Non-carcinogenic health risk assessment and predicting of pollution indexing of groundwater around Osisioma, Nigeria, using artificial neural networks and multi-linear modeling principles
Obinna Chigoziem Akakuru,Uzoma Benedict Njoku,Annabel Uchechukwu Obinna-Akakuru,Bernard E. B. Akudinobi,Philip Njoku Obasi,Godwin Okumagbe Aigbadon,Uzochi Bright Onyeanwuna
DOI: https://doi.org/10.1007/s00477-023-02398-0
IF: 3.821
2023-03-11
Stochastic Environmental Research and Risk Assessment
Abstract:Non-carcinogenic health risk assessment and prediction of pollution indexing of groundwater around Osisioma, Nigeria, using artificial neural networks and multi-linear modeling principles has been done. Thirty groundwater samples were collected systematically and analyzed for organic and heavy metal pollutants. The results of the analysis showed that the heavy metals and organic pollutants contributed to the pollution of groundwater resources in the locality. 63.3% of the entire water samples had As above the WHO standard, same as Fe (60%), Cr (100%), Pb (56.7%), E (16.7%), X (13.3%), B (40%). Correlation matrix results indicated a weak correlation. For the Principal Component Analysis, PC1 showed that 60% of the entire variable had loadings, PC2 had 40%, PC3 had 30%, PC4 had 10% loadings of parameters within the study area, and that organic pollutants were major contributors to the loadings. The Contamination factor, Pollution load index, Metal pollution index, Geoaccumulation index, Potential ecological risk index, Elemental Contamination Index, and Overall Metal Contamination Index showed no significant pollution, whereas the Heavy Metal Evaluation Index, Pollution Index of Groundwater results showed the worrisome impact of the anthropogenic activities on the groundwater quality. Health risk assessment showed that children are more at risk than adults as it related to taking polluted with a Hazard Quotient and Hazard Index trend is Cr > As > T > E > m-X > o-X > B > Pb > Cu > Fe. This trend is the same for both children and adults. Seven mathematical models were generated for the prediction of pollution indices. Based on the results, this study recommends regular monitoring of groundwater resources and the integration of ANN and MLR modeling approaches for the prediction of pollution indices.
environmental sciences,engineering, environmental,water resources, civil,statistics & probability