The investigation collected 50 random water samples from wells and bore holes in the five wards. In the meantime, the Water Quality Index (WQI) in this region was assessed using a novel machine learning model. In this sphere of science, the Emotional Artificial Neural Network (EANN) was used as an innovative technique. The training dataset comprised 80% of the available data, while the remaining 20% was used to assess the performance of the network. The laboratory analysis revealed that the levels of magnesium (0.581 mg/L), mercury (0.0143 mg/L), iron (0.82 mg/L), lead (0.69 mg/L), calcium (2.03 mg/L), and total dissolved solid (105 mg/L) in the water sample were quite high and exceeded the maximum permissible limits established by the National Standard Water Quality (NSWQ) and Water Quality Association (WQA). Except for magnesium, mercury, iron, and lead, all physicochemical parameters are below the utmost permissible limit. Results showed that hydrogeological effects and anthropogenic activities, such as waste management and land use, impact groundwater pollution in the Chikun Local Government Area of Kaduna State up to 60 m deep. The results of the EANN showed that R2 index and normalized root mean square error (RMSENormalized) values for the training and test stages are 0.89 and 0.18, and 0.83 and 0.23, respectively.

  • Groundwater quality is examined in the Chikun Local Government Area of Kaduna State and how it can be used for water supply to improve water management.

  • Machine learning outperformed the WQI in water quality estimation.

  • The EANN satisfactory performance when applied to studies characterized by limited data availability.

Water is a one-of-a-kind resource with no substitute and a fixed quantity, though its quantity and quality vary across space and time. Population growth has continued, particularly in the developing world, as has the demand for water for various purposes (Javidan et al. 2022). This increase in demand has resulted in water scarcity in many parts of the world, and water scarcity remains an endemic problem anywhere (Kubiak et al. 2021; Konin et al. 2023). Aside from domestic water needs, other human activities such as fishing and farming rely on water (Lapworth et al. 2017; Maroeto & Santoso 2023).
Figure 1

Geology of the Chikun Local Government. Source: GIS Lab. Department of Environmental Management, Kaduna State University, (2019).

Figure 1

Geology of the Chikun Local Government. Source: GIS Lab. Department of Environmental Management, Kaduna State University, (2019).

Close modal

The formation of the Earth and the regulation of water and climate are significantly influenced by this factor (Ualiyeva et al. 2023). The demand for freshwater has experienced a substantial increase as a result of the rapid expansion of the global population and by the process of industrialization. Hence, the scarcity of potable water has emerged as a prominent obstacle in the preceding century. The preservation of water resources, particularly in arid and semi-arid regions, is of significant significance due to their limited availability in terms of both quality and quantity. In recent years, there has been a significant surge in the utilization of underground water as humans have sought to augment their agricultural lands, resulting in a corresponding decline in the utilization of surface water resources (Ashraf et al. 2023; Heyi et al. 2023).

Groundwater has recently emerged as one of the most valuable natural resources in many countries (Alfaleh et al. 2023). Water is generally of high quality and an essential natural resource in its native state because it is naturally purified as it slowly percolates through the soil (Ben Khedher et al. 2023). Groundwater has many advantages over surface water, including higher quality, better protection from surface contaminants, less susceptibility to drought, and much more even distribution over large areas. Groundwater is the sole source of water supply in some countries, such as Denmark, Malta, and Saudi Arabia, while it is the most significant component of total water resources in others. Groundwater, for example, accounts for 95% of total water resources in Tunisia, 83% in Belgium, and 75% in the Netherlands, Germany, and Morocco (Dhaka & Bhaskar 2017).

Utilizing a Water Quality Index (WQI) model enables the transformation of extensive water quality data into a single numerical representation known as the index score. The WQI model is comprised of five phases of development, including the selection of parameters, generation of sub-index functions, establishment of parameter weights, aggregation of sub-index values, and determination of classification schemes (Gradilla-Hernández et al. 2020).

Nigeria has seen an increase in concern in recent years about the depletion of groundwater resources and water contamination as a result of rapid demographic changes (Komolafe 2014). These changes have occurred concurrently with the development of human settlements that lack adequate water supply and sanitation infrastructure. This is especially true in peri-urban areas like Chikun Local Government Area, the study's focal point, which includes the country's major metropolitan towns (Vivan et al. 2021). Despite the Kaduna State government's efforts to ensure clean water for the local population, the Chikun Local Government Area has made little progress in establishing a reliable water supply network. The current issue is the inability to meet the entire region's water demand due to its escalation for various purposes. Furthermore, when confronted with unfavorable and unpredictable climatic conditions, the vulnerability of relying heavily on a solitary water source becomes clear. Aside from the observation that groundwater levels in various regions of the Chikun Local Government Area are rapidly declining and deteriorating in quality, this phenomenon can be attributed to increased groundwater extraction and the indiscriminate proliferation of wells and boreholes for domestic water consumption. This trend can be attributed to the country's rapid population growth and residents' changing lifestyles (Samira et al. 2015).

The first empirical evidence of groundwater pollution in the Chikun Local Government Area was presented in a World Bank-sponsored study in 1988. This study concentrated on the pollution of surface and groundwater in several wards including Mararaban Rido, Kakau, Nassarawa, and Sabon Tasha, as well as the Kaduna River. The findings revealed that, of the sampling sites examined, the location where the River Romi meets the Kaduna River had the highest pollution load. This can be attributed to the discharge of refinery effluents into the Romi River. It is worth noting that River Romi contributes significantly to groundwater replenishment in various areas of the Chikun Local Government Area. As a result, there has been a dearth of scholarly investigations into groundwater contamination in the Chikun Local Government Area. Samira et al. (2015) conducted a study on the assessment of water quality in hand dug wells in Bayan Dutse, Narayi, within the Chikun Local Government Area, which was a notable culmination of scholarly investigations in this field. The study focused on the analysis of physicochemical parameters derived from collected samples. The study discovered that hand dug wells are vulnerable to water contamination from sewage system runoff, owing to their geographical location. In a study conducted by Ishaku (2011), an assessment was made regarding the groundwater quality in the Jimeta-Yola region located in northeastern Nigeria. The findings revealed that the WQI exhibited a lower value during the dry season as opposed to the wet season, indicating a comparatively superior water quality within this particular area.

The presence of uncertainty in the WQI might be attributed to the variability in parameter choices and their respective weightings, which may differ across various WQI models. Various models may possess distinct parameter sets and weightings, which are contingent upon their particular aims and geographical contexts. Consequently, the interpretation and comparison of WQI results across various models and locales might provide challenges (Uddin et al. 2021). According to Gupta & Gupta (2021), the utilization of multi-criteria decision-making methodologies is recommended for incorporating into WQI models. Multi-criteria decision-making methodologies enable the concurrent evaluation of numerous factors and criteria, hence mitigating the issue of eclipsing. Yilma et al. (2018) employed an artificial neural network (ANN) methodology to forecast the WQI in Ethiopia's highly contaminated rivers, utilizing data collected from 27 sample locations. Furthermore, in order to reduce the time and effort required for repeated calculation of the WQI, a successful strategy involves the utilization of the ANN for WQI determination. In a separate study, Ismael et al. (2021) utilized the ANN to forecast the WQI in the Red sea region of Sudan. Specifically, they employed the feed forward back propagation technique inside the ANN framework to optimize their predictions. ANNs have the capability to discern the correlation between input and output parameters without the need for explicit physical deliberation. The ANN model demonstrated a much higher level of accuracy in predicting the WQI during the modeling process. Uddin et al. (2023) present a comprehensive and rigorous approach to evaluate the uncertainties associated with the WQI model. A total of eight WQI models are taken into consideration. The use of the Monte Carlo simulation approach was employed to quantify model uncertainty, while the Gaussian Process Regression algorithm was utilized to forecast uncertainties in the WQI models at each sample site.

The existing body of literature on groundwater prospecting in the specified study area indicates a lack of previous research on the application of the WQI methodology for evaluating groundwater quality in the Chikun Local Government Area of Kaduna State. The application of the WQI approach in evaluating the groundwater status in the Chikun Local Government Area exhibits potential in tackling the issue of water shortage resulting from the exploitation of unreliable water sources, as well as in limiting health risks that might potentially result in deaths. The present study employed the novel Emotional Artificial Neural Network (EANN) model to assess the water quality in the designated area, thereby capitalizing on its inherent benefits. Subsequently, the outcomes derived from employing this methodology were juxtaposed with those acquired through the utilization of the WQI approach. Hence, it is imperative to conduct an evaluation of the groundwater quality in the Chikun Local Government Area in order to safeguard the integrity of the groundwater sources, which are currently susceptible to pollution.

