curve (AUC) and receiver operating characteristics (ROC).

This section provides an overview of the field methodology used to determine the coordinates of each borehole well, to collect borehole water samples and measure corresponding physical, geochemical and hydrocarbon parameters across the study sites. After all coordinate points needed for study site assessment were identified, the next step was calculating the proximity of boreholes within the nearest sources of contaminant as identified from the Kano case study analysis. Buffer Analysis in Arc Geographic Information System (GIS) 10.3 was used to determine the 70 different borehole well locations relative to locations of nearby residential dumpsites, underground storage tanks at petrol stations, automobile garages and depots within the study.

For obtaining groundwater samples from boreholes, a rubber tube with a long rope attached was lowered into the borehole well. The required sample volume was then drawn from the well and deposited into the sample bottle which was marked with the date and name. Sample bottles were then transferred immediately into a cooler container and subsequently taken to the laboratory for analysis (Figure 2).

Groundwater samples were collected from 60 boreholes within the proximal range to 67 petrol-dispensing stations within the Kano Metropolis. The greatest measured distance between a borehole and a petrol station was 4.6 km and the closest was identified at 0.1 km. In Figure 5, the black line identifies the distance to each borehole and its closest petrol station. Petrol station borehole monitoring took place on 11–20 August 2018 and 11–20 March 2020; sampling was usually between 8.00 am and 17.30 pm. Groundwater samples collected at these boreholes were analysed for the hydrocarbon concentrations.

The sampling standard methods prescribed by the American Public Health Association (APHA 2005) were followed carefully for the groundwater samples collection. The groundwater was pumped out for about 10–15 min before taking the actual water samples. The reason was to ensure that the stagnant polluted water in the borehole wells was replaced by the fresh water from the aquifer, thus ensuring collecting a ‘true representative’ of groundwater samples in the surrounding aquifer and not the stagnant polluted water that resides in the well (APHA 2005). Prior to the pumping of water from the borehole wells, the water containers (plastic polyethylene bottles) were washed with nitric acid (HNO₃) and rinsed with distilled water. This was to eliminate contaminants precipitated at the surface of the sampling bottles (New Zealand Water Quality Sampling Part I 1998).

Inorganic pollutants affecting groundwater quality were considered due to the presence of different sources of contamination near borehole wells serving as sources of drinking water for the communities around the Kano region. Water quality is assessed based on parameters characterizing potential issues with taste, colour and odour. Groundwater usually contains appreciable levels of chemical ions such as calcium (Ca), magnesium (Mg), chloride (Cl), nitrogen (N), iron (Fe), etc. due to local geology. These ions were evaluated based on standards of drinking water defined by WHO (2017a) and Standards Organization of Nigeria (2007). Boreholes close to potential sources of organic hydrocarbon contamination were considered for this assessment.

Commonly identified petroleum hydrocarbons are benzene, toluene, ethylbenzene, and xylene, jointly identified as BTEX; all of which are major solvents from petroleum-related sources such as under-storage tanks within petroleum-dispensing stations, automobile shops and storage tanks at depots. The locations of hydrocarbons varied over the sampling intervals. All these parameters were used in assessing the concentration level of organic pollutants in the study area. Samples collected from the selected borehole well sites that were near hydrocarbon sources (e.g., automobile garages, petrol stations and depots) were sent to the Yobe State University Analytical Chemistry Laboratory for gas chromatography-mass spectrometry (GCMS) analysis. Analytical standards containing 50 mg/L each of benzene, ethylbenzene, toluene, xylene (BTEX) was used to first prepare calibration standards obtained from the University of Bath. An approximate proportion of mass relative to the volume used was filtered with a nylon membrane. 50 mL of groundwater sample was then shaken vigorously for 30 min with an equal proportion of the BTEX solution. The resultant aqueous layer was further separated and extracted again using an additional 25 mL of BTEX solution. 10 mL of extracted BTEX solution was then placed in a 30-mL vial and allowed to separate for 1 min. The organic layer and combined extracts obtained were then stored at 4 °C. A centrifuge was used to spin the extracts for 10 min at a speed of 4,000 rpm. Further analysis was performed using an Agilent GCMS 7890b at temperatures of 30 °C–50 °C. The supernatant 1.0 UL was injected into the GCMS; the GCMS oven was set at an initial temperature of 40 °C for 1 min and then increased at a rate of 2 °C/min to 65 °C for 3 min and then at a rate of 35 °C/min to 190 °C for 3 min. Following GCMS analysis completion, sample components were identified by spectra comparison within a corresponding BTEX library.

where WQI indicates the water quality index; W_i is the unit weight of the parameter; Q_i is the water quality rating of each parameter; n is the number of parameters assessed.

