Projects

Combining physically-based hydrology with machine learning to improve edge-of-field and watershed-scale runoff and risk prediction.

Coupling agent-based models of farmer decision-making with groundwater systems to study irrigation behavior and water security.

Agent-based models coupled with nutrient-load modeling to simulate how farmers adopt best management practices under different policy and information scenarios.

Using explainable AI to identify the minimal set of solutes that preserves hydrochemical source dynamics across diverse watersheds.

An open inventory of U.S. coastal pump stations across 55 cities for compound-flood-risk and energy-intensity analysis, aimed at a published dataset.

Coupling the 1D EPA-SWMM drainage network with the 2D Synxflow surface model for high-fidelity urban inundation mapping.

High-resolution groundwater modeling of tidal propagation and nuisance flooding beneath Lewes, Delaware.

A deep-learning model that detects non-stationarity in coastal storm surge and predicts extreme events across 208 US tide-gauge stations.

Studying how coastal communities perceive flood hazards and integrating local knowledge into resilience and hazard-mitigation planning.

A variable-density groundwater model of tidal flow and salinity transport in a coastal aquifer.

A private, locally-hosted multi-tenant LLM + RAG assistant that answers hydrologic-modeling questions (WRF-Hydro, SCHISM, BMI, PyMT, MPI, GIS) at postdoc-tier depth with citations.

CNN–LSTM surrogate models that emulate WRF-Hydro streamflow to accelerate scenario testing and explainable hydrologic analysis.

Geospatial analysis, hydrologic modeling, and machine learning to quantify how land-use and climate change drive non-point source pollution risk.

A high-resolution digital twin of Lewes, Delaware coupling urban drainage with surface topography to simulate sunny-day tidal flooding.

Generating an improved gridded precipitation dataset for Delaware to better capture flooding and water-quality risk under a changing climate.

Evaluating urban rain-garden performance through an integrated social–ecological–hydrological framework.

Estimating extreme skew-surge probabilities from tide-gauge and atmospheric data, with nonstationarity detection and storm clustering.

A BMI wrapper for WRF-Hydro that couples it one-way with the SCHISM coastal-ocean model via PyMT for compound-flooding simulation — including the first MPI-aware BMI processing element.

Diagnosing the street-to-pipe dynamics of compound rain and tailwater flooding in coastal combined sewer systems.

AWS SageMaker ML pipelines using National Water Model data to forecast surface-water runoff for the Great Lakes Region, plus a production web app for real-time and historical runoff-risk analysis.

Modeling the impacts of intensive pumping on regional groundwater and coastal aquifers, and evaluating seawater-intrusion control measures.

Analytical and numerical models of seawater intrusion under different coastal land-reclamation scenarios.

A mobile-first PWA that lets Low Impact Development coordinators in Washington DC run door-to-door field campaigns with real-time cross-team sync on a shared map.

Using FUNWAVE-TVD to study how intense rainfall affects nearshore wave dynamics, runup, and wetting patterns.

A phase-aware scheduling wrapper for PESTPP-GLM that reallocates idle compute threads during distributed hydrologic model calibration.

Developing river water-quality management policies through water-quality analysis, environmental modeling, and policy-oriented assessment.

Simulating surface water and groundwater resources from local to catchment scales, and the interactions between terrestrial water systems.

Technical assessment of Tehran's stormwater management master plan with the Tehran Urban Planning & Research Center.

Simulating watershed runoff and streamflow in coastal watersheds to study rainfall–runoff response and inland contributions to compound flooding.