Quality of flood extents delineated by a non-hydrodynamic GIS tool

Abstract

Delineation of flood extents is an important task in various disciplines. Traditional sophisticated hydrodynamic or alternative non-hydrodynamic approaches can be used depending on the availability and quality of input data and, in particular, the purpose of the delineation. This work focuses on testing the accuracy of flooded areas computed by GIS-based non-hydrodynamic AIZM tool, which is undemanding in terms of input data and technological requirements. The flood delineation is carried out on the basis of flood height at a single point and its spreading along a stream line over a digital elevation model (DEM). In order to assess the tool's performance, flood inundations have been computed along 60 river reaches in the Czech Republic and compared with benchmarks inundations: (1) real flood events that occurred in 1997, 2002 and 2006, as mapped from aerial and satellite imagery and (2) flood hazard maps designed for flood events of 5-, 20- and 100-year return periods. The accuracy of the outputs is examined and interpreted in the context of the river reaches' specificities to identify generally favourable conditions as well as the key limitations of the model. The results outline that under certain conditions, the AIZM tool can be used as a reliable tool for flood extent delineation in data scarce areas (e.g. lacking hydrodynamic constants and channel bathymetry), for reconstruction of past flood events or for preliminary estimation of flooded areas.