The probability prediction of storm surge disaster along south-east coasts of China

The research of long-term risk analysis method of typhoon induced storm surge disaster is always a compelling topic of marine science and engineering. Traditional approach is using univariate probability model based on yearly maximum sampling to describe the characteristics parameters of typhoon and water level. According to the probability analysis results, samples can be generated by using Monte Carlo method and long-term risk of storm surge disaster with some return period is calculated. In this paper, a double layer multi-objective probability model based on multivariate compound extreme value distribution is proposed. The annual typhoon frequency and correlations among different key parameters of typhoon are considered in the new model. This model is applied to the long-term risk analysis of typhoon induced storm surge disaster along the coast of China and it indicates reasonable results.