Skip to content Where Legends Are Made
Cooperative Institute for Research to Operations in Hydrology

CIROH Training and Developers Conference 2024 Abstracts

Authors: Juli Scamardo, Scott Lawson, Rebecca Diehl, Kristen Underwood, Beverley Wemple – University of Vermont

Title: Incorporating Floodplain Topographic Features to Improve Channel Routing

Abstract: River floodplains can attenuate overbank flows, where the rate and magnitude of attenuation is in part dependent on floodplain topography. Representing the forms and variability in floodplain topography within routing models can therefore influence the prediction of peak flow magnitudes and timing throughout a watershed. However, incorporating high-resolution topography from digital elevation models (DEMs) into large-scale routing models requires significant amounts of data. To balance the need for floodplain representation with computational demand, we developed a method to identify and extract hydraulically relevant floodplain features from high-resolution height above nearest drainage (HAND) layers. The basis of this method relies on identifying Energy Dissipation Zones (EDZs) within floodplain topographic cross-sections and defining key features of the EDZ that influence peak flood discharge and flood wave celerity. The method was tested by extracting floodplain features for over 3,000 NHDplus HR reaches across six topographically diverse basins in Vermont, USA. Feature extraction was paired with a reach-based Muskingum-Cunge routing analysis to describe hydraulic behavior and attenuation capacity of each floodplain. For unconfined, low-gradient settings, the KMedoids clustering algorithm was used to identify six unique floodplain classes based on their extracted topographic features. Class routing behavior was then evaluated by the average celerity and attenuation capacity of the reaches within it. The hydraulic behavior of each floodplain class was compared to that of a two-stage channel using the National Water Model (NWM) Retrospective 2.1 output for multiple high flow events across the basins. The incorporation of hydraulically relevant floodplain features into modeled cross-sections could improve predictions of the timing and magnitude of flood peaks, particularly for watersheds containing a greater percentage of wide and/or laterally sloped floodplains.