Impact of cross‐sectional orientation in one‐dimensional hydrodynamic modeling on flood inundation mapping

Abstract

AbstractFlood management activities require development of flood maps that depict the spatial and temporal extent of floods with the help of hydrodynamic models. Two‐dimensional (2‐D) hydrodynamic models are frequently employed for flood inundation modeling and mapping with the help of high‐end computational resources and high‐resolution terrain information such as LiDAR (light detection and ranging) data. However, LiDAR data are either unavailable or not freely accessible in many parts of the world, especially in nations belonging to Global South. Hence, one‐dimensional (1‐D) models are still in practice owing to their lesser computational cost and data requirement. Nonetheless, the successful application of a 1‐D model depends mainly on the representation of the natural river and floodplain geometry, which is the primary input in the form of discrete cross‐sections. The assumed flow directions on floodplains while orienting the cross‐sections in 1‐D models induce some uncertainty in the model simulations. This study aims to evaluate the variability in the model simulations caused by the cross‐sectional orientations. Flood simulations were performed using a 1‐D hydrodynamic model for six different cross‐sectional realizations for three river reaches, having distinct morphological and topographical characteristics, and were compared with the simulations of a 2‐D hydrodynamic model and available reference inundation maps. The study suggests that the simulations of flood inundation extent and maximum flow depth variation are influenced by the cross‐sectional orientation on flood plains in river reaches characterized by broad flood plains with complex local topographical variations. In contrast, for reaches with relatively less wide and complex terrains, 1‐D models can generate robust simulations of flood inundation extent and spatial variation in maximum flood depth (R2 and NSE greater than 0.89) for high flood events.