Key points are not available for this paper at this time.
Abstract Fluvial flooding is characterized by flood discharges, which cause overflowing of the river banks and endangering the adjacent floodplains. In this study, we studied the 3.1 km section of the Gidra River (western Slovakia), which belongs to critical river sections prone to fluvial flooding. We aimed to compare two approaches for fluvial flood hazard mapping. Hydraulic approach was performed with the use of MIKE + model while the GIS-based approach was carried out using the Height Above the Nearest Drainage (HAND) model. The research question was whether the HAND model is suitable for flood inundation modeling and comparable to hydraulic modeling in the specific valley type of terrain. To determine the peak discharges for 100 and 1000 return periods, we used the rainfall-runoff procedure via the STORAGE and COSMO4SUB models. Accuracy of HAND models, as respect to the hydraulic MIKE + models, was assessed with the use of the following metrics: Critical Success Index, F1-score, Bias, Precision, Recall, and Overall Accuracy. Based on the results, the HAND model underestimated the hydraulically modeled flood extents in case of both flood scenarios with Bias values of 0.84 (Q 1000 ) and 0.87 (Q 100 ). The Overall Accuracy resulted in values of 0.93 (Q 1000 ) and 0.94 (Q 100 ). We concluded that the HAND-based modeled flood extents for the studied river section had an acceptable accuracy in case of both flood scenarios. Focusing on the river located in a valley type of terrain, the usage of the HAND model is mainly for rapid and preliminary fluvial flood hazard mapping due to lower data and computational requirements in comparison to hydraulic models.
Vojtek et al. (Wed,) studied this question.