A new joint probability appraisal of flood risk

Abstract

Prior to the construction of the Cardiff Bay Barrage the risk of flooding of the Cardiff Bay waterfront was from inundation by surge tides. The barrage has substantially eliminated the probability of tidal flooding, and the residual flood risk now depends upon the ability of the inland bay to provide sufficient storage for river floodwater during periods of high tide. In the course of the feasibility and design studies the actual reduction in flood hazard was not quantified, but the design was carried out against specific combinations of tidal levels and fluvial flows. Recent research for MAFF at HR Wallingford has now made viable the assessment of the actual degree of flood risk, where flooding depends on the combination of two or more events. This paper presents a new joint probability appraisal of the standard of flood defence for the land around the inland bay, and demonstrates the conservatism incorporated in the original design methodology for flood risks. An examination of individual floods and surge tides and of the daily flow and tide records for a concurrent 18-year period has shown no positive correlation between concurrent high tides and river floods. An ISIS flow model of the Rivers Taff and Ely, the inland bay and the operation of the barrage structure was used to identify combinations of hydro-meteorological events that give specific levels in the inland bay. A statistical frequency assessment of these levels in the inland bay has provided ‘return periods’ for the still water level in the inland bay exceeding, for example, 7·5 m OD and 8·0 m OD. This frequency assessment was undertaken for sea levels with and without an allowance for future rise in mean sea level attributable to global warming. The ISIS modelling contained some conservative assumptions, notably the assumed coincidence between the peak of the river flow and high tide in the Bristol Channel, and the effect of these assumptions was investigated by some sensitivity tests. Finally, a best estimate of the return period for these levels in the inland bay was made using the sensitivity testing and the results of a study in 1994 of the effects of the operation of the barrage. It is concluded that the construction of the barrage is a significant flood defence benefit to Cardiff, which now has greater protection from flooding from the inland bay than the Thames Barrier provides for London. Flood risk from impoundment of the inland bay for the waterfront area remains very low even for the currently used scenarios for mean sea level rise due to global warming to the year 2100.