Xxxi Iahr Congress 3573 Resolution Issues of Elevation Data during Inundation Modeling of River Floods

A case study of the Eskilstuna River in Sweden is presented. This study is carried out within the project KRIS-GIS, a Swedish initiative of handling crisis situations, including flooding. The purpose is to show how different resolutions in input elevation data affect the resulting inundation maps. Terrain elevation points at the sides of the river were gathered from an airborne laser altimetry survey, and river bed elevations were gathered from an echosounding survey. The terrain model was constructed in ArcView GIS as a triangulated irregular network (TIN), which served as the base for all later modeling. The hydraulic modeling was done as one-dimensional steady flow in HEC-RAS flow routing software. High-resolution elevation data resulted in better inundation delineation than did lowresolution elevation data. If the mean water discharge was used in the modeling and if the river is narrow, a low resolution could even lead to that the river itself would not be marked as inundated. At high water discharges, the river was usually inundated, but there was great uncertainty if the riparian areas really would be flooded or not. With steep side slopes, the delineation of inundation becomes more certain, while at gentler side slopes, the flow is distributed on a larger surface with a risk that the raster cells will be incorrectly marked regarding inundation. Finally, the use of high-resolution elevation data compared with lowresolution data, makes estimates of friction factor, Manning’s n, relatively more important for correct results in inundation studies.