The Chikun Local Government Area lies geographically between Latitude 10° N and 10° 50″ N and Longitude 6° 4″ E and 7° 5″ E (Figure 1). It is located on the Southern part of Kaduna State and share common boundaries with Kaduna North Local Government and Igabi in the North. In the Southwestern part, it shares border with Niger State and in the East and with Kajuru and Kachia Local Government Area. At present the Chikun Local Government Area has Kujama as its administrative headquarters, Gonin Gora, Narayi, Nassarawa, Trikania, Sabon Tasha, Ungwar Romi, Ungwar Sunday, Ungwar Yelwa, Karatudu and part of Barnawa as it component area covering a total land size of 4,801 km2 (Njoku & Akpan 2013; Hassan Ibrahim et al. 2017).

The assessment of groundwater quality for Kaduna State's Chikun Local Government Area was designed and carried out in the following stages: pre-field preparation, which included the creation of maps for the study area; reconnaissance study of the area; sampling technique used to collect water samples from hand-dug wells and boreholes throughout the study area; and data collection and analysis. Boreholes and hand-dug wells are the two most common types of groundwater abstraction structures in the Chikun Local Government Area. Groundwater samples from the five selected divisions were collected throughout the peak of the dry and wet seasons in 2019 and 2020, as shown in Tables 13. Groundwater samples were collected between 7 a.m. and 1 p.m. during the research period. To maintain the composition of the water samples, about two liters of water sample from each source (wells and boreholes) was collected in separate 2-L plastic cans and carried to the laboratory for examination. Prioritization-based parameter analysis was performed. The WQI was analyzed using 16 water quality parameters covering four hazard classifications (salinity hazard, permeability/infiltration hazard, specific ion toxicity hazard, and miscellaneous hazard). The closeness of a well or borehole to a previously selected well or borehole in the area, as well as the owner's willingness to make the well or borehole available for study, influenced the selection of a well or borehole. Water samples were collected from various sites within the study region using new 2-L acid-washed plastic canisters in accordance with the random sampling procedure instructions.

Table 1

Selected wards in the study area

S/NoWardsSelected wards
Rural wards 
Chikun Gwagwada 
Gwagwada Kunai 
Kakau  
Kunai  
Kuriga  
Urban wards 
Kujama Mararaban Rido 
Mararaban Rido Nassarawa 
Narayi Sabon Tasha 
Nassarawa  
10 Sabon Gari  
11 Sabon Tasha  
12 Yelwa  
S/NoWardsSelected wards
Rural wards 
Chikun Gwagwada 
Gwagwada Kunai 
Kakau  
Kunai  
Kuriga  
Urban wards 
Kujama Mararaban Rido 
Mararaban Rido Nassarawa 
Narayi Sabon Tasha 
Nassarawa  
10 Sabon Gari  
11 Sabon Tasha  
12 Yelwa  

Source: Field Survey, 2019.

Table 2

Wells and boreholes sampled in the study area

S/NoSettlementWellsBoreholes
Gwagwada 
Kunai 
Mararaban Rido 
Nassarawa 
Sabon-Tasha 
S/NoSettlementWellsBoreholes
Gwagwada 
Kunai 
Mararaban Rido 
Nassarawa 
Sabon-Tasha 

Source: Field Survey, 2019.

Table 3

Details of sampling locations in the Chikun Local Government Area

Sample area (Ward)Sample codeCoordinates
Elevation (m)Depth (m)
Lat. (N)Long. (E)
Gwagwada GD(W1) 10° 21′55.4″ 07°10′56.9″ 590.3 9.41 
 GD(W2) 10° 19′58.2″ 07°12′54.4″ 594.4 10.38 
 GD(W3) 10° 17′58.3″ 07°11′01.7″ 591.9 10.74 
 GD(W4) 10° 18′59.1″ 07°13′10.5″ 592.5 11.96 
 GD(W5) 10° 15′03.6″ 07°16′19.7″ 597.1 11.13 
 GD(BH1) 10° 11′53.8″ 07°18′57.6″ 591.6  
 GD(BH2) 10° 10′56.3″ 07°11′54.7″ 601.3  
 GD(BH3) 10° 08′56.9″ 07°15′01.0″ 589.6  
 GD(BH4) 10° 07′00.2″ 07°11′12.7″ 594.8  
 GD(BH5) 10° 05′15.7″ 07°10′23.9″ 563.7  
Kunai KN(W1) 10° 22′31.8″ 06°53′33.5″ 629.1 12.33 
 KN(W2) 10° 25′44.1″ 06°55′22.4″ 631.6 12.78 
 KN(W3) 10° 29′58.3″ 06°59′55.8″ 633.1 11.36 
 KN(W4) 10° 28′55.6″ 07°01′56.3″ 622.4 12.11 
 KN(W5) 10° 30′73.2″ 07°02′12.9″ 624.6 10.95 
 KN(BH1) 10° 29′31.6″ 07°02′36.0″ 621.0  
 KN(BH2) 10° 31′37.8″ 07°05′42.9″ 627.0  
 KN(BH3) 10° 33′33.4″ 07°06′34.6″ 630.9  
 KN(BH4) 10° 35′01.1″ 07°07′53.8″ 627.4  
 KN(BH5) 10° 34′24.6″ 07°08′24.5″ 622.1  
Mararaban Rido MR(W1) 10° 25′42.0″ 07°31′32.2″ 698.6 11.36 
 MR(W2) 10° 26′02.4″ 07°31′42.9″ 639.1 12.67 
 MR(W3) 10° 26′08.1″ 07°31′52.5″ 648.7 13.04 
 MR(W4) 10° 25′49.9″ 07°31′34.4″ 663.4 9.64 
 MR(W5) 10° 25′52.7″ 07°31′37.5″ 642.2 10.12 
 MR(BH1) 10° 25′34.2″ 07°31′02.4″ 658.3  
 MR(BH2) 10° 25′23.9″ 07°31′56.2″ 650.6  
 MR(BH3) 10° 26′06.4″ 07°31′50.7″ 642.4  
 MR(BH4) 10° 25′54.3″ 07°31′40.7″ 641.2  
 MR(BH5) 10° 25′30.1″ 07°30′57.3″ 632.3  
Nassarawa NS(W1) 10° 26′57.5″ 07°11′29.4″ 621.8 3.41 
 NS(W2) 10° 26′57.5″ 07°11′34.2″ 614.3 3.94 
 NS(W3) 10° 26′51.9″ 07°13′13.8″ 623.1 6.49 
 NS(W4) 10° 27′44.3″ 07°12′06.2″ 609.8 7.83 
 NS(W5) 10° 27′37.9″ 07°14′11.4″ 638.4 6.98 
 NS(BH1) 10° 28′55.0″ 07°14′52.6″ 633.7  
 NS(BH2) 10° 28′31.4″ 07°13′06.2″ 628.5  
 NS(BH3) 10° 28′40.9″ 07°12′33.3″ 622.1  
 NS(BH4) 10° 26′05.2″ 07°11′51.7″ 631.9  
 NS(BH5) 10° 29′41.8″ 07°11′′01.6″ 629.3  
Sabon Tasha ST(W1) 10° 26′58.5″ 07°27′48.4″ 611.6 11.31 
 ST(W2) 10° 26′54.4″ 07°27′43.9″ 609.8 12.01 
 ST(W3) 10° 27′01.7″ 07°27′20.1″ 611.5 10.12 
 ST(W4) 10° 26′33.9″ 07°27′05.3″ 610.7 11.78 
 ST(W5) 10° 26′11.7″ 07°27′21.8″ 611.9 10.93 
 ST(BH1) 10° 26′56.5″ 07°27′56.9″ 609.5  
 ST(BH2) 10° 26′05.2″ 07°27′44.8″ 610.2  
 ST(BH3) 10° 26′37.1″ 07°27′06.3″ 611.9  
 ST(BH4) 10° 26′55.6″ 07°27′58.0″ 609.4  
 ST(BH5) 10° 27′01.4″ 07°27′11.5″ 607.1  
Sample area (Ward)Sample codeCoordinates
Elevation (m)Depth (m)
Lat. (N)Long. (E)
Gwagwada GD(W1) 10° 21′55.4″ 07°10′56.9″ 590.3 9.41 
 GD(W2) 10° 19′58.2″ 07°12′54.4″ 594.4 10.38 
 GD(W3) 10° 17′58.3″ 07°11′01.7″ 591.9 10.74 
 GD(W4) 10° 18′59.1″ 07°13′10.5″ 592.5 11.96 
 GD(W5) 10° 15′03.6″ 07°16′19.7″ 597.1 11.13 
 GD(BH1) 10° 11′53.8″ 07°18′57.6″ 591.6  
 GD(BH2) 10° 10′56.3″ 07°11′54.7″ 601.3  
 GD(BH3) 10° 08′56.9″ 07°15′01.0″ 589.6  
 GD(BH4) 10° 07′00.2″ 07°11′12.7″ 594.8  
 GD(BH5) 10° 05′15.7″ 07°10′23.9″ 563.7  
Kunai KN(W1) 10° 22′31.8″ 06°53′33.5″ 629.1 12.33 
 KN(W2) 10° 25′44.1″ 06°55′22.4″ 631.6 12.78 
 KN(W3) 10° 29′58.3″ 06°59′55.8″ 633.1 11.36 
 KN(W4) 10° 28′55.6″ 07°01′56.3″ 622.4 12.11 
 KN(W5) 10° 30′73.2″ 07°02′12.9″ 624.6 10.95 
 KN(BH1) 10° 29′31.6″ 07°02′36.0″ 621.0  
 KN(BH2) 10° 31′37.8″ 07°05′42.9″ 627.0  
 KN(BH3) 10° 33′33.4″ 07°06′34.6″ 630.9  
 KN(BH4) 10° 35′01.1″ 07°07′53.8″ 627.4  
 KN(BH5) 10° 34′24.6″ 07°08′24.5″ 622.1  
Mararaban Rido MR(W1) 10° 25′42.0″ 07°31′32.2″ 698.6 11.36 
 MR(W2) 10° 26′02.4″ 07°31′42.9″ 639.1 12.67 
 MR(W3) 10° 26′08.1″ 07°31′52.5″ 648.7 13.04 
 MR(W4) 10° 25′49.9″ 07°31′34.4″ 663.4 9.64 
 MR(W5) 10° 25′52.7″ 07°31′37.5″ 642.2 10.12 
 MR(BH1) 10° 25′34.2″ 07°31′02.4″ 658.3  
 MR(BH2) 10° 25′23.9″ 07°31′56.2″ 650.6  
 MR(BH3) 10° 26′06.4″ 07°31′50.7″ 642.4  
 MR(BH4) 10° 25′54.3″ 07°31′40.7″ 641.2  
 MR(BH5) 10° 25′30.1″ 07°30′57.3″ 632.3  
Nassarawa NS(W1) 10° 26′57.5″ 07°11′29.4″ 621.8 3.41 
 NS(W2) 10° 26′57.5″ 07°11′34.2″ 614.3 3.94 
 NS(W3) 10° 26′51.9″ 07°13′13.8″ 623.1 6.49 
 NS(W4) 10° 27′44.3″ 07°12′06.2″ 609.8 7.83 
 NS(W5) 10° 27′37.9″ 07°14′11.4″ 638.4 6.98 
 NS(BH1) 10° 28′55.0″ 07°14′52.6″ 633.7  
 NS(BH2) 10° 28′31.4″ 07°13′06.2″ 628.5  
 NS(BH3) 10° 28′40.9″ 07°12′33.3″ 622.1  
 NS(BH4) 10° 26′05.2″ 07°11′51.7″ 631.9  
 NS(BH5) 10° 29′41.8″ 07°11′′01.6″ 629.3  
Sabon Tasha ST(W1) 10° 26′58.5″ 07°27′48.4″ 611.6 11.31 
 ST(W2) 10° 26′54.4″ 07°27′43.9″ 609.8 12.01 
 ST(W3) 10° 27′01.7″ 07°27′20.1″ 611.5 10.12 
 ST(W4) 10° 26′33.9″ 07°27′05.3″ 610.7 11.78 
 ST(W5) 10° 26′11.7″ 07°27′21.8″ 611.9 10.93 
 ST(BH1) 10° 26′56.5″ 07°27′56.9″ 609.5  
 ST(BH2) 10° 26′05.2″ 07°27′44.8″ 610.2  
 ST(BH3) 10° 26′37.1″ 07°27′06.3″ 611.9  
 ST(BH4) 10° 26′55.6″ 07°27′58.0″ 609.4  
 ST(BH5) 10° 27′01.4″ 07°27′11.5″ 607.1  