Often numerical model such as DRASTIC requires data for developing accurate predictions of needed parameters using the GIS interface. It is critical to include accurate estimations of intrinsic parameters (e.g., depth of groundwater, net recharge and aquifer media) to improve the reliability of these models as these parameters dictate the migration of contaminants at the surface of most aquifers. For example, zones with high fracture aquifers are more vulnerable to contamination because they serve as pathways for pollutant migration.

The statistical significances were determined using AUC to measure the dimensional area of organic and inorganic pollutants within an area, while the ROC indicates the probability of occurrences. The plotting on the y-axis is the concentration of the pollutants and the x-axis is the groundwater quality parameters values that will produce the two curves (Carter et al. 2016) for eight of the key physiochemical parameters including nitrate (NO₃), magnesium (Mg), total dissolved solids (TDS) comprises of magnesium, potassium and sodium measured in parts per million(ppm), which ranges from 50 to 1,000 ppm, calcium (Ca), chloride (Cl), total alkalinity, pH, hardness and hydrocarbons parameters of toluene, benzene, ethylbenzene and xylene. Water quality parameters have been taken as a reference point for preparing ROC. The influential parameters are depth of water, net recharge, and aquifer media. For groundwater vulnerability analysis. ROC curve is a popular validation method, which is widely used on groundwater vulnerability terms to contamination (Singh et al. 2015). The AUC was calculated using ROC and judged based on the acceptability of the models such as DRASTIC. The AUC is a measure of the likelihood of the proper classification of a phenomenon by the model obtained. The AUC nearer to 0.5 indicates a good prediction within the pollutants and groundwater quality parameters to random sampling of the groundwater parameters, and AUC nearer to 1 indicates a good prediction (Arabameri et al. 2020) of the organic and inorganic pollutants.

The physiochemical parameters of groundwater are considered the most important principles in identifying and defining the quality of water (Mohamed 2005). Statistical analysis contributed immensely to defining the monitored physiochemical parameters relative to standard guidelines and recommended limits by WHO and the Nigerian government (Standards Organization of Nigeria 2007; WHO 2017b) for this study. The average concentration of the parameters reveal that TDS, 155.7 mg/L; total alkalinity, 2.96 mg/L; hardness, 127.1 mg/L; K, 37.7 mg/L; Na, 8.9 mg/L; Mg, 63 mg/L; Ca, 65.4 mg/L; NO₃, 29.5 mg/L; Cl, 24.5 mg/L; and pH level was 4.8. Relative to the values presented in Table 4, all the average values are within the standard for drinking water.

The spatial distribution for nitrate is depicted in Figure 8(a). Nitrate in groundwater is believed to come from different sources, including waste discharges and indiscriminate use of primary artificial fertilizers. Human activities associated with specific land use patterns such as dumpsites may be highly efficient for anthropogenic discharge of nitrate pollutants into the subsurface (Ali & Young 2014). Results obtained for the concentration of nitrate in borehole samples yield average an value of 29.5 mg/L which is within the standard for drinking water WHO (2017b) and (Standards Organization of Nigeria 2007) of 45 mg/L. Similar results for nitrates were obtained at 8.6 mg/L for Gwale LGA (Idris et al. 2018) . Groundwater samples expected to be influenced by leachates from a dumpsite around one of the local study areas called Tarauni were found to have nitrate concentration ranges of 1.1–8.7 mg/L, still within the permissible limits (Ismaila et al. 2020). TDS (which contains potassium and sodium) evaluates the number of dissolved salts in a sample. These salts often result in inferior palatability and may induce an unfavourable physiological reaction in the transient consumer. High TDS concentrations may also be responsible for increasing hardness, turbidity, odour, taste, colour, and alkalinity. The estimated average value for TDS across the study location is 155.7 mg/L. These values were within the limit of water quality standards of 500 mg/L as defined by (WHO 2017b) and Nigerian water quality standards in Table 1.

