Combined Influence of Storm Rainfall and Tidal Anomaly on Coastal Flooding

In inter-tidal zones, flooding may be a result of high sea levels, high surge or extreme rainfall. To provide adequate levels of defence against flooding, accurate predictions of the probability of exceedence must be determined. A method using the Gibbs sampler for generating sea levels and rainfall variables above some threshold is presented. This is combined with a stochastic emulation of a catchment hydraulic flow routing model (HFRM). The method allows for correlation between sea levels and rainfall. The use of a threshold concentrates the fitting of the distributions of these variables to values likely to cause floods. The method is also significantly faster than time series techniques and requires less calibration. Flood risk estimates for a catchment, loosely based on Port Adelaide, are compared with the simulated floods obtained by using a time series input to a HFRM. The Gibbs sampler approach gives close agreement with the simulated results. This paper also demonstrates the need to allow for correlation between the causes of flooding.