Source: Field Survey, 2019.

Table 4

Mean values of pH concentration in groundwater of the Chikun Local Government Area (2019–2020)

WardWell
Borehole
WHO (MP)6.6–8.5MeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  7.073 0.29732 0.066484 0.04203 6.825 0.097899 0.021891 0.014344 
Kunai  6.9485 0.142433 0.03184 0.020498 6.426 0.095278 0.021305 0.014827 
Mararaban Rido  6.829 0.118495 0.026496 0.017352 6.852 0.035184 0.007867 0.005135 
Nassarawa  6.476 0.176707 0.039513 0.027286 6.564 0.263986 0.059029 0.040217 
Sabon Tasha  6.8255 0.172458 0.03856 0.025267 6.43 0.113025 0.025273 0.016457 
WardWell
Borehole
WHO (MP)6.6–8.5MeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  7.073 0.29732 0.066484 0.04203 6.825 0.097899 0.021891 0.014344 
Kunai  6.9485 0.142433 0.03184 0.020498 6.426 0.095278 0.021305 0.014827 
Mararaban Rido  6.829 0.118495 0.026496 0.017352 6.852 0.035184 0.007867 0.005135 
Nassarawa  6.476 0.176707 0.039513 0.027286 6.564 0.263986 0.059029 0.040217 
Sabon Tasha  6.8255 0.172458 0.03856 0.025267 6.43 0.113025 0.025273 0.016457 

Source: Field Survey and Laboratory Analysis (2019 and 2020).

Table 5

Mean values of turbidity in groundwater of the Chikun Local Government Area (2019–2020)

WardWell
Borehole
WHO (MPL)50 (NTU) mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  48.35 6.351751 1.420295 0.13137 34.99 12.34951 2.761435 0.352944 
Kunai  46.5 1.877849 0.4199 0.040384 16.8 0.756724 0.169209 0.014827 
Mararaban Rido  49.35 2.796144 0.625237 0.056659 22.65 0.873951 0.195421 0.038636 
Nassarawa  41.7 4.079474 0.912198 0.097829 15.63 8.14849 1.822059 0.521337 
Sabon Tasha  48.1 2.125039 0.475173 0.04418 48.48 2.444241 0.546549 0.050418 
WardWell
Borehole
WHO (MPL)50 (NTU) mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  48.35 6.351751 1.420295 0.13137 34.99 12.34951 2.761435 0.352944 
Kunai  46.5 1.877849 0.4199 0.040384 16.8 0.756724 0.169209 0.014827 
Mararaban Rido  49.35 2.796144 0.625237 0.056659 22.65 0.873951 0.195421 0.038636 
Nassarawa  41.7 4.079474 0.912198 0.097829 15.63 8.14849 1.822059 0.521337 
Sabon Tasha  48.1 2.125039 0.475173 0.04418 48.48 2.444241 0.546549 0.050418 

Source: Field Survey and Laboratory Analysis (2019 and 2020).

Table 6

Mean values of electrical conductivity in groundwater in the Chikun Local Government Area (2019–2020)

WardWell
Borehole
WHO (MPL)15,000 (μS/cm)MeanStandard DeviationStandard Error MeanCoefficient of VariationMeanStandard DeviationStandard Error MeanCoefficient of Variation
Gwagwada  639.165 98.74404 22.07984 0.154489 690.348 4.312581 0.965322 0.006247 
Kunai  554.693 75.69895 16.9268 0.13647 596.967 4.685292 1.047663 0.007848 
Mararaban Rido  570.144 12.68637 28.36759 0.222512 688.179 10.81259 2.417769 0.015712 
Nassarawa  664.402 43.93596 9.82438 0.066129 572.345 6.065617 1.356313 0.010598 
Sabon Tasha  694.801 8.289031 1.853484 0.01193 697.549 10.26037 2.294289 0.014709 
WardWell
Borehole
WHO (MPL)15,000 (μS/cm)MeanStandard DeviationStandard Error MeanCoefficient of VariationMeanStandard DeviationStandard Error MeanCoefficient of Variation
Gwagwada  639.165 98.74404 22.07984 0.154489 690.348 4.312581 0.965322 0.006247 
Kunai  554.693 75.69895 16.9268 0.13647 596.967 4.685292 1.047663 0.007848 
Mararaban Rido  570.144 12.68637 28.36759 0.222512 688.179 10.81259 2.417769 0.015712 
Nassarawa  664.402 43.93596 9.82438 0.066129 572.345 6.065617 1.356313 0.010598 
Sabon Tasha  694.801 8.289031 1.853484 0.01193 697.549 10.26037 2.294289 0.014709 

Source: Field Survey and Laboratory Analysis (2019 and 2020).

Table 7

Mean values of total dissolved solids concentration in groundwater of the Chikun Local Government Area (2019–2020)

WardWell
Borehole
WHO (MPL)100 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  85 5.129892 1.147079 0.060352 34.99 12.34951 2.761435 0.352944 
Kunai  90 10.25978 2.294157 0.113998 16.8 0.756724 0.169209 0.014827 
Mararaban Rido  90 10.25978 2.294157 0.113998 22.65 0.873951 0.195421 0.038636 
Nassarawa  105 5.129892 0.912198 0.097829 15.63 8.14849 1.822059 0.521337 
Sabon Tasha  48.1 2.125039 0.475173 0.04418 48.48 2.444241 0.546549 0.050418 
WardWell
Borehole
WHO (MPL)100 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  85 5.129892 1.147079 0.060352 34.99 12.34951 2.761435 0.352944 
Kunai  90 10.25978 2.294157 0.113998 16.8 0.756724 0.169209 0.014827 
Mararaban Rido  90 10.25978 2.294157 0.113998 22.65 0.873951 0.195421 0.038636 
Nassarawa  105 5.129892 0.912198 0.097829 15.63 8.14849 1.822059 0.521337 
Sabon Tasha  48.1 2.125039 0.475173 0.04418 48.48 2.444241 0.546549 0.050418 

Source: Field Survey and Laboratory Analysis (2019 and 2020).

