Leveraging High-resolution Topography to Advance Parsimonious Reach-scale Flood Inundation Mapping

Forecasting flood inundation requires accurate flood discharges, topography, and a means to map discharge to water level. In the continental United States, the National Water Model forecasts flooded areas using the Height Above Nearest Drainage (HAND) approach. HAND utilizes a Synthetic Rating Curve (SRC) and forecasted discharge to compute flood inundation within a river reach. Within the National Water Model, SRCs are currently developed using reach-averaged hydraulic geometry extracted from a 10-m Digital Elevation Model; however, the increasing availability of high-resolution topography and hydrodynamic modeling tools poses a unique opportunity to explore how the full complexity of channel morphology can be leveraged to assess and improve the HAND-SRC approach. We utilize high-resolution topo-bathymetry coupled with satellite images of flood inundation to calibrate 2D hydraulic models for three reaches of the South Fork Eel River in northern California. To assess the impact of reach-scale hydraulic variability on flooding, we construct an ensemble of rating curves based on high-resolution cross stream elevation transects spaced every 10-15 meters. We generate inundation extents using the ensemble of rating curves and HAND-SRCs and assess their performance against the calibrated model inundation extents. We find that a single rating curve can produce accurate reach-scale FIM and that the relatively poor performance of the HAND-SRC approach can be improved by accounting for inaccuracies within the SRC treatment of hydraulic geometry and flow resistance. Improving the HAND-SRC approach has the potential to reduce uncertainty in flood prediction and provide advanced warning of flood extent across the United States.