LiDAR-Derived Flood-Inundation Maps for Real-Time Flood-Mapping Applications, Tar River Basin, North Carolina

Abstract

Flood-inundation maps were created for selected streamgage sites in the North Carolina Tar River basin. Light detection and ranging (LiDAR) data with a vertical accuracy of about 20 centimeters, provided by the Floodplain Mapping Information System of the North Carolina Floodplain Mapping Program, were processed to produce topographic data for the inundation maps. Bare-earth mass point LiDAR data were reprocessed into a digital elevation model with regularly spaced 1.5-meter by 1.5-meter cells. A tool was developed as part of this project to connect flow paths, or streams, that were inappropriately disconnected in the digital elevation model by such features as a bridge or road crossing. The Hydraulic Engineering Center-River Analysis System (HEC-RAS) model, developed by the U.S. Army Corps of Engineers, was used for hydraulic modeling at each of the study sites. Eleven individual hydraulic models were developed for the Tar River basin sites. Seven models were developed for reaches with a single gage, and four models were developed for reaches of the Tar River main stem that receive flow from major gaged tributaries, or reaches in which multiple gages were near one another. Combined, the Tar River hydraulic models included 272 kilometers of streams in the basin, including about 162 kilometers on the Tar River main stem. The hydraulic models were calibrated to the most current stage-discharge relations at 11 long-term streamgages where rating curves were available. Medium- to high-flow discharge measurements were made at some of the sites without rating curves, and high-water marks from Hurricanes Fran and Floyd were available for high-stage calibration. Simulated rating curves matched measured curves over the full range of flows. Differences between measured and simulated water levels for a specified flow were no more than 0.44 meter and typically were less. The calibrated models were used to generate a set of water-surface profiles for each of the 11 modeled reaches at 0.305-meter increments for water levels ranging from bankfull to approximately the highest recorded water level at the downstream-most gage in each modeled reach. Inundated areas were identified by subtracting the water-surface elevation in each 1.5-meter by 1.5-meter grid cell from the land-surface elevation in the cell through an automated routine that was developed to identify all inundated cells hydraulically connected to the cell at the downstream-most gage in the model domain. Inundation maps showing transportation networks and orthoimagery were prepared for display on the Internet. These maps also are linked to the U.S. Geological Survey North Carolina Water Science Center real-time streamflow website. Hence, a user can determine the near real-time stage and water-surface elevation at a U.S. Geological Survey streamgage site in the Tar River basin and link directly to the flood-inundation maps for a depiction of the estimated inundated area at the current water level. Although the flood-inundation maps represent distinct boundaries of inundated areas, some uncertainties are associated with these maps. These are uncertainties in the topographic data for the hydraulic model computational grid and inundation maps, effective friction values (Manning's n), model-validation data, and forecast hydrographs, if used. The Tar River flood-inundation maps were developed by using a steady-flow hydraulic model. This assumption clearly has less of an effect on inundation maps produced for low flows than for high flows when it typically takes more time to inundate areas. A flood in which water levels peak and fall slowly most likely will result in more inundation than a similar flood in which water levels peak and fall quickly. Limitations associated with the steady-flow assumption for hydraulic modeling vary from site to site. The one-dimensional modeling approach used in this study resulted in good agreement between measurements and simulations. T