Answers are grounded in this paper and cite their sources.
A coupled storm surge and overland flow model is used to simulate extreme coastal flooding of Canvey Island, a dense urban area located at the mouth of the Thames Estuary, U.K. The flood model is based on a shock‐capturing numerical scheme and resolves the terrain and buildings of the study area with high‐resolution topographic data. Repeat simulation is used to propagate uncertainties in model inputs and parameters through to uncertainties in model outputs, and the main sources of uncertainties are described. The greatest uncertainties originate from the forcing inputs to the flood model, including coastal water levels and sea defense failures, rather than its internal boundaries, such as the model terrain and bottom friction. This is consistent with studies in rural and semideveloped floodplains and reflects a combination of high sensitivity to, and uncertainty in, the forcing inputs. However, other numerical and physical variables introduce significant uncertainties, of which many are specific to, or altered by, the urban floodplain. In particular, model predictions are sensitive to the spatial resolution (Δs) of the numerical grid. This includes the effects of Δs on boundary features (walls, hedges, etc.), which are described in the bed friction coefficient, and on the representation of buildings, which are described in the model terrain. Finally, model predictions are integrated across multiple combinations of breach width and location to explore the “relative” hazards (i.e., spatial and temporal variations in hazard) associated with the floodplain geography.
No takes yet. Share an insight, caveat, or question.
Brown et al. (2007) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: