National Floodplain Mapping: Datasets and Methods – 160,000 km in 12 months

Abstract

This paper describes two projects requiring production of national floodplain maps for England and Wales -some 80,000 km of river. The novel solutions developed have brought together a national Digital Elevation Model (DEM), automatically-generated peak flow estimates at intervals along the watercourses and two alternative methods of calculating the outlines: normal depth calculation; and a purpose-built 2-dimensional raster-based floodplain model, JFLOW. The DEM was derived using Interferometric Synthetic Aperture Radar (IFSAR) techniques and has a vertical precision of ±0.5 m-1.0 m (RMSE) and a 5 m horizontal resolution. The flow estimates were derived by automating Flood Estimation Handbook (FEH) techniques. The normal depth calculations are applied at a number of discrete cross-sections with linear interpolation between to form a 3-dimensional water surface. This is overlain on the DEM to produce the flood outline. Careful manual checking is required at a number of stages. The JFLOW model is based on a discretised form of the 2-dimensional diffusive wave equation and directly simulates the flood outline in a series of overlapping short (1 km) reaches. Flood outlines from the overlapping reaches are merged to produce the overall flood envelope. The model has been written to work as a screen-saver, allowing distributed processing across all computers in an office and manual intervention is minimal. In simple valley situations both methods give similar results, but show differences in more complex areas. Each has advantages and disadvantages, but both have been shown to be a practicable solution to allow production of 160,000 km of flood outline in 12 months.

The challenge of forecasting impacts of flash floods: test of a simplified hydraulic approach and validation based on insurance claim data

“…A temporal resolution of 15 min has been used here for the computations. The effective rainfall is then propagated onto both the hillslopes and the river network using either the kinematic wave model (Borah et al, 1980), or the Hayami solution for diffusive wave model (Moussa, 1996) to represent the flood wave attenuation in river reaches with slopes less than 0.6 %.…”

“…Such resolutions are not suited to the representation of floodplains of small streams. On the other hand, detailed flood inundation mapping approaches are available at higher resolutions (Bradbrook et al, 2005;Sanders, 2007;Nguyen et al, 2015), but require large computational resources which limit the implementation possibility at a large scale. In both cases, most of the proposed mapping approaches would not be compatible with application in real time.…”

The potential of digital filtering of generic topographic data for geomorphological research

“…These datasets are now available, at a relatively low cost, in a large, rapidly growing database for general access. The elevation dataset, which will include complete coverage of western Europe and the USA by the end of 2008 has already been widely used in hydrological and geomorphological research in the UK (Bates et al, 2005;Bradbrook et al, 2005;Smith et al, 2006). These data have a quoted vertical RMSE of ±0·5 to ±3 m and sample spacing of 5 m (Mercer, 2004).…”

Analysis of the French insurance market exposure to floods: a stochastic model combining river overflow and surface runoff

“…The K-Manning values are estimated from land use (Corine land cover). As mentioned by Bradbook et al (2005), the lack of precision in the elevation grid will induce approximations in the expansion of the flood, which will have higher impacts than the uncertainties in river flow estimations.…”