Table 8

Mean values of calcium concentration in groundwater of the Chikun Local Government Area (2019–2020)

WardWell
Borehole
WHO (MPL)2.0 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  2.03 0.093302 0.020863 0.045962 1.899 0.097489 0.021799 0.051337 
Kunai  0.404 0.106646 0.023846 0.263649 1.83 0.317714 0.070914 0.172453 
Mararaban Rido  1.937 0.120092 0.026853 0.061999 1.972 0.147098 0.032892 0.074593 
Nassarawa  1.532 0.135436 0.030284 0.088405 2.212 0.132012 0.029519 0.05968 
Sabon Tasha  1.933 0.111596 0.024953 0.057732 1.961 0.150749 0.0337085 0.076874 
WardWell
Borehole
WHO (MPL)2.0 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  2.03 0.093302 0.020863 0.045962 1.899 0.097489 0.021799 0.051337 
Kunai  0.404 0.106646 0.023846 0.263649 1.83 0.317714 0.070914 0.172453 
Mararaban Rido  1.937 0.120092 0.026853 0.061999 1.972 0.147098 0.032892 0.074593 
Nassarawa  1.532 0.135436 0.030284 0.088405 2.212 0.132012 0.029519 0.05968 
Sabon Tasha  1.933 0.111596 0.024953 0.057732 1.961 0.150749 0.0337085 0.076874 

Source: Field Survey and Laboratory Analysis (2019 and 2020).

Table 9

Mean values of magnesium concentration in groundwater of the Chikun Local Government Area (2019–2020)

WardWell
Borehole
WHO (MPL)0.5 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  0.581 0.016189 0.005808 0.027865 0.515 0.012354 0.002762 0.023989 
Kunai  0.299 0.047514 0.010624 0.158646 0.698 0.070007 0.015654 0.100297 
Mararaban Rido  0.524 0.041218 0.009216 0.078661 0.526 0.017888 0.004 0.034009 
Nassarawa  0.486 0.030157 0.006743 0.062052 0.563 0.024730 0.005529 0.043926 
Sabon Tasha  0.571 0.058873 0.013164 0103015 0.534 0.028727 0.006423 0.053797 
WardWell
Borehole
WHO (MPL)0.5 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  0.581 0.016189 0.005808 0.027865 0.515 0.012354 0.002762 0.023989 
Kunai  0.299 0.047514 0.010624 0.158646 0.698 0.070007 0.015654 0.100297 
Mararaban Rido  0.524 0.041218 0.009216 0.078661 0.526 0.017888 0.004 0.034009 
Nassarawa  0.486 0.030157 0.006743 0.062052 0.563 0.024730 0.005529 0.043926 
Sabon Tasha  0.571 0.058873 0.013164 0103015 0.534 0.028727 0.006423 0.053797 

Source: Field Survey and Laboratory Analysis (2019 and 2020).

Table 10

Mean values of sulfate concentration in groundwater of Chikun Local Government Area (2019–2020)

WardWell
Borehole
WHO (MPL)100 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  45.665 12.317565 2.7542911 0.269738 37.85 5.6834478 1.270857 0.150157 
Kunai  16.415 0.9980375 0.223168 0.0608 23.39 1.9490619 0.435823 0.083329 
Mararaban Rido  28.215 6.8476869 1.5311893 0.242697 43.5 1.2354415 0.276253 0.028401 
Nassarawa  28.435 0.657167 0.14497 0.023111 33.92 0.6740295 0.150717 0.019871 
Sabon Tasha  35.955 2.8863517 0.6454079 0.080277 34.98 3.3516061 0.749441 0.095815 
WardWell
Borehole
WHO (MPL)100 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  45.665 12.317565 2.7542911 0.269738 37.85 5.6834478 1.270857 0.150157 
Kunai  16.415 0.9980375 0.223168 0.0608 23.39 1.9490619 0.435823 0.083329 
Mararaban Rido  28.215 6.8476869 1.5311893 0.242697 43.5 1.2354415 0.276253 0.028401 
Nassarawa  28.435 0.657167 0.14497 0.023111 33.92 0.6740295 0.150717 0.019871 
Sabon Tasha  35.955 2.8863517 0.6454079 0.080277 34.98 3.3516061 0.749441 0.095815 

Source: Field Survey and Laboratory Analysis (2019 and 2020).

Table 11

Mean values of iron concentration in groundwater of Chikun Local Government Area (2019–2020)

WardWell
Borehole
WHO (MPL)0.3 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  0.517 0.031305 0.007 0.060551 0.596 0.296832 0.066373 0.498041 
Kunai  0.225 0.069772 0.015601 0.309411 0.31 0.031455 0.007033 0.101471 
Mararaban Rido  0.437 0.078731 0.017604 0.179958 0.589 0.065284 0.014598 0.11084 
Nassarawa  0.827 0.25701 0.056058 0.302963 0.675 0.068094 0.015226 0.100881 
Sabon Tasha  1.149 0.381808 0.085375 0.332297 0.828 0.375900 0.084053 0.453986 
WardWell
Borehole
WHO (MPL)0.3 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  0.517 0.031305 0.007 0.060551 0.596 0.296832 0.066373 0.498041 
Kunai  0.225 0.069772 0.015601 0.309411 0.31 0.031455 0.007033 0.101471 
Mararaban Rido  0.437 0.078731 0.017604 0.179958 0.589 0.065284 0.014598 0.11084 
Nassarawa  0.827 0.25701 0.056058 0.302963 0.675 0.068094 0.015226 0.100881 
Sabon Tasha  1.149 0.381808 0.085375 0.332297 0.828 0.375900 0.084053 0.453986 

Source: Field Survey and Laboratory Analysis (2019 and 2020).

Table 12

Mean values of lead concentration in groundwater of Chikun Local Government Area (2019–2020)

WardWell
Borehole
WHO (MPL)0.01 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  0.009 0.001099 0.000246 0.110452 0.083 0.020026 0.004478 0.241281 
Kunai  0.007 0.001333 0.000298 0.168835 0.009 0.000852 0.000190 0.093648 
Mararaban Rido  0.056 0.051205 0.011449 0.902296 0.092 0.011964 0.002675 0.130053 
Nassarawa  0.056 0.050965 0.011396 0.901254 0.098 0.051052 0.011415 0.52094 
Sabon Tasha  0.069 0.064189 0.014353 0.918968 0.107 0.025975 0.005808 0.242764 
WardWell
Borehole
WHO (MPL)0.01 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  0.009 0.001099 0.000246 0.110452 0.083 0.020026 0.004478 0.241281 
Kunai  0.007 0.001333 0.000298 0.168835 0.009 0.000852 0.000190 0.093648 
Mararaban Rido  0.056 0.051205 0.011449 0.902296 0.092 0.011964 0.002675 0.130053 
Nassarawa  0.056 0.050965 0.011396 0.901254 0.098 0.051052 0.011415 0.52094 
Sabon Tasha  0.069 0.064189 0.014353 0.918968 0.107 0.025975 0.005808 0.242764 

Source: Field Survey and Laboratory Analysis (2019 and 2020).

Table 13

Mean values of mercury concentration in groundwater of Chikun Local Government Area (2019–2020)

WardWell
Borehole
WHO (MPL)0.001 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  0.0143 0.01628 0.003641 1.138741 0.00138 0.000832 0.000186 0.602899 
Kunai  0.000855 0.000128 0.028505 0.14924 0.0234 0.021934 0.004905 0.937346 
Mararaban Rido  0.00139 0.000827 0.000185 0.594604 0.0016 0.000821 0.000184 0.513 
Nassarawa  0.002 0.000973 0.000218 0.48665 0.00127 0.000673 0.000151 0.529843 
Sabon Tasha  0.00185 0.00104 0.000233 0.562162 0.00128 0.000668 0.000149 0.521719 
WardWell
Borehole
WHO (MPL)0.001 mg/LMeanStandard deviationStandard error meanCoefficient of variationMeanStandard deviationStandard error meanCoefficient of variation
Gwagwada  0.0143 0.01628 0.003641 1.138741 0.00138 0.000832 0.000186 0.602899 
Kunai  0.000855 0.000128 0.028505 0.14924 0.0234 0.021934 0.004905 0.937346 
Mararaban Rido  0.00139 0.000827 0.000185 0.594604 0.0016 0.000821 0.000184 0.513 
Nassarawa  0.002 0.000973 0.000218 0.48665 0.00127 0.000673 0.000151 0.529843 
Sabon Tasha  0.00185 0.00104 0.000233 0.562162 0.00128 0.000668 0.000149 0.521719 

Source: Field Survey and Laboratory Analysis (2019 and 2020).

