Utility of Remote Sensed Topography to Support Flood Inundation Modelling Under Parametric and Rating Curve Uncertainty (PAP014897)

Código

PAP014897

Autores

Micah Mukungu Mukolwe, Kun Yan, Giuliano Di Baldassarre, Dimitri P. Solomatine

Tema

Land use and Floods, landslides and erosion

Resumo

Over the last few years, remote sensed data have gotten more readily available. This has created opportunities for synthesis and adoption to support flood modelling. In this context, an important recurring question is: what is the added value (and associated level of uncertainty) in flood modelling resulting from the use of these products? In particular, the recently released EU-DEM, which is a weighted fusion of SRTM (Shuttle Radar Topography Mission) and ASTER - GDEM (Advanced Space borne Thermal Emission and Reflection Radiometer - Global Digital Elevation Map) data, was compared to a relatively accurate LiDAR topography and the original SRTM dataset on a 98-km reach of the Po river, Italy. Rating curve uncertainty and model parameter uncertainty were explicitly taken into account as well as model structural uncertainty. Our results show an improvement in the 2D hydraulic model of the EU-DEM dataset compared to SRTM in terms of the spatial simulation of flood inundation patterns. These outcomes are discussed in view of the difficult question of handling uncertainty in model outputs faced by stakeholders and decision-makers. Our study provides a contribution to the scientific community, which is continuously evaluating different approaches and communication mechanisms to reduce ambiguity and increase utility of different sources of data in flood modelling.