Water Resources Research Act Program

Details for Project ID CT-2025-Witharana

Predictive Modelling of Domestic Well Water Quality Using Artificial Intelligence

Institute: Connecticut
Year Established: 2025 Start Date: 2026-01-01 End Date: 2026-12-31
Total Federal Funds: $30,000 Total Non-Federal Funds: $30,000

Principal Investigators: Chandi Witharana

Project Summary: We propose to develop a novel AI model (Well Water Quality (W2Q) Transformer) for private well water quality prediction. To the best of our knowledge, no prior study has employed Transformers for groundwater quality prediction tasks. We will use accessible private well water quality datasets and an array of explanatory variables extracted from multi-modal/ multi-source environmental, hydrological, geological, land use, soil, terrain, demographic datasets in the AI model development and training. Our specific study objectives are to; 1) Design a Transformer-based model (W2Q Transformer) to predict private well water quality, integrating spatiotemporal and contextual data, 2) Compare proposed W2Q Transformer's prediction capabilities against conventional statistical and machine learning algorithms, and 3) Identify key spatiotemporal drivers of domestic well water quality variation in the state.