Carcinogenic health risks and water quality assessment of groundwater around lead–zinc mining areas of Ebonyi state Nigeria: a data-driven machine learning approach
Abstract This study evaluates carcinogenic health risks and groundwater quality in lead–zinc mining areas of Ebonyi State, Nigeria, using machine learning to predict contamination levels, providing crucial insights for public health and environmental management in a region heavily impacted by mining...
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2024-11-01
|
| Series: | Discover Civil Engineering |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s44290-024-00134-3 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850064859917451264 |
|---|---|
| author | Obinna Chigoziem Akakuru Moses Oghenenyoreme Eyankware Godwin O. Aigbadon Ayatu Ojonugwa Usman Alexander Iheanyi Opara Kizito Ojochenemi Musa Micheal Akaninyene Okon Okechukwu Pius Aghamelu Gabriel Ehriga Odesa Ifeyinwa Juliana Ofoh Annabel U. Obinna-Akakuru |
| author_facet | Obinna Chigoziem Akakuru Moses Oghenenyoreme Eyankware Godwin O. Aigbadon Ayatu Ojonugwa Usman Alexander Iheanyi Opara Kizito Ojochenemi Musa Micheal Akaninyene Okon Okechukwu Pius Aghamelu Gabriel Ehriga Odesa Ifeyinwa Juliana Ofoh Annabel U. Obinna-Akakuru |
| author_sort | Obinna Chigoziem Akakuru |
| collection | DOAJ |
| description | Abstract This study evaluates carcinogenic health risks and groundwater quality in lead–zinc mining areas of Ebonyi State, Nigeria, using machine learning to predict contamination levels, providing crucial insights for public health and environmental management in a region heavily impacted by mining activities. A systematic and purposive sampling approach was employed to collect 42 groundwater samples from various sources across the study area and analyzed using standard methods. Machine learning was preferred over traditional methods for its ability to capture complex, non-linear relationships in the data, allowing for more accurate predictions of contamination levels and health risks, which is crucial for targeted monitoring and efficient groundwater management in mining-impacted areas. The pH levels ranged from 3.64 to 7.36, with a mean of 5.53, indicating significant acidity below the World Health Organization (WHO) standard. Heavy metal concentrations were alarming: cadmium ranged from 0.06 to 0.69 (mean 0.37), copper from 0.03 to 2.43 (mean 1.18), nickel from 0.03 to 1.81 (mean 0.85), mercury from 0.16 to 2.19 (mean 1.24), and arsenic from 0.09 to 4.15 (mean 1.94). All exceeded WHO limits, posing severe health risks. The correlation analysis of heavy metals identified a statistically significant negative relationship between Cadmium (Cd) and Chromium (Cr) (r = − 0.317, p = 0.041), and a marginally significant negative correlation between Mercury (Hg) and Selenium (Se) (r = − 0.273, p = 0.08). Most other correlations, including those between pH and metals such as Copper (Cu) and Zinc (Zn), were not statistically significant, indicating weak or non-existent linear relationships among these variables. The Principal Component Analysis (PCA) results suggest that groundwater quality in the mining area is influenced by multiple factors, including pH levels and the presence of various heavy metals. The Contamination Factor (CF) and Heavy Metal Evaluation Index (HEI) highlight severe contamination, particularly with Hg. The Pollution Load Index (PLI), Water Quality Index (WQI), and Heavy Metal Pollution Index (HPI), further demonstrate the widespread nature of the pollution, with numerous locations exhibiting significantly degraded water quality. The Geo-accumulation index (Igeo) and Nemerow Pollution Index (NPI) indicate moderate yet persistent contamination, reflecting the ongoing impact of mining activities throughout the region. Carcinogenic risk assessments highlighted greater risks for children, with chromium and arsenic levels reaching up to 0.36 mg/L and 0.42 mg/L, respectively. Specific sites, such as GW12, GW23, and GW30, were identified as high-risk areas. The Artificial Neural Networks, Support Vector Machines, and Multi-Linear Regression (MLR) models effectively predicted pollution indices, with MLR showing the most reliable performance (perfect R2 value). Seven predictive models were generated for key contamination metrics, offering valuable insights for future monitoring and management. The novelty of this study lies in its application of machine learning models to accurately predict contamination levels in groundwater around lead–zinc mining areas in Ebonyi State, Nigeria, providing a data-driven approach to enhance environmental monitoring and public health protection in a highly impacted region. |
