Flood analysis supported by low‐cost geometric modelling

Abstract

[EN] Today, it is increasingly clear that non-stationarity hydrological and hydraulic variables and processes are occurring largely because of global warming. Accordingly, extreme hydrological events are becoming more common over time, and their effects are creating greater negative impacts on the environment (fluvial geomorphology and floodplains) and society (flood damage). Given this situation, the implementation of adaptation-mitigation measures is vital, as well as an increased knowledge of the interaction between water and physical environments. In the binominal water-terrain, having a reliable digital elevation model (DEM) is essential because of its important influence on fluvial modelling. However, this is frequently a technical-economic problem. The aim of this paper is first to evaluate the compatibility between hydraulics and geometrics for fluvial applications and second to determine the quality of a novel DEM by robust estimators. This was obtained through the photogrammetric processing of digital aerial images acquired from a low-cost camera mounted on an alternative aerial platform. Flood modelling and hydraulic parameters were obtained with the assistance of photogrammetric DEM (mesh size: 0.15 m, vertical accuracy: 0.102±0.081 m, point density: ≈40 point/m2). Finally, our other goal is to develop a comparative analysis between light detection and ranging and digital photogrammetry on-demand. This comparison revealed that flood modelling by photogrammetric DEM was considerably more detailed than that by light detection and ranging-DEM, mainly because of higher point density and vertical accuracy. Consequently, flood analysis assisted by this novel geometric modelling approach qualifies as a reliable and competitive approach.