41 - Ensemble Flood Inundation Mapping under Changing Climate Conditions: A Case Study Driven by Downscaled CMIP6 Projections

Climate change poses a substantial threat to critical infrastructures due to the heightened vulnerability to floods. In this study, we introduce an open-source, high-resolution hierarchical modeling framework capable of generating ensemble flood inundation maps under changing climate conditions for sites and watersheds of interest. The framework integrates climate, hydrologic, and a GPU-accelerated two-dimensional hydrodynamic model – TRITON (https://triton.ornl.gov/) that can accurately simulate both fluvial and pluvial inundation events. Driven by downscaled and bias-corrected hydroclimate projections from the Coupled Model Intercomparison Project phase 6, an ensemble of annual maximum flood events is simulated and formed the basis for constructing high-resolution probabilistic inundation maps. These maps enable the identification of areas of potential flood risk under historic and future climate conditions that cannot be revealed by the conventional, deterministic methods. Further, we use the Hazus-MH model in conjunction with probabilistic inundation maps to conduct a flood vulnerability assessment, which informs effective flood mitigation measures. We demonstrate the application of the framework at selected U.S. Air Force Bases where the smoother topography generally represents a major challenge for inundation modeling. The proposed framework has the potential to be scalable beyond the selected sites and can support broader climate change-driven vulnerability assessments. We anticipate this new capability will equip flood risk managers with robust and process-based insights enabling informed decision-making to safeguard our critical infrastructures against climate-induced flood risks.