As the Gulf Coast heads into the most active stretch of the Atlantic hurricane season - August through September - forecasters warn the region could face heightened storm activity this year, fueled by warm ocean waters and a changing climate. Rice University researchers are at the forefront of using artificial intelligence (AI) to address these urgent climate challenges. The Advancing AI for Climate Risk and Urban Resilience (AI4UrbanResilience) research group, led by civil and environmental engineering scholar James Doss-Gollin and supported by the Ken Kennedy Institute, unites climate scientists, engineers and AI experts to develop transformative solutions for:
- Extreme weather modeling and disaster response
- Climate hazard prediction and risk assessment
- Flood modeling and mitigation strategies
- Urban energy and transportation resilience
While traditional hazard models offer valuable insights, they are often limited in resolution, efficiency and adaptability to different regions. AI4UrbanResilience integrates AI and machine learning with physics-based models to produce open-source, high-resolution and computationally efficient tools for managing complex, interconnected systems under extreme weather stress. Key focus areas include:
- Synthetic Hazard Generation: Producing large AI-generated datasets of realistic synthetic weather patterns for urban downscaling.
- Infrastructure Systems Response: Assessing hazard impacts on critical systems with AI- and optimization-enhanced methods.
- Multiscale Earth Observation and Data Assimilation: Merging diverse datasets to improve model accuracy and enable real-time initialization.
- Trustworthiness and Validation: Ensuring transparency and reliability of physics-informed AI through rigorous evaluation.
This work aligns with Rice's initiative to generate sustainable futures, positioning the university as a nexus for AI-powered climate adaptation. By connecting domain experts with AI researchers, the AI4UrbanResilience group aims to help communities in Houston and worldwide prepare for, withstand and recover from climate-driven disasters.
