Effect of topographic data, geometric configuration and modeling approach on flood inundation mapping

Technological aspects of producing, delivering and updating of flood hazard maps in the US have has gone through a revolutionary change through Federal Emergency Management Agency’s Map Modernization program. In addition, the use of topographic information derived from Light Detection and Ranging (LIDAR) is enabling creation of relatively more accurate flood inundation maps. However, LIDAR is not available for the entire United States. Even for areas, where LIDAR data are available, the effect of other factors such as cross-section configuration in one-dimensional (1D) models, mesh resolution in two-dimensional models (2D), representation of river bathymetry, and modeling approach is not well studied or documented. The objective of this paper is to address some of these issues by comparing newly developed flood inundation maps from LIDAR data to maps that are developed using different topography, geometric description and modeling approach. The methodology involves use of six topographic datasets with different horizontal resolutions, vertical accuracies and bathymetry details. Each topographic dataset is used to create a flood inundation map for twelve different cross-section configurations using 1D HEC-RAS model, and two mesh resolutions using 2D FESWMS model. Comparison of resulting maps for two study areas (Strouds Creek in North Carolina and Brazos River in Texas) show that the flood inundation area reduces with improved horizontal resolution and vertical accuracy in the topographic data. This reduction is further enhanced by incorporating river bathymetry in topography data. Overall, the inundation extent predicted by FESWMS is smaller compared to prediction from HEC-RAS for the study areas, and that the variations in the flood inundation maps arising from different factors are smaller in FESWMS compared to HEC-RAS.