| format | Article |
| id | doaj-art-12a653b7ad4a47bd873f7a330548ddc7 |
| institution | DOAJ |
| issn | 2948-1546 |
| language | English |
| publishDate | 2024-11-01 |
| publisher | Springer |
| record_format | Article |
| series | Discover Civil Engineering |
| spelling | doaj-art-12a653b7ad4a47bd873f7a330548ddc72025-08-20T02:49:09ZengSpringerDiscover Civil Engineering2948-15462024-11-011114010.1007/s44290-024-00134-3Carcinogenic health risks and water quality assessment of groundwater around lead–zinc mining areas of Ebonyi state Nigeria: a data-driven machine learning approachObinna Chigoziem Akakuru0Moses Oghenenyoreme Eyankware1Godwin O. Aigbadon2Ayatu Ojonugwa Usman3Alexander Iheanyi Opara4Kizito Ojochenemi Musa5Micheal Akaninyene Okon6Okechukwu Pius Aghamelu7Gabriel Ehriga Odesa8Ifeyinwa Juliana Ofoh9Annabel U. Obinna-Akakuru10Department of Geology, Federal University of TechnologyDepartment of Geology, Faculty of Science, Dennis Osadebay UniversityDepartment of Geology, University of BeninDepartment of Physics, Geology and Geophysics, Alex Ekwueme Federal UniversityDepartment of Geology, Federal University of TechnologyDepartment of Geology, Faculty of Sciences, Federal University LokojaDepartment of Soil Science and Technology, Federal University of TechnologyDepartment of Physics, Geology and Geophysics, Alex Ekwueme Federal UniversityDepartment of Geology, Faculty of Science, Dennis Osadebay UniversityDepartment of Geology, Federal University of TechnologyCollege of Allied Health Sciences, University of CincinnatiAbstract This study evaluates carcinogenic health risks and groundwater quality in lead–zinc mining areas of Ebonyi State, Nigeria, using machine learning to predict contamination levels, providing crucial insights for public health and environmental management in a region heavily impacted by mining activities. A systematic and purposive sampling approach was employed to collect 42 groundwater samples from various sources across the study area and analyzed using standard methods. Machine learning was preferred over traditional methods for its ability to capture complex, non-linear relationships in the data, allowing for more accurate predictions of contamination levels and health risks, which is crucial for targeted monitoring and efficient groundwater management in mining-impacted areas. The pH levels ranged from 3.64 to 7.36, with a mean of 5.53, indicating significant acidity below the World Health Organization (WHO) standard. Heavy metal concentrations were alarming: cadmium ranged from 0.06 to 0.69 (mean 0.37), copper from 0.03 to 2.43 (mean 1.18), nickel from 0.03 to 1.81 (mean 0.85), mercury from 0.16 to 2.19 (mean 1.24), and arsenic from 0.09 to 4.15 (mean 1.94). All exceeded WHO limits, posing severe health risks. The correlation analysis of heavy metals identified a statistically significant negative relationship between Cadmium (Cd) and Chromium (Cr) (r = − 0.317, p = 0.041), and a marginally significant negative correlation between Mercury (Hg) and Selenium (Se) (r = − 0.273, p = 0.08). Most other correlations, including those between pH and metals such as Copper (Cu) and Zinc (Zn), were not statistically significant, indicating weak or non-existent linear relationships among these variables. The Principal Component Analysis (PCA) results suggest that groundwater quality in the mining area is influenced by multiple factors, including pH levels and the presence of various heavy metals. The Contamination Factor (CF) and Heavy Metal Evaluation Index (HEI) highlight severe contamination, particularly with Hg. The Pollution Load Index (PLI), Water Quality Index (WQI), and Heavy Metal Pollution Index (HPI), further demonstrate the widespread nature of the pollution, with numerous locations exhibiting significantly degraded water quality. The Geo-accumulation index (Igeo) and Nemerow Pollution Index (NPI) indicate moderate yet persistent contamination, reflecting the ongoing impact of mining activities throughout the region. Carcinogenic risk assessments highlighted greater risks for children, with chromium and arsenic levels reaching up to 0.36 mg/L and 0.42 mg/L, respectively. Specific sites, such as GW12, GW23, and GW30, were identified as high-risk areas. The