Authors: Seth Bryant
Title: Bridging Global Flood Models and Local Decision-Making with Automated 2 m DEM-Consistent Mapping
Presentation Type: Lightning Talk
Abstract: Operational FIM is now fast and scalable, but still too coarse to resolve the street-scale controls that determine real-world flood impacts. Here we present the results of a seed grant to develop a scalable machine-learning driven super-resolution CLI floodsr and demonstrate how it can be used to generate 1m resolution street-level flood depths from fathom’s global flood 3.0. The super-resolution approach acts as a lightweight post-processing enhancement for coarse inundation products, transforming them into sharper, terrain-informed flood maps while preserving the scalability required for operational flood forecasts. We show how floodsr packages this capability in an easy-to-install Python tool with clear documentation and hands-on tutorials, and how the regional pipeline extends the same approach to large-domain flood products. Together, these developments offer a faster, more accessible research-to-operations path for next-generation FIM.