PROBABILISTIC METHODS FOR FLOOD RISK ASSESSMENT

In this paper a comparison of probabilistic techniques for risk assessment of sea defences is presented. Where a long set of simultaneous records of wave conditions and water levels are available a time series of the key design quantity can be generated and analysed with standard procedures. An alternative is to use the available data to estimate the statistical characteristics of the design variable directly. This can be done very swiftly, using analytical techniques such as the point estimate method to give approximate answers where incomplete information about the loading variables exists. In this paper the respective advantages and disadvantages of different techniques are discussed. Example calculations based on twenty-one years of data are presented to provide a comparison of the different methods. Initial results suggest that useful quantifiable bounds on likely sea defence performance can be obtained from incomplete data.