Boreholes with water lifting motors were allowed to operate for 5 min, while those with manual pumps were allowed to run for 15 min to flush off stagnant water. Several volumes of water were poured through each sample container prior to collection. Water's chemical makeup varies as it transitions from its natural habitat to a new environment during sampling due to its dynamic nature. The ability of the water to adapt to its new environment has the greatest influence on these alterations. Notably, the content of water, particularly organic compounds, changes at variable speeds. The outlet was opened and closed numerous times to eliminate dirt particles before sampling from taps/hand pumps, and the tips of the taps/hand pumps were cleaned for a sufficient period of time to assure sterility. Before filling the bottle, the water was allowed to run freely for around 5 min. Under sterile conditions, the sample container is sealed and labeled. When the samples arrived, they were refrigerated at 4 °C. The chemical properties, including metals, were then assessed using standard water and wastewater analysis protocols (APHA 2017). A total of 50 samples were collected for laboratory analysis, or 10 samples (five from wells and five from boreholes) were taken in each of the five communities every season and year.

Groundwater modeling

The package used for this work is MODFLOW, developed by the United States Geological Survey (Boyce et al. 2020). The MODFLOW is capable of doing steady-state analysis by applying the principle of mass conservation. This principle posits that the inflow of groundwater into an aquifer is equal to the outflow rate from the aquifer. The governing equations are represented by Equation (1).
(1)
where Kxx, Kyy, Kzz are hydraulic conductivity along the x, y, and z axes (parallel to the major axes of hydraulic conductivity); h is piezometric (hydraulic) head; Q is volumetric flux per unit volume representing source/sink terms; Ss is specific storage.
The dispersion of pollutants at the dump site was modeled over a period of 30 years. The movement of solutes inside the saturated zone is regulated by the advection–dispersion equation. This equation, denoted as Equation (2), describes the behavior of solute transport in a porous medium with a uniform distribution of porosity.
(2)
where c = concentration of the solute; Rc = sources or sinks; Dij = dispersion coefficient tensor; Vi = velocity tensor.
  • Conceptual model

The groundwater-flow model was developed using a grid-based conceptual model approach. x = 1,104 m, y = 1,582 m, and z = 80 m were estimated from the Digital Elevation Model (DEM) of the study area and the borehole log of the area, respectively. The direction of groundwater flow was determined using well hydraulic heads. The area's DEM was used as the top elevation, while the borehole log was used to designate the remaining two layers. The hydraulic conductivity values of each stratum were derived from Guideal et al. (2011). The model's recharge rate of 638.46 mm/year (0.001749 m/day) was estimated based on the average annual precipitation in the Chikun Local Government Area between 2004 and 2014. This investigation utilized dispersion, longitudinal dispersivity, transverse and vertical dispersivity as estimated by Schulze-Makuch (2005). The longitudinal dispersivity was determined to be 8.5 m, while the transversal and vertical dispersivities were determined to be 0.85 and 0.085 m, respectively.

Emotional artificial neural network

The EANN represents a progression in the field of ANNs. This mechanism enables neurons to generate agents capable of modifying cognitive, emotional, and executive functions as needed (Sharghi et al. 2018; Molajou et al. 2021). To clarify, EANN models represent an advancement in the field of ANN models. These models introduce an artificial sensing unit capable of releasing hormones to regulate the functioning of nodes (neurons). Additionally, the hormone weights within EANN models can be adjusted according to the input and output values of the nodes (Sharghi et al. 2019).

In this study, the EANN model, known for its strong performance in hydrological investigations, was employed to examine the WQI. One notable benefit of this particular model lies in its satisfactory performance when applied to studies characterized by limited data availability (Nourani et al. 2019). Given the limited data available to researchers from groundwater quality studies, it can be considered a viable approach for predicting the WQI. Consequently, the EANN method has been employed in the present study.

The objective of the suggested EANN is to attain an ideal neural network characterized by reduced computational complexity, hence enabling the identification of intricate and non-linear systems.

Some existing systems are complex and non-linear in actuality, making it difficult to develop a suitable mathematical model for them (Romero Ugalde et al. 2015). In addition, the design of the controller requires a suitable model of the system, particularly for model-based controllers are identical to model-based predictive control (MPC), which necessitates an accurate system model (Kittisupakorn et al. 2009). In fact, the precision of this model will impact the system's efficacy. Nevertheless, you should go. Also monitored was the quantity of computational load, as achieving high accuracy necessitates increased computational load and complexity (Prakash & Srinivasan 2009). In order to accomplish this simplicity and reduce the complexity of calculations, they typically employ linear models, despite the fact that real systems exhibit completely nonlinear behavior.

Existing plants whose equations cannot be accessed can be evaluated based on input and output data without sufficient knowledge of the plant's internal structure. ANNs are the most well-known of these techniques because they can identify a system whose inputs and outputs can be measured and provide a model for it (Chen & Billings 1992). Multi-layer perceptron neural networks are extensively utilized. The neural network model for each output is described in Equation (3).
(3)
Due to the presence of a concealed layer, the Equation (3) has a computational complexity of O(m × n) when viewed from the perspective of complexity. N is the number of inputs, and m is the number of hidden layer neurons. Inputs to neural networks can include both current and historical data, as well as historical outputs. Figure 2 demonstrates that information is transmitted in the form of feedback between the input and output units in each hidden node of EANN. These nodes provide active hormones of Ha, Hb, and Hc, and these parameters are designed in the training phase of the model based on input and output values, followed by the training process.
Figure 2

The unit of the EANN model (Sharghi et al. 2018).

Figure 2

The unit of the EANN model (Sharghi et al. 2018).

Close modal
In the course of training, hormonal constants influence other nodal units. In the given EANN, the output of the ith node containing the three hormones Ha, Hb, and Hc is computed as Equation (4).
(4)
where
(5)

First term of Equation (5) represents the activation function weights. This incorporates both the fixed neural weight and the fluctuating hormone weight. The second term is the weight used in the sum function. Third term specifies the weight associated with ijx. Fourth term illustrates the sum function's tendencies. Factors must regulate the distribution of total hormone level (Hh) among hormones.

Consequently, the glandularity factor should be calibrated during the EANN training phase so that the glands receive an adequate amount of hormone. It is possible to implement schemes for activating Hh hormone values based on input samples, such as the training sample's average input values. The hormone values are then updated in the learning process based on the output network and the relationship Equations (4) and (5) in order to find a suitable match between the objectives of the time series.

The results of the hydro-chemical analyses are presented in Tables 413, highlighting the results of laboratory analysis of water samples, statistical analysis as well as comparing the concentration of the physico-chemical parameters of groundwater with that of World Health Organization (WHO) (2022) standards for portable water are discussed.

  • pH

The pH value of water samples in the study area indicated a minimum statistical mean value of 6.4 in the sample collected at Nassarawa and a maximum statistical mean value of 7.07 in the sample collected at Gwagwada for wells and with a maximum statistical mean value of 6.85 in the samples collected at Mararaban Rido. A minimum value of 6.43 was observed in samples from Sabon Tasha for boreholes as presented in Table 4.

The varying pH values in the groundwater system may be attributed to the variation of photosynthetic activity, disposal of untreated wastewaters, agricultural and anthropogenic activities (Kanchanapiya & Tantisattayakul 2022).

The standard values of pH for drinking water as per WHO is between 6.5–8.5% and 95.63% of the samples analyzed from the entire study area during both rainy and dry seasons, have pH values within the permissible limits of WHO (2022) and could be classified as suitable for drinking purpose. However, pH alone cannot be taken as a criterion for determining portability of water.

  • Turbidity

The turbidity values (NTU) for the groundwater samples is presented in Table 5. The values obtained for wells indicated a minimum mean value of 46.5 and a maximum mean value of 49.35 NTU and that of boreholes indicated a minimum mean value of 15.63 and a maximum mean value of 48.48 NTU in the samples collected for both rainy and dry seasons in the study area. Furthermore, it was observed that the turbidity values of groundwater samples during the rainy season have indicated an increasing trend when compared to dry season. However, all the samples have turbidity values falling within the permissible limits of WHO (2022) (maximum 50 NTU).

  • Electrical conductivity

Table 6 displays the statistical mean and electrical conductivity (EC) (μS/cm) data. The observed EC values in the area during the study period, with a maximum statistical mean of 694.80 s/cm and a minimum statistical mean of 554.69 s/cm in groundwater samples collected from wells and a maximum statistical mean of 697.54 s/cm and a minimum statistical mean of 572.34 s/cm in groundwater samples collected from borehole for both seasons during the study period. The EC values in boreholes have shown a rising tendency as compared to wells, owing to the fact that the dissolution of salts, minerals, and other soil elements increases as the groundwater table rises. The majority of inorganic salts, such as NaCl, are responsible for raising groundwater EC values. The findings found that groundwater in the entire research area falls into the acceptable range. This result compares favorably with that of Vivan et al. (2012) who opined that the EC is a useful parameter for water quality indicating salinity hazards. In general waters with conductivity values below 750 (μs/cm) are satisfactory; conductivity values ranging between 250 and 750 (μS/cm) are widely used for crop growth. Akpoborie (2011) observed that a sudden rise in conductivity in the water indicates addition of some pollutants to it, and that the area having higher EC also has high pH. Groundwater has normally a large amount of dissolved inorganic matter and therefore high values are not unexpected.

