Deciphering geochemical fingerprints and health implications of groundwater fluoride contamination in mica mining regions using machine learning tactics.

The contribution of mica mining activities to fluoride (F-) contamination in groundwater has been chased in this study. For the purpose, groundwater samples (n = 40, replicated thrice) were collected during the post-monsoons (September-October) from a mica mining area in the Tisri block of Giridih district, Jharkhand. The study has employed a synergy of classical aquifer chemistry, statistical approaches, different indices, Self-Organising Maps (SOM), and Sobol sensitivity index (SSI) to unveil the underlying aquifer chemistry, identify the impacts of mining activities on groundwater quality and its associated health hazard. Fluoride levels varied from 0.34 to 2.8 ppm, with 40% of samples exceeding the World Health Organization's permissible limit (1.5 ppm). Physicochemical analysis revealed significant differences in electrical conductivity (EC), total dissolved solids (TDS), total hardness (TH) and major ion concentrations (Na+, HCO3-, Ca2+) between fluoride-contaminated (FC) and fluoride-uncontaminated (FU) groups. Higher Na+ and HCO3- associated with F- contaminated samples, were indicative of silicate weathering and carbonate dissolution as primary geogenic sources for this ion. Health risk assessment (HRA) revealed hazard quotient (HQ) values exceeding unity, indicating non-carcinogenic risks, particularly for children in most samples from group FC. The mean Water Quality Index (WQI) of FC group (156.76 ± 7.30) was significantly higher (p < 0.05) than group FU indicating of its unsuitability. SOM could accurately (80%) predict presence of fluoride in water samples based on other major ions. Sobol sensitivity analysis successfully identified fluoride concentration and body weight as most impactful parameters affecting human health. The integration of advanced modelling techniques and geospatial analysis as Inverse Distance Weightage (IDW) maps has provided a robust framework for ongoing groundwater quality monitoring in mining-affected regions and can help proactive intervention in risk-prone areas. Overall, this comprehensive study takes us a step ahead towards ensuring safe drinking water access for the global community.

From fuzzy-TOPSIS to machine learning: A holistic approach to understanding groundwater fluoride contamination.

Fluoride (F-) contamination of groundwater is a prevalent environmental issue threatening public health worldwide and in India. This study targets an investigation into spatial distribution and contamination sources of fluoride in Dhanbad, India, to help develop tailored mitigation strategies. A triad of Multi Criteria Decision Making (MCDM) models (Fuzzy-TOPSIS), machine learning algorithms {logistic regression (LR), classification and regression tree (CART), Random Forest (RF)}, and classical methods has been undertaken here. Groundwater samples (n = 283) were collected for the purpose. Based on permissible limit (1.5 ppm) of fluoride in drinking water as set by the World Health Organization, samples were categorized as Unsafe (n = 67) and Safe (n = 216) groups. Mean fluoride concentration in Safe (0.63 ± 0.02 ppm) and Unsafe (3.69 ± 0.3 ppm) groups differed significantly (t-value = -10.04, p < 0.05). Physicochemical parameters (pH, electrical conductivity, total dissolved solids, total hardness, NO3-, HCO3-, SO42-, Cl-, Ca2+, Mg2+, K+, Na+ and F-) were recorded from samples of each group. The samples from 'Unsafe group' showed alkaline pH, the abundance of Na+ and HCO3- ions, prolonged rock water interaction in the aquifer, silicate weathering, carbonate dissolution, lack of Ca2+ and calcite precipitation which together facilitated the F- abundance. Aspatial distribution map of F- contamination was created, pinpointing the "contaminated pockets." Fuzzy- TOPSIS identified that samples from group Safe were closer to the ideal solution. Among these models, the LR proved superior, achieving the highest AUC score of 95.6 % compared to RF (91.3 %) followed by CART (69.4 %). This study successfully identified the primary contributors to F- contamination in groundwater and the developed models can help predicting fluoride contamination in other areas. The combination of different methodologies (Fuzzy-TOPSIS, machine learning algorithms, and classical methods) results in a synergistic effect where the strengths of each approach compensate for the limitations of the other.