Magnesium is generally less commonly found in appreciable quantities within the Kano groundwater system even though evidence for anomalous magnesium occurrences in groundwater is marked by the presence of an undesirable taste. Physical evidence of magnesium occurrences in groundwater is often defined by the presence of black specks that settle out of the water. The groundwater sample analyses carried out indicate a maximum concentration of manganese in groundwater for the study up to 52.7 mg/L was just over the 50 mg/L limit (WHO 2017b) and (Standards Organization of Nigeria (2007),Figure 8(b). Concentrations for a majority of the Kano study region remained at relatively low levels, with an average of 33.6 mg/L well within the regulatory limit (Amoo et al. 2018). Alkaline compounds that are present in water, like hydroxides and carbonates, eliminate H⁺ ions from the water, which lowers the acidity, of the water and results in a higher pH.

As stated by WHO (2017a) and Nigerian Water Quality Standards (NWQS) allowable limit for drinking water is 200 mg/L. In this study, alkalinity values ranged from 75 mg/L values obtained from boreholes. Concentrations for a majority of the Kano study region remained within the regulatory limit.

Calcium is generally considered a non-essential element with wide distribution across the earth. Its occurrence in groundwater is generally linked to local geology and/or to anthropogenic activities. As shown in Figure 9(e), maximum concentrations of 49.8 mg/L were observed within the Kano study location; observed values are lower than the regulatory limit of 75 mg/L (Table 4). High amounts of calcium in groundwater can be carcinogenic and problematic concentrations on the order of 220 mg/L have been observed in other studies within the Gwale area of Kano (Idris et al. 2018). As shown in Figure 4(f), the concentration of chloride in the Kano groundwater ranged from 0 to 10.9 mg/L; concentrations were well below the limit of 250 mg/L (Table 4) at all borehole sampling locations. Excess concentration of chloride in groundwater can result in a salty taste to water; in this study the limit was 292 mg/L, areas in Kano. Hardness in groundwater measures the concentration of dissolved calcium and magnesium salts. In Kano, the physical evidence of hardwater is often identified by an increased chalky taste; Kano Metropolis groundwater hardness ranges from 0 to 431 mg/L (Figure 9). The Nigerian water standard threshold value is set to 300 mg/L so observed hardness levels were above this limit considerably in some regions (Figure 9). Increased concentrations of hardness have been measured as high as 590 mg/L in several studies focused on the Kano Metropolis (Hamza et al. 2017; Idris et al. 2018). Observed values of pH ranged from 6.7 to 7.7 (Figure 9); similar values were obtained from groundwater quality assessments of the basement complex within Kano (Adamu et al., 2013). These values fall within the recommended limit of 8.5 by the WHO (Table 4). From the 42 borehole wells sampled close to dumpsites, 10 boreholes illustrate high concentration of hardness which is represented in Figure 9 with red colour across the study location.

Sodium salts (e.g., sodium chloride) are found in drinking water. Although concentrations of sodium in potable water are typically less than 20 mg/L, they can greatly exceed this in some localities. Concentrations of more than 200 mg/L may give rise to an unacceptable taste. The water analysis indicates a sodium concentration in groundwater of the study area ranging from 125.6 to 0.70 mg/L, which is the lowest concentration. The samples have values which are far below those of the Drinking Water Standards (WHO 2004), as well as Nigerian Standards for Drinking Water (Standards Organization of Nigeria 2007), with recommended and maximum permissible limits of 150 and 200 mg/L, respectively.

It can occur naturally in minerals and from soils. High levels in surface water, especially in areas where there are agricultural activities as indicative of the introduction of potassium due to the application of fertilizers. The concentration of potassium in groundwater in the study area ranges from 97 to 5.94 mg/L. Most of the samples analysed are above the maximum permissible limit (15 mg/L) with respect to the WHO standard.