  • Total dissolved solids

The total dissolved solids (TDS) values for the groundwater samples are given in Table 7.

Table 7 indicated that TDS value varies from a minimum statistical mean value of 48 mg/L in the groundwater samples collected at Sabo Tasha ward to a maximum statistical mean value of 105 mg/L in samples collected at Nassarawa ward for wells had a maximum statistical mean of 48 mg/L and for samples collected at Sabo Tasha, while a minimum value of 15 mg/L was obtained in samples collected at Nassarwa ward for boreholes. Further TDS values have exhibited an increasing trend in concentration during rainy season compared to dry season. This may be due to the dissolution of more quantity of constituents of soil particles as groundwater table increases during rainy season.

  • Calcium

The values of calcium obtained for the five (5) settlements from wells and boreholes in the study area with minimum and maximum mean values are presented in Table 8.

Table 8 reveals that the calcium concentration varies from a minimum statistical mean of 0.40 mg/L in the groundwater samples collected at Kunai to a maximum statistical mean of 2.03 mg/L in samples collected at Gwagwada for wells, while the concentration varies from a minimum statistical mean of 0.001 mg/L in the groundwater samples collected at Sabon Tasha, to a maximum statistical mean of 1.97 mg/L in samples collected at Mararaban Rido for boreholes. It was observed that most of the samples have exhibited an increasing trend in calcium concentration boreholes compared to wells. Tse & Adamu (2012) have expressed opinion that the high concentrations of calcium have no health hazard. Yusuf (2015) attested to this as he reported that calcium is an essential macro element owing to its functions in bone structure, muscle contraction, blood clotting, etc. Excess of calcium has a teratogenic action in chicks and depresses the functioning of muscles and nerve tissues. However, it should be noted that in human beings, hyper-calcimea causes coma and death if serum calcium rises to 160 mg/L. Besides, it is important to note that calcium has indicated strong significant correlation with total hardness and total dissolved solid.

  • Magnesium

The values of magnesium obtained for the five (5) selected wards from wells and boreholes in the study area with minimum and maximum mean values are presented in Table 9.

The magnesium concentration varies from a minimum statistical mean of 0.05 mg/L in the groundwater samples collected in Kunai ward; to a maximum statistical mean of 0.58 mg/L in samples collected in Gwagwada ward. It was observed that most of the samples have exhibited an increasing trend in Magnesium concentration boreholes compared to wells.

Magnesium is also an essential macro nutrient for human beings. It forms part of structure of the body. It plays a critical role in cell metabolism. Magnesium toxicity in higher doses greater than 400 mg/L causes nausea, muscular weakness and paralysis in humans and mammals (Vivan et al. 2012). Newborn infants develop hyper magnesemia if mother is treated with MgSO4 drugs.

The results of magnesium analysis have revealed that most of the samples have exceeded the permissible limits of 0.5 mg/L.

  • Sulfate

The concentration of sulfate in groundwater in the study area is presented on Table 10. The sulfate concentration varied from a minimum statistical mean of 16.4 mg/L in the groundwater samples collected at Kunai ward, to a maximum statistical mean of 45.66 mg/L in the samples collected at Nasarawa ward for wells and with a minimum statistical mean of 37.85 mg/L in Gwagwada ward for boreholes. It was noticed that the sulfate values have exhibited an increasing trend in concentration boreholes compared to wells. This may be attributed to the dissolution of more quantity sulfate minerals at increased depth due to rise in the groundwater table by recharge process. Considerable quantity of sulfate has also been added to the hydrologic cycle from precipitation (rainfall). The agriculture run-off and irrigation drainage carry these sulfate minerals in soil and due to variation in the temperature conditions, the breakdown of organic substances in soil, leachable sulfates present in fertilizers and other human interferences are the expected causes for the high concentration of sulfates (Asiwaju-Bello & Ololade 2013). Generally, the concentration of sulfate in all the groundwater samples (boreholes and wells) collected fall within the permissible limits of 100 mg/L.

  • Iron

The analysis of groundwater in the study area revealed that the concentration of dissolved iron ranged from a minimum statistical mean of 0.22 mg/L to a maximum statistical mean of 1.14 mg/L for wells over the course of the study. Similarly, the concentration of dissolved iron in boreholes within the same study area ranged from a minimum statistical mean of 0.22 mg/L to a maximum statistical mean of 0.82 mg/L, as shown in Table 11.

The upper threshold for iron concentration is set at 1.0 mg/L, exceeding which can result in alterations in taste and appearance, as well as negative consequences for household applications, including the potential for staining clothes and utensils. When the concentration of iron in water surpasses 0.3 mg/L, it has adverse effects on water supply infrastructure and facilitates the growth of iron bacteria. The analysis revealed that the iron concentration in the samples exceeded the acceptable thresholds.

  • Lead

The distribution of concentration lead in the study area is presented in Table 12.

The concentration of lead ranged from a minimum statistical mean of 0.0024 mg/L to a maximum statistical mean of 0.05 mg/L for wells and a minimum statistical mean value of 0.97, a maximum statistical mean value of 1.91 for boreholes. Thus, the concentration of lead observed is well above the safe limit for most of the groundwater samples in the study area. As the maximum permissible limit is 0.01 mg/L.

  • Mercury

The concentration of mercury ranges from a minimum statistical mean of 0.008 mg/L to a maximum statistical mean of 0.0143 mg/L for wells and a minimum statistical mean value of 0.00127, a maximum statistical mean value of 0.0234 for boreholes as presented on Table 13. The concentration of mercury observed is well above the safe limit for most of the groundwater samples in the study area. The concentration ranged between 1.07 and 1.67 mg/L, and 1.11 and 1.77 mg/L for borehole and hand dug well water samples, respectively (Kalip et al. 2022). Mean concentrations were 1.16 mg/L for boreholes and 1.76 mg/L for wells. While the average values are within the maximum permissible limits set by USEPA, but were far greater than the 0.001 mg/L WHO world average. However, several incident values from wells and boreholes exceeded the USEPA maximum permissible limits, while the annual effective doses of all samples were within the recommended limits.

The physico-chemical analysis conducted on the groundwater in the Chikun Local Government Area, Kaduna State indicated that a significant proportion of the water samples examined exhibited substandard quality. According to the findings of the study, it was observed that 45% of the total samples collected from the study area were determined to be non-portable when comparing the laboratory results with the drinking water standards set by the WHO (2022). Among the parameters contributing to the lack of portability, heavy metals, total hardness, and TDS emerged as the three prominent factors.

The analysis of the WQI conducted for the groundwater in the Chikun Local Government Area indicated that a significant proportion of the samples demonstrated inadequate water quality, rendering them unsuitable for consumption. The WQI findings presented in Tables 14 and 15 indicate that samples obtained from boreholes exhibit a poor rating, with a score of 106.521. Conversely, samples collected from wells demonstrate a good rating, with a score of 85.450. Nine wells were randomly selected for sampling, and the concentrations of 10 physico-chemical parameters were measured. The objective was to evaluate the appropriateness of the groundwater in the area for human consumption. The disparity between the laboratory measurements of the samples and the potable water standards set by the WHO (2022) was assessed through laboratory analysis and inferential statistics. The findings indicate a notable disparity between the concentrations of the selected parameters in the samples and the potable water standards established by the WHO (2022). The study recommends disinfecting water from hand-dug open wells in the area prior to human consumption.

