Flood Projections Across Scales: A Local, Regional, and Continental Approach

Flooding is one of the most damaging natural hazards worldwide, with increasing risk driven by climate??'induced changes and exposure in vulnerable communities. Meeting emerging design and planning challenges requires scalable flood modeling tools that incorporate future climate conditions while quantifying the uncertainty inherent in these projections. Typically, national flood information often lacks sufficient spatial detail, does not reflect nonstationary climate influences, and provides deterministic outputs that obscure uncertainty that hinder effective and equitable risk management.

To bridge this gap, this seminar introduces an integrated modeling framework designed to quantify how flood hazard and risk may evolve under warming across multiple spatial scales. The framework combines high??'resolution rainfall??"runoff modeling, two??'dimensional hydraulic simulations, and pseudo global warming scenarios to generate consistent flood estimates across the contiguous United States. A regional application in Iowa further evaluates how uncertainties in climate model forcing propagate through hydrologic and hydraulic processes, revealing nonlinear sensitivities. This has significant implications for engineering risk assessment, scenario selection, and adaptation planning.

An analysis for the northeastern United States applies large ensembles of flood inundation simulations to quantify interval variability and develop probabilistic bounds on potential future flood inundation. Collectively, these efforts demonstrate how physically based models, modern climate datasets, and uncertainty??'aware methods can support resilient infrastructure design, hazard mitigation, and community??'level adaptation in a changing climate.

Speaker: Alexander Michalek, Princeton University

Attend in person or watch online (see weblink)