My research focuses on developing and advancing numerical, statistical, and machine learning models to improve understanding of subsurface systems and to bridge the gap between scientific insight and real-world decision-making. I integrate field observations, laboratory experiments, and computational modeling into predictive frameworks that advance fundamental science while informing applications in energy storage, groundwater management, mineral exploration, and environmental protection.
Basin-Scale Frameworks and Subsurface Energy Storage
- Focus: Develop basin-scale hydrogeologic frameworks to guide the safe and efficient use of the subsurface for carbon capture and storage (CO₂), hydrogen, geothermal energy, brine management, mineral exploration, and groundwater resources, while ensuring the protection of underground sources of drinking water (USDWs).
- Approach: Integrate geologic structure, hydrostratigraphy, and geochemical datasets into coupled flow–transport–reactivity models. Use telescopic mesh refinement to bridge basin-scale dynamics with site-specific reservoir evaluations. Apply probabilistic geologic frameworks (e.g., TPROGS) to capture heterogeneity, quantify uncertainty, and assess risks of fluid migration, pressure interference, and mechanical integrity.
- Tools: TPROGS, MODFLOW family, MT3DMS, SEAWAT, TOUGH, coupled thermal–hydrological simulators, geomechanical models, geostatistical mapping, causal ML, and multi-objective optimization frameworks.
Reactive Transport Modeling
- Focus: Advance predictive understanding of multiphase, multicomponent transport and geochemical reactions in heterogeneous formations.
- Approach: Integrate field datasets and laboratory experiments (batch and column) with multiscale simulations to constrain reactivity and transport. Upscale processes from lab and site studies into regional-scale predictive models of contaminant fate and geochemical evolution.
- Tools: TOUGH, PFLOTRAN, MT3DMS, PHREEQC, and hybrid physics–ML frameworks to capture nonlinear reactivity and heterogeneity.
Hydrology and Climate Systems
- Focus: Quantify surface and ground water availability, hydrological resilience, and drought vulnerability under climate and land-use stressors.
- Approach: Combine distributed hydrological models with monitoring networks to evaluate surface–groundwater interactions and system stress. Incorporate large-scale climate teleconnections (ENSO, AMO, PDO) into predictive models to evaluate water resource variability.
- Tools: SWAT/SWAT-VSA, MODFLOW, drought indices (SPI, Palmer Drought Index), climate teleconnection analysis, and probabilistic forecasting frameworks.
Data Science and Model Advancement
- Focus: Push the frontier of model predictability, interpretability, and decision support by integrating physics-based hydrogeologic models with advanced AI/ML.
- Approach: Develop physics-informed machine learning, causal inference, and surrogate models to reduce uncertainty and improve risk assessments. Apply FAIR-compliant data platforms to deliver transparent, reproducible, and actionable science for decision-makers.
- Tools: LSTM, CNN, Random Forest, Bayesian networks, SHAP/LIME for explainability, Monte Carlo methods, and interactive dashboards.