Table 14

WQI of wells in Chikun Local Government Area

ParametersTest results (Vn)Standard permissible value (Si)Relative weight (Wi)Quality rating (Qi)Weighted value {(Wi)*(Qi)}
pH 6.8 6.5–8.5 0.04000 88.4 3.54 
Turbidity 46.78 50 0.11764 6.66 0.784 
Total hardness (NTU) 27.32  0.03333 326.6 10.88 
TDS (mg/L) 93  0.00200 24 0.048 
Electrical Conductivity (μS/cm) 624.7 15,000 0.13333 94.36 12.58 
CO2 (mg/L) 30.52 50 1.00000 2.8 2.8 
Nitrite (mg/L) Nil 0.2 0.10000 42.1 4.21 
Sulfate (mg/L) 30.9 100 1.00000 
Copper (mg/L) 0.912 1.0 0.00000 
Iron (mg/L) 0.626 0.3 0.02500 135 3.375 
Cadium (mg/L) 0.502 0.01 0.20000 70.2 14.04 
Calcium (mg/L) 1.15 2.0 10.0000 121 1200 
Mercury (mg/L) 0.004 2.0 0.00500 4.75 0.0237 
Lead (mg/L) 0.039 0.001 1.00000 212 212 
Magnesium (mg/L) 0.437 0.01 0.20000 28.6 5.72 
Coliform bacteria (MPN/mL) 0.00 1.0 0.00000 
  Sum 13.8563 – 1,476.001 
ParametersTest results (Vn)Standard permissible value (Si)Relative weight (Wi)Quality rating (Qi)Weighted value {(Wi)*(Qi)}
pH 6.8 6.5–8.5 0.04000 88.4 3.54 
Turbidity 46.78 50 0.11764 6.66 0.784 
Total hardness (NTU) 27.32  0.03333 326.6 10.88 
TDS (mg/L) 93  0.00200 24 0.048 
Electrical Conductivity (μS/cm) 624.7 15,000 0.13333 94.36 12.58 
CO2 (mg/L) 30.52 50 1.00000 2.8 2.8 
Nitrite (mg/L) Nil 0.2 0.10000 42.1 4.21 
Sulfate (mg/L) 30.9 100 1.00000 
Copper (mg/L) 0.912 1.0 0.00000 
Iron (mg/L) 0.626 0.3 0.02500 135 3.375 
Cadium (mg/L) 0.502 0.01 0.20000 70.2 14.04 
Calcium (mg/L) 1.15 2.0 10.0000 121 1200 
Mercury (mg/L) 0.004 2.0 0.00500 4.75 0.0237 
Lead (mg/L) 0.039 0.001 1.00000 212 212 
Magnesium (mg/L) 0.437 0.01 0.20000 28.6 5.72 
Coliform bacteria (MPN/mL) 0.00 1.0 0.00000 
  Sum 13.8563 – 1,476.001 

Table 15

WQI of boreholes in the Chikun Local Government Area

ParametersTest results (Vn)Standard permissible value (Si)Relative weight (Wi)Quality rating (Qi)Weighted value {(Wi) *(Qi)}
pH 6.65 6.5–8.5 0.04000 88.4 3.54 
Turbidity 23.4 50 0.11764 6.6 0.78 
Total hardness (NTU) 60  0.03333 50.0 1.6 
TDS (mg/L) 80  0.00200 6.0 0.012 
Electrical conductivity (μs/cm) 699.8 15,000 0.13333 44.79 5.971 
CO2 (mg/L) 33.0 50 1.00000 97 97 
Nitrite (mg/L) Nil 0.2 0.10000 30.1 3.01 
Sulfate (mg/L) 44.9 100 1.00000 5.0 
Copper (mg/L) 1.70 1.0 0.00000 
Iron (mg/L) 0.71 0.3 0.02500 55 1.375 
Cadium (mg/L) 0.1 0.01 0.20000 76 15.2 
Calcium (mg/L) 0.11 2.0 10.0000 98 980.0 
Mercury (mg/L) 2.13 2.0 0.00500 0.415 0.021 
Lead (mg/L) 0.1 0.001 1.00000 70 70.0 
Magnesium (mg/L) 0.5 0.01 0.20000 2.6 0.52 
Coliform bacteria (MPN/mL) 0.0 1.0 0.00000 
  Sum 13.8563  1,184.029 
ParametersTest results (Vn)Standard permissible value (Si)Relative weight (Wi)Quality rating (Qi)Weighted value {(Wi) *(Qi)}
pH 6.65 6.5–8.5 0.04000 88.4 3.54 
Turbidity 23.4 50 0.11764 6.6 0.78 
Total hardness (NTU) 60  0.03333 50.0 1.6 
TDS (mg/L) 80  0.00200 6.0 0.012 
Electrical conductivity (μs/cm) 699.8 15,000 0.13333 44.79 5.971 
CO2 (mg/L) 33.0 50 1.00000 97 97 
Nitrite (mg/L) Nil 0.2 0.10000 30.1 3.01 
Sulfate (mg/L) 44.9 100 1.00000 5.0 
Copper (mg/L) 1.70 1.0 0.00000 
Iron (mg/L) 0.71 0.3 0.02500 55 1.375 
Cadium (mg/L) 0.1 0.01 0.20000 76 15.2 
Calcium (mg/L) 0.11 2.0 10.0000 98 980.0 
Mercury (mg/L) 2.13 2.0 0.00500 0.415 0.021 
Lead (mg/L) 0.1 0.001 1.00000 70 70.0 
Magnesium (mg/L) 0.5 0.01 0.20000 2.6 0.52 
Coliform bacteria (MPN/mL) 0.0 1.0 0.00000 
  Sum 13.8563  1,184.029 

Figure 3 depicts scatterplots demonstrating the improved performance of the EANN models over the basic water WQI. The training dataset comprised 80% of the available data, with the remaining 20% used to evaluate the network's performance. Scatter plots show a higher degree of concordance between predicted and observed values, implying that hybrid models are more precise and reliable.
Figure 3

EANN model results.

Figure 3

EANN model results.

Close modal
The R2 index and normalized root mean square error (RMSENormalized) values for the training and test stages, respectively, are 0.89 and 0.18, and 0.83 and 0.23 (Figure 4). Based on the current situation, the use of the EANN model demonstrates its potential for evaluating water quality due to its predictive abilities even when faced with insufficient data. Nonetheless, increasing the model's training dataset leads to an improvement in the overall quality and precision of the generated output. It is critical to recognize that many parts of the world suffer from data poverty. To improve the assessment process, it is critical to use models with a high level of reliability, thereby addressing this limitation. The scatterplots for the combined models show a notable tendency for data points to tightly cluster along the diagonal line, indicating a high level of agreement between predicted and observed values. Scatterplots depicting simple WQI, on the other hand, show greater dispersion and deviations from the diagonal line, indicating a lower correlation between projected and actual measurements. The findings of this study highlight the effectiveness of incorporating artificial intelligence and decision tree methodologies into the traditional WQI framework for evaluating groundwater quality. EANN models have been found to have improved levels of accuracy, reliability, and predictive capability, making them extremely valuable tools in the realm of water quality management and decision-making procedures.
Figure 4

EANN model performance.

Figure 4

EANN model performance.

Close modal

Water is a unique resource with a fixed quantity, but its quantity and quality vary over time and space. Population growth, especially in developing countries, and water demand have continued. This increase in demand has caused water scarcity in many countries, and it is endemic everywhere Other than domestic water needs, fishing and farming require water. The primary objective of this study was to evaluate the quality of groundwater and develop a strategy for managing groundwater resources in the Chikun Local Government Area. This study indicates that the current system places a significant emphasis on excluding residents from groundwater management. Therefore, the present study utilized the method of realistic evaluation to establish that the groundwater in the Chikun Local Government Area is classified as hard to very hard water. In addition, the findings indicate that the groundwater extracted from the majority of bore wells in the study area is unfit for human consumption according to current standards and protocols. The influence of pit latrines, open waste sites, and other non-point sources is the primary cause of groundwater contamination in the area of the case study. This investigation's primary objective was to identify and evaluate the relative significance of potential sources of groundwater contamination in the Chikun Local Government Area. For the ground water modeling the MODFLOW and EANN was used. The EANN model performance was evaluated using 20% of the data, while 80% was used for training. The ultimate aim of this study was to develop effective strategies for mitigating the adverse impacts of these contamination sources on the underlying aquifer. The findings of the groundwater quality analysis suggest that the water is appropriate for both domestic use, such as consumption and household applications. Based on the petrographic analyses conducted, it can be inferred that the upper horizon of the sedimentary units in the case study area is primarily composed of fine-grained materials. These fine-grained materials are believed to offer more efficient physico-chemical barriers compared to the coarse sands found at the base of the sedimentary units. The R2 index and RMSENormalized for training and test are 0.89 and 0.18, for train phase and 0.83 and 0.23 for test phase. According to the current situation, the EANN model can predict water quality even with limited data. However, increasing the model's training dataset improves output quality and precision. Note that data poverty exists in many countries. To overcome this limitation, reliable models must be used to improve assessment. The combined models' scatterplots show data points clustering tightly along the diagonal line, indicating high agreement between predicted and observed values.

This study demonstrates that incorporating EANN into the WQI framework for evaluating groundwater quality works. EANN models enhance accuracy, reliability, and predictive power, making them useful tools for water quality management and decision-making. Within the confines of the case study, it is possible that the aforementioned variables contributed to a reduction in the level of pollutants in the groundwater. To effectively address societal concerns and achieve sustainable management practices, future sustainable management of vulnerable aquifers necessitates the conduct of hydrochemical research.

We would like to acknowledge the Scientific Research Deanship at University of Ha'il – Saudi Arabia for funding this project with number RG-20 220.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.

