Assessing Climate Risks at Urban Scales Using Large-Scale Hydrodynamic Models (invited)

Floods have devastating consequences for urban communities. Increasingly common heavy precipitation events – like cloudbursts – have led to recurrent damage and disruption, impacting the bottom lines of homeowners, businesses, and local and state governments. Widespread recognition of the potential for even more extreme storms in the future has led to a growing demand for climate risk estimates at the property scale. Yet questions remain as to how to best use climate model output to inform flood hazard models and ultimately, estimate economic impacts. While large scale hydrodynamic models have grown increasingly cost efficient – requiring fewer computational resources to model large regions or a greater number of storm events – model performance is highly variable, even across a single urban area.

In this presentation, we discuss recent work towards developing stochastic approaches to flood hazard estimation and ultimately, climate risk assessment. Using Greater Houston, Texas USA as a case study, we apply the large-scale hydrodynamic model SFINCS to assess flood risk from pluvial and fluvial flooding at the parcel-scale under current and future conditions. The model is validated against recent flood events using gage data, high water marks, and observed flood damage. We use traditional approaches based on NOAA Atlas 14 precipitation estimates to quantify current flood hazards and scale them based on GEV distributions of 24-hour precipitation derived from CMIP 6 to derive a distribution of water levels at each property under future climate conditions. We then translate modeled water levels into damages using the standard USACE depth-damage curves provided in the HAZUS model, considering uncertainty in the hydrodynamic model performance and in the precipitation estimates. Our results demonstrate the range inherent in climate hazard assessment, especially at the property level. We caution against using deterministic estimates of flood hazards to inform policy, development decisions, or property valuation and provide recommendations for paths toward more robust estimates of climate risk.