Research

We are a mission-oriented research group addressing fundamental challenges in infrastructure resilience, risk assessment, and climate adaptation. Our research vision is to advance climate risk analysis and adaptive planning for the complex, interdependent infrastructure systems that underpin modern society.

To achieve this, we integrate Earth science, risk analysis, and probabilistic machine learning. While motivated by the downstream need for robust adaptation, our primary technical focus is on improving the upstream characterization of hazards and their interaction with complex systems.

Our work is organized around three interrelated questions:

How can we simulate nonstationary extremes using physics-informed AI?
How do these hazards propagate to create risk in critical infrastructure?
What strategies robustly meet competing objectives under deep uncertainty?

We explore these questions through four specific research themes:

Research Themes

Probabilistic Hydroclimate Hazard

We develop advanced statistical and machine learning methods, including hierarchical Bayesian models and generative diffusion models, to characterize the hazard associated with nonstationary hydroclimate extremes.

AI-Accelerated Flood Modeling

To translate rainfall into urban flood hazard at scale, we are developing physics-informed machine learning (e.g., Graph Neural Networks) to rapidly and accurately predict flood extent and depth.

Infrastructure and Systems Risk

We quantify the risks these hazards pose to infrastructure systems and the people who use them.

Adaptive Decision Making Under Uncertainty

We develop decision frameworks that identify robust and adaptive pathways for managing climate risks, enabling planners to balance multiple objectives under deep uncertainty.

Project Pages

Rice AI for Climate Risk and Urban Resilience cluster

Research Support

We would like to thank the following organizations for their research funding and/or in-kind support.

Active Support

Eleven Labs

Past Support