Comparison of varying complexity numerical models for the prediction of flood inundation in Greenwich, UK

In recent years, urban flood modelling studies have taken the form of either validation against observations of a single event or benchmarking models of varying complexity for a synthetic flood event. In this paper, the latter framework is undertaken for hypothetic flooding scenarios at Greenwich on the River Thames, UK. A simple diffusion wave storage cell model, LISFLOOD-FP, is compared to a fully 2D hydrodynamic model, SOBEK, at multiple resolutions for a 1-in-100 year return period event. The models yield consistent results at the highest resolution (5 m) with minor differences accounted for by inertial effects and model schematisation. Secondly, both models predict significant degradation in model results at coarse resolutions compared to the high resolution benchmark. Specifically, results at 25 and 50 m suggest different flow structures emerge as the representation of urban structures becomes more coarse. As a result of this finding, the models are setup using a digital terrain model (DTM) with the buildings removed to investigate these emergent flow structures.