Kavya Agrawal

Research Portfolio

Work So Far

Ongoing and completed research directions across groundwater contamination, geospatial AI, exposure assessment, and policy diagnostics.

Uranium Risk Mapping

India-scale machine learning framework for predicting uranium contamination probability and translating spatial risk into exposure and governance indicators.

Fluoride DL Benchmarking

Multi-resolution deep learning benchmarking to evaluate how spatial scale affects groundwater contamination model performance and variable sensitivity.

Arsenic Transferability

Temporal transferability, robustness testing, and interpretable AI for arsenic contamination prediction.

Current Ph.D. Research Framework

• Prediction and process understanding of groundwater contamination patterns
• Probabilistic transition regime identification
• Model validation and spatial reliability assessment
• Population exposure assessment across contamination probability classes
• Urban and peri-urban vulnerability analysis
• Governance diagnostics using MEI, GBSI, EII, and TSI
• Composite policy priority zoning for risk-based monitoring and intervention