Groundwater quality drivers in the drought-prone Thakurgaon District, Northwestern Bangladesh: An integrated fuzzy logic and statistical modeling approach

Abstract

Groundwater quality in the drought-prone Thakurgaon District, Northwestern Bangladesh, is deteriorating due to a combination of natural and anthropogenic factors. This study evaluates the key drivers of groundwater quality degradation by employing ecotoxicological risk indices, such as the Heavy Metal Pollution Index (HPI), Heavy Metal Evaluation Index (HEI), and Nemerow's Pollution Index (NPI). An innovative fuzzy logic approach is used to integrate these indices and reduce uncertainty, while Automatic Linear Modeling (ALM) predicts the primary impacts on the Fuzzy Groundwater Quality Index (FGWQI). Additionally, Monte Carlo simulations assess probabilistic health risks and sensitivity. Groundwater samples from 40 wells were analyzed for physicochemical parameters and heavy metal concentrations. The results show that 25 % of the samples are unsuitable for drinking, and 17.5 % are unfit for household use, based on HPI and HEI values. Fuzzy analysis reveals that 22.5 %, 47.5 %, and 30 % of the samples exhibit excellent, good, and poor quality, respectively. The overlay of FGWQI with Land Use/Land Cover (LULC) maps identifies areas with excellent groundwater quality in the southern parts of the region, while the northern areas suffer from poor quality due to overexploitation. One-way ANOVA indicates that rainfall, water discharge, and LULC significantly affect FGWQI. The ALM results highlight HEI (0.62) and HPI (0.38) as the main factors influencing FGWQI. Health risk analysis reveals elevated non-carcinogenic risks due to arsenic and lead ingestion, particularly for children. These findings emphasize the need for targeted policies and interventions to mitigate health risks and ensure the well-being of the community.