Predicting floodplain inundation: raster-based modelling versus the finite-element approach

We compare two approaches to modelling floodplain inundation: a raster-base d approach, which uses a relatively simple process representation, with cha nnel flows being resolved separately from the floodplain using either a kin ematic or diffusive wave approximation, and a finite-element hydraulic mode l aiming to solve the full two-dimensional shallow-water equations. A flood event on a short (c. 4 km) reach of the upper River Thames in the UK is si mulated, the models being validated against inundation extent as determined from satellite synthetic aperture radar (SAR) imagery. The unconstrained f riction parameters are found through a calibration procedure, where a measu re of fit between predicted and observed shorelines is maximized. The raste r and finite-element models offer similar levels of performance, both class ifying approximately 84% of the model domain correctly, compared with 65% f or a simple planar prediction of water surface elevation. Further discrimin ation between models is not possible given the errors in the validation dat a. The simple raster-based model is shown to have considerable advantages i n terms of producing a straightforward calibration process, and being robus t with respect to channel specification. Copyright (C) 2001 John Wiley & So ns, Ltd.