Statistical and probabilistic modeling of groundwater contamination: A data-driven uncertainty and risk assessment in Barind Tract, Rajshahi, Bangladesh

• 244 groundwater samples analyzed in wet and dry seasons to assess pollution severity. • Potentially toxic elements were evaluated using indices, statistics, and risk models. • Monte Carlo simulation showed children face higher health risks, mainly in dry season. • PCA, HCA, ANOVA, and correlation analyses identified contamination sources and trends. Groundwater contamination by potentially toxic elements (PTEs) poses serious health risks in Bangladesh, yet limited studies have addressed their seasonal and spatial variability. In the Barind Tract, Rajshahi, groundwater is a critical resource for drinking and irrigation, emphasizing the need for a comprehensive evaluation of contamination patterns and health implications. A total of 244 groundwater samples (122 per season) were analyzed for As, Al, Cu, Mn, Cr, and B using Atomic Absorption Spectrophotometry and VIS Spectrophotometry. Contamination was assessed using Metal Evaluation Index (MEI), Nemerow Pollution Index (NPI), Contamination Degree (CD), and Poseidon Index (PoS). Statistical analyses included ANOVA, effect size (η², f²), PCA, and HCA. Monte Carlo simulation (10,000 iterations, 95% CI) estimated probabilistic health risks for adults and children. Boron (1.034 mg L⁻¹) and manganese (0.824 mg L⁻¹) dominated during the wet season, while manganese (0.735 mg L⁻¹) and aluminum (0.605 mg L⁻¹) were highest in the dry season. MEI, NPI, and CD indices indicated significant contamination in Sirajganj, Chapai Nawabganj, and Natore, with Chapai Nawabganj (PoS 9.01) and Natore (PoS 8.223) showing extensive groundwater degradation. Statistical analyses confirmed significant seasonal variations, with aluminum, copper, chromium, and boron showing the most fluctuation due to hydrological and geochemical changes. Health risk assessments identified 64 samples unsafe for children and 32 for adults, with Monte Carlo simulations indicating hazard indices reaching up to 4.68 for children in certain hotspots. Pronounced seasonal variations in groundwater contamination underscore the necessity of targeted, season-specific monitoring and mitigation measures. Tailored interventions are essential to safeguard public health, especially for children who face higher exposure risks. The findings offer critical insights for policy-makers and stakeholders, promoting sustainable groundwater management in the Barind Tract.