Focus: Simulating and predicting extreme rainfall events using high-resolution numerical weather prediction (NWP) via the Weather Research and Forecasting (WRF) model.

Challenge: Managing the high computational cost of running physics-rich simulations while accurately capturing the complex spatial, temporal, and dynamical variability of extreme precipitation.

Approach: Evaluating optimal WRF physical configurations, analyzing rainfall time-series dynamics, and building AI/ML surrogate models that emulate high-resolution atmospheric simulations at a fraction of the computational expense.

Focus: High-accuracy solar irradiance estimation and multi-horizon forecasting to support grid decarbonization.

Challenge: Overcoming sparse ground stations and the spatial-temporal resolution limits of satellite retrievals under dynamic cloud cover.

Approach: Developing machine learning models that fuse satellite imagery, NWP simulations, and reanalysis data into a unified, hybrid prediction pipeline.