Akpoborie
I. A.
2011
Aspects of the hydrology of the western Niger delta wetlands: groundwater conditions in the neogene (recent) deposits of the Ndokwa Area
.
African Geosciences Review
18
(
3
),
25
36
.
APHA
2017
Standard Methods for Examination of Water and Wastewater
, 23rd edn.
APHA, AWWA, WPCF, Washington, DC, USA.
Ashraf
S.
,
Ali
M.
,
Sohail
M.
&
Eldin
S. M.
2023
Assessing the environmental impact of industrial pollution using the complex intuitionistic fuzzy ELECTREE method: a case study of pollution control measures
.
Frontiers in Environmental Science
11
,
610
.
Asiwaju-Bello
Y. A.
&
Ololade
J. O.
2013
Groundwater potential of basement aquifers in part of southwestern Nigeria
.
American International Journal of Contemporary Research
3
(
3
),
124
141
.
Boyce
S. E.
,
Hanson
R. T.
,
Ferguson
I.
,
Schmid
W.
,
Henson
W.
,
Reimann
T.
,
Mehl
S. M.
&
Earll
M. M.
2020
One-water Hydrologic Flow Model: A MODFLOW Based Conjunctive-use Simulation Software
.
U.S. Geological Survey Techniques and Methods
.
Chen
S.
&
Billings
S.
1992
Neural networks for nonlinear dynamic system modelling and identification
.
International Journal of Control
56
,
319
346
.
Dhaka
S. K.
&
Bhaskar
N.
2017
Assessment of ground water quality in terms of water quality index and regression analysis of water quality parameters
.
Journal of Basic and Applied Engineering Research
4
(
4
),
339
342
.
Gradilla-Hernández
M. S.
,
de Anda
J.
,
Garcia-Gonzalez
A.
,
Montes
C. Y.
,
Barrios-Piña
H.
,
Ruiz-Palomino
P.
&
Díaz-Vázquez
D.
2020
Assessment of the water quality of a subtropical lake using the NSF-WQI and a newly proposed ecosystem specific water quality index
.
Environmental Monitoring and Assessment
192
(
5
),
296
.
Guideal
R.
,
Bala
A. E.
&
Ikpokonte
A. E.
2011
Preliminary estimates of the hydraulic properties of the quaternary aquifer in N'Djaména area, Chad Republic
.
Journal of Applied Sciences
11
(
3
),
542
548
.
Available from
:
https://www.scialert.net/abstract/?doi = jas.2011.542.548
.
Hassan Ibrahim
A.
,
Alhaji Yaro
N.
&
Olufemi Adebola
A.
2017
Assessing the socio-economic impact of gully erosion in Chikun Local Government Area, Kaduna State, Nigeria
.
Science World Journal
12
(
1
),
42
47
.
Heyi
E. A.
,
Dinka
M. O.
&
Mamo
G.
2023
Assessing the impact of climate change on water resources of upper Awash River sub-basin, Ethiopia
.
Journal of Water and Land Development
52
,
232
244
.
Ishaku
J. M.
2011
Assessment of groundwater quality index for Jimeta-Yola area, Northeastern Nigeria
.
Journal of Geology and Mining Research
3
(
9
),
219
231
.
Javidan
P.
,
Baghdadi
M.
,
Torabian
A.
&
Aminzadeh Goharrizi
B.
2022
A tailored metal–organic framework applicable at natural pH for the removal of 17α-ethinylestradiol from surface water
.
Desalination and Water Treatment
264
,
259
269
.
Kalip
A.
,
Yusuf
J.
&
Aremu
S. O.
2022
Assessment of radon and heavy metals in groundwater sources from Kaduna and Environs, Nigeria
.
International Journal of Innovative Science, Engineering & Technology
7
(
12
),
102
112
.
Kanchanapiya
P.
&
Tantisattayakul
T.
2022
Wastewater reclamation trends in Thailand
.
Water Science and Technology
86
(
11
),
2878
2911
.
Kittisupakorn
P.
,
Thitiyasook
P.
,
Hussain
M. A.
&
Daosud
W.
2009
Neural network based model predictive control for a steel pickling process
.
Journal of Process Control
19
(
4
),
579
590
.
Komolafe
A. A.
,
Abdul-Azeez
A. S.
,
Biodun
A. A. Y.
,
Omowonuola
A. F.
&
Rotimi
A. D.
2014
Air pollution and climate change in Lagos, Nigeria: needs for proactive approaches to risk management and adaptation
.
American Journal of Environmental Sciences
10
(
4
),
412
423
.
Konin
N. J. C.
,
N'go
Y. A.
,
Soro
G. E.
&
Goula
B. T. A.
2023
Impact of rainfall trends on flood in Agnéby watershed
.
Journal of Water and Land Development
52
,
9
20
.
Kubiak
J.
,
Machula
S.
,
Czerniejewski
P.
,
Brysiewicz
A.
&
Wawrzyniak
W.
2021
Long term changes in the quality and water trophy of Lake Ińsko – the effect of the re-oligotrophication?
Journal of Water and Land Development
51
,
30
37
.
Lapworth
D. J.
,
Nkhuwa
D. C. W.
,
Okotto-Okotto
J.
,
Pedley
S.
,
Stuart
M. E.
,
Tijani
M. N.
&
Wright
J.
2017
Urban groundwater quality in sub-Saharan Africa: current status and implications for water security and public health
.
Hydrogeology Journal
25
(
4
),
1093
1116
.
Maroeto
M.
&
Santoso
W.
2023
Dynamic model engineering of sustainable agriculture as a pragmatic concept of critical land management
.
Journal of Water and Land Development
(
55
),
56
59
.
Molajou
A.
,
Nourani
V.
,
Afshar
A.
,
Khosravi
M.
&
Brysiewicz
A.
2021
Optimal design and feature selection by genetic algorithm for emotional artificial neural network (EANN) in rainfall-runoff modeling
.
Water Resources Management
35
(
8
),
2369
2384
.
Nourani
V.
,
Molajou
A.
,
Uzelaltinbulat
S.
&
Sadikoglu
F.
2019
Emotional artificial neural networks (EANNs) for multi-step ahead prediction of monthly precipitation; case study: northern Cyprus
.
Theoretical and Applied Climatology
138
(
3–4
),
1419
1434
.
Romero Ugalde
H. M.
,
Carmona
J.-C.
,
Reyes-Reyes
J.
,
Alvarado
V. M.
&
Mantilla
J.
2015
Computational cost improvement of neural network models in black box nonlinear system identification
.
Neurocomputing
166
,
96
108
.
Samira
K.
,
Gadiga
B.
&
Martins
A.
2015
An assessment of water quality from hand-dug wells in Bayan Duste Narayi in Chikun Local Government of Kaduna State
.
International Journal of Engineering Research & Technology
4
(
2
),
924
927
.
Sharghi
E.
,
Nourani
V.
,
Molajou
A.
&
Najafi
H.
2019
Conjunction of emotional ANN (EANN) and wavelet transform for rainfall-runoff modeling
.
Journal of Hydroinformatics
21
,
136
152
.
Tse
A. C.
&
Adamu
C. I.
2012
Assessment of anthropogenic influence on groundwater in hand dug wells in Makurdi metropolis
.
North Central Nigeria. Ife Journal of Science
14
(
1
),
21
32
.
Ualiyeva
R. M.
,
Kukusheva
A. N.
,
Insebayeva
M. K.
,
Akhmetov
K. K.
,
Zhangazin
S. B.
&
Krykbayeva
M. S.
2023
Agrotechnological methods of plant feeders applying for spring wheat agrocenoses – North-Eastern Kazakhstan varieties
.
Journal of Water and Land Development
55
,
28
40
.
Uddin
M. G.
,
Nash
S.
&
Olbert
A. I.
2021
A review of water quality index models and their use for assessing surface water quality
.
Ecological Indicators
122
,
107218
.
Vivan
E. L.
,
Caleb
A. I.
&
Ndahi
A. K.
2012
Effects of effluent discharge of Kaduna refinery on the water quality of river Romi
.
Journal of Research in Environmental Science and Toxicology
1
(
3
),
41
46
.
Vivan
E. L.
,
Ali
A. Y.
,
Obasi
M. T.
,
Emmanuel
J. N.
,
Daloeng
H. M.
,
Giwa
C. Y.
,
Watson
S. S.
&
Yakubu
M. T.
2021
Effect of artisanal mining on groundwater quality in Antang District, Jema'al Local Government Area of Kaduna State, Nigeria
.
Bayero Journal of Pure and Applied Sciences
13
(
1
),
169
180
.
WHO
2022
Guidelines for Drinking-Water Quality: Fourth Edition Incorporating the First and Second Addenda
.
World Health Organization
,
Geneva, Switzerland
.
Yilma
M.
,
Kiflie
Z.
,
Windsperger
A.
&
Gessese
N.
2018
Application of artificial neural network in water quality index prediction: a case study in Little Akaki River, Addis Ababa, Ethiopia
.
Modeling Earth Systems and Environment
4
(
1
),
175
187
.
Yusuf
A. K.
2015
Groundwater resource management strategy in the Nigerian sector of the Chad basin
.
Journal of Natural Sciences Research
5
(
14
),
56
63
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).