The ability to accurately model and predict environmental conditions is essential for tackling climate-related health and economic challenges, particularly in vulnerable regions like Hampton Roads, VA. As climate change poses significant risks, it is crucial to develop neighborhood-level weather forecasts that can aid in enhancing community adaptation and resilience. Traditional weather models have limitations, but new developments like the WRF-LES (Large Eddy Simulation) allow for much finer resolution, potentially predicting weather at scales of less than 200 meters. The integration of AI and sophisticated modeling is essential for localized forecasting, enabling the connection between weather patterns and geospatial attributes – including health and economic indicators – to inform effective solutions, and to support improved public health and economic resilience.
This project aims to focus on southeastern coastal Virginia, deploying high-resolution weather modeling integrated with air quality predictions for pollutants such as PM2.5 and nitrogen oxides.
Through the AI-driven integration of weather forecasting with health and environmental analyses, this project aims to inform strategies that promote economic resilience and to strengthen support of communities at risk.
With partnerships from institutions like the National Center for Atmospheric Research, the initiative will address long-term climate-health concerns and explore other applications, such as flash flood predictions. The project’s outcomes are relevant to vital sectors including offshore wind farm operations, port activities, and power grid management.
AI is central to this endeavor, enabling the fusion of complex weather models with dispersed geospatial and demographic data to create impactful predictions. AI also assists in aligning model predictions with actual measurements and enhancing the efficiency of weather modeling processes - areas of great interest for companies such as NVIDIA, Google, and Microsoft. Current work by the School of Data Science and Drs. Maryam Golbazi and Heather Richter at the Joint Institute for Advanced Computing for Environmental Studies includes developing a prototype WRF-LES model for urban heat island predictions, with additional research planned for broader applications.
Frank Liu, Ph.D.
Director, ±¬ÁϹÏ’s School of Data Science; Centennial Professor, Computer Science; ±¬ÁϹÏ