Based on these spatial representations of the eight key water quality parameters, corresponding WQI values were estimated for the borehole monitoring locations and a spatial representation was created (Figure 11), following Equations (1)–(3). As part of the WQI calculations, weightings for each parameter characterize the relative importance of the parameter (e.g., pH, hardness) to the overall quality of water for drinking water purposes. According to the five quantile classes obtained from the final WQI map in Figure 5, 45% of the study area is estimated to fall under good classification with 25% excellent, 15% poor, 5% very poor and 10% unsuitable for drinking water. These results indicate that substantial coverage of the Kano study area can be classified as good to excellent quality per the WQI ranking.

Organic hydrocarbon BTEX compounds were analysed within the boreholes close to known hydrocarbon sources such as under-storage tanks at petrol stations, automobile shops and the primary depot within the study area. Chemical analysis was done following the methodology to determine BTEX concentrations for the borehole samples and compare these to established guidelines by WHO (2017b) and Standards Organization of Nigeria (2007). Aromatic BTEX hydrocarbons tend to be the most water-soluble component of crude oil and other petroleum compounds. Benzene is the most soluble of the BTEX compounds; it is thus the primary groundwater contaminant of concern at petroleum release sites because of its high toxicity and mobility as compared to the other petroleum hydrocarbons. BTEX compounds are considered the most volatile of the organic compounds, also known as VOC. Hydrocarbons are generally colourless with mild odours; when consumed they cause illness that can sometimes lead to death within communities. Studies on BTEX contamination of groundwater have been performed globally, with concentrations of 24–28 mg/L obtained for Seoul, Korea (Yu et al. 2017), 121 mg/L within Italy, (Pietroletti et al. 2010) 15 mg/L in Thailand (Yeesang & Cheirsilp 2011) and 45 mg/L in Australia (Abraham et al. 2018).

The datasets obtained from the chemical analyses as well as the environmental anthropogenic factors have been quantified and assessed. Correlation analysis was applied to these concentrations and supporting environmental data using the statistical software package (SPSS) to estimate correlation coefficient R. The method was conducted to test whether there was a significant correlation between the different indicators, based on a quasi measure of a two-variable relationship (Gaál et al. 2015). Chemical parameters were compared with groundwater quality index parameters to determine whether there is a statistically significant correlation and what can be deduced from these results. There was a correlation when all chemical parameters were compared with general water quality guidelines (WHO 2017a). The measured AUC under ROC specifically defined the appropriateness of all models as the AUC of each model is nearer to 1% or above 65.6%. A number of concentration ranges were found to have a positive trend with WQI values and inorganic pollutants after normalization (Table 5 and Figure 16).

This study performed correlation analysis between the groundwater vulnerability index with organic and inorganic pollutants and developed a vulnerability risk map in Kano Metropolis, North-western Nigeria. A combination of multivariate statistics, geospatial techniques, and the WQI were used to provide a comprehensive understanding of groundwater pollution risk in Kano Metropolis, North-western Nigeria. The following conclusions were reached. Most of the higher concentrations for all parameters were attributed to anthropogenic activities such as dumpsites, based on the respective borehole locations around the study area. Quantification of the four major BTEX hydrocarbons analysed yielded maximum values of toluene at 47.5 mg/L, benzene at 24.7 mg/L, and relatively lower concentrations for ethylbenzene at 6.8 mg/L, and xylene at 9.1 mg/L; these were monitored at borehole locations in specific proximity to known point sources of hydrocarbon pollution including petrol-dispensing stations, automobile shops, and depots. The corresponding WQI quantification of the borehole groundwater samples revealed that 45% of the groundwater supplies may be classified as sufficiently good for drinking. The measured AUC under ROC revealed positive correlation between groundwater vulnerability with organic (63.7%) and inorganic pollutants (65.6%). Mapping based on hydrocarbon concentrations revealed a high level of pollution risk especially in the core centre of the metropolitan Kano. This suggests that groundwater in the area is highly vulnerable to hydrocarbon pollution. Close monitoring of groundwater quality is recommended in the study area for proper management.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.