Artificial Neural Networks, Support Vector Machines, and Multi-Linear Regression (MLR) models effectively predicted pollution indices, with MLR showing the most reliable performance (perfect R2 value). Seven predictive models were generated for key contamination metrics, offering valuable insights for future monitoring and management. The novelty of this study lies in its application of machine learning models to accurately predict contamination levels in groundwater around lead–zinc mining areas in Ebonyi State, Nigeria, providing a data-driven approach to enhance environmental monitoring and public health protection in a highly impacted region.https://doi.org/10.1007/s44290-024-00134-3GroundwaterArtificial neural networkPollutionSupport vector machineMulti-linear regressionRisk assessment |
| spellingShingle | Obinna Chigoziem Akakuru Moses Oghenenyoreme Eyankware Godwin O. Aigbadon Ayatu Ojonugwa Usman Alexander Iheanyi Opara Kizito Ojochenemi Musa Micheal Akaninyene Okon Okechukwu Pius Aghamelu Gabriel Ehriga Odesa Ifeyinwa Juliana Ofoh Annabel U. Obinna-Akakuru Carcinogenic health risks and water quality assessment of groundwater around lead–zinc mining areas of Ebonyi state Nigeria: a data-driven machine learning approach Discover Civil Engineering Groundwater Artificial neural network Pollution Support vector machine Multi-linear regression Risk assessment |
| title | Carcinogenic health risks and water quality assessment of groundwater around lead–zinc mining areas of Ebonyi state Nigeria: a data-driven machine learning approach |
| title_full | Carcinogenic health risks and water quality assessment of groundwater around lead–zinc mining areas of Ebonyi state Nigeria: a data-driven machine learning approach |
| title_fullStr | Carcinogenic health risks and water quality assessment of groundwater around lead–zinc mining areas of Ebonyi state Nigeria: a data-driven machine learning approach |
| title_full_unstemmed | Carcinogenic health risks and water quality assessment of groundwater around lead–zinc mining areas of Ebonyi state Nigeria: a data-driven machine learning approach |
| title_short | Carcinogenic health risks and water quality assessment of groundwater around lead–zinc mining areas of Ebonyi state Nigeria: a data-driven machine learning approach |
| title_sort | carcinogenic health risks and water quality assessment of groundwater around lead zinc mining areas of ebonyi state nigeria a data driven machine learning approach |
| topic | Groundwater Artificial neural network Pollution Support vector machine Multi-linear regression Risk assessment |
| url | https://doi.org/10.1007/s44290-024-00134-3 |
| work_keys_str_mv | AT obinnachigoziemakakuru carcinogenichealthrisksandwaterqualityassessmentofgroundwateraroundleadzincminingareasofebonyistatenigeriaadatadrivenmachinelearningapproach AT mosesoghenenyoremeeyankware carcinogenichealthrisksandwaterqualityassessmentofgroundwateraroundleadzincminingareasofebonyistatenigeriaadatadrivenmachinelearningapproach AT godwinoaigbadon carcinogenichealthrisksandwaterqualityassessmentofgroundwateraroundleadzincminingareasofebonyistatenigeriaadatadrivenmachinelearningapproach AT ayatuojonugwausman carcinogenichealthrisksandwaterqualityassessmentofgroundwateraroundleadzincminingareasofebonyistatenigeriaadatadrivenmachinelearningapproach AT alexanderiheanyiopara carcinogenichealthrisksandwaterqualityassessmentofgroundwateraroundleadzincminingareasofebonyistatenigeriaadatadrivenmachinelearningapproach AT kizitoojochenemimusa carcinogenichealthrisksandwaterqualityassessmentofgroundwateraroundleadzincminingareasofebonyistatenigeriaadatadrivenmachinelearningapproach AT michealakaninyeneokon carcinogenichealthrisksandwaterqualityassessmentofgroundwateraroundleadzincminingareasofebonyistatenigeriaadatadrivenmachinelearningapproach AT okechukwupiusaghamelu carcinogenichealthrisksandwaterqualityassessmentofgroundwateraroundleadzincminingareasofebonyistatenigeriaadatadrivenmachinelearningapproach AT gabrielehrigaodesa carcinogenichealthrisksandwaterqualityassessmentofgroundwateraroundleadzincminingareasofebonyistatenigeriaadatadrivenmachinelearningapproach AT ifeyinwajulianaofoh carcinogenichealthrisksandwaterqualityassessmentofgroundwateraroundleadzincminingareasofebonyistatenigeriaadatadrivenmachinelearningapproach AT annabeluobinnaakakuru carcinogenichealthrisksandwaterqualityassessmentofgroundwateraroundleadzincminingareasofebonyistatenigeriaadatadrivenmachinelearningapproach |