A novel framework for groundwater quality evaluation in industrial zone using unsupervised machine learning methods.

Area of Science:

• Environmental Science
• Hydrogeology
• Geochemistry

Background:

• Rapid population growth, urbanization, and industrialization pose significant threats to freshwater ecosystems.
• Assessing groundwater quality in industrial zones is crucial for sustainable water resource management.

Purpose of the Study:

• To assess groundwater quality in a South Indian industrial zone using integrated methods.
• To identify key factors influencing groundwater chemistry and human health risks.

Main Methods:

• Collected 55 groundwater samples across seasons, considering availability, population density, and industrial activity.
• Employed unsupervised machine learning (PCA, HCA, k-means), groundwater pollution index (GPI), entropy water quality index (EWQI), and human health risk assessment.
• Utilized Piper Trilinear and Gibbs diagrams to analyze hydrogeochemical processes.

Main Results:

• Calcium-chloride and mixed calcium-magnesium-chloride water types dominate.
• Evaporation, water-rock interaction, and anthropogenic activities are major drivers of groundwater chemistry.
• Nitrate is identified as the primary parameter affecting groundwater sustainability and human health.

Conclusions:

• Integrated assessment reveals water-rock interaction, evaporation, and agricultural practices (fertilizer use) significantly impact groundwater quality.
• Machine learning techniques effectively delineate groundwater characteristics and influencing factors.
• Findings provide critical insights for enhancing water resource management strategies in the study region.