Hydraulic Modeling of a River Network for Predicting Flood Inundation using HEC-RAS and GIS Models - A Case Study in Southern Virginia

Abstract

A flood inundation study is presented for a watershed located in south central Virginia. A HEC-RAS hydrodynamic model of the main river network was developed to assess the impact of a number of hydrologic events, including the Probable Maximum Flood (PMF), in the area of interest. The primary goal of the study was to transform discharge hydrographs produced by HEC-HMS, an event-based hydrologic model, into water surface elevations and flood inundation spatial extents. Initially, a river terrain model was constructed using data from publicly available sources and filed survey campaigns. HEC-GeoRAS and ArcGIS were used to document and integrate the considerable amount of data required for building the model. Then, a calibration process was performed using stage-discharge predictor curves. The HEC-RAS unsteady flow component was employed for routing the discharge hydrographs through the modeled river network. Flood inundation maps, as well as longitudinal water surface elevation and channel velocity profiles were generated for the study reaches. As part of the flood inundation study, an uncertainty quantification analysis was carried out on the boundary roughness of the floodplains. The objective was to measure the extent to which flood inundated areas, water surface elevations, and channel velocities were influenced by variations on this empirically-based model coefficient. Finally, the impact of various hydraulic characteristics of the modeled river on the sediment transport process is examined. This characterization is intended to provide a better understanding of a subsequent sediment transport modeling effort to be performed under severe flooding conditions.