A STUDY ON THE JOINT PROBABILITY OF WAVES AND WATER LEVELS DURING TYPHOONS

Wave and water level are important factors to induce coastal flooding. Sea dyke is still the main construction to defense the sea impacts. The estimation of sea dyke height is therefore an important issue. They are commonly estimated by summation of significant wave height of certain return period, astronomical tide, storm surge. This is due to the assumption of max. significant wave heights and the water levels happen simultaneously. The objective of this paper is to study the probability and uncertainty of joint occurrences of high waves and high water levels. 111 typhoons data sets were collected from four field stations around Taiwan Waters. For the traditional frequency analysis method (FAM), we find the best model to simulate the significant wave height is Weibull Distribution. The best model for simulating water level is Extreme Type I Distribution. Traditional design wave heights and water levels are thus estimated by various frequency years. In this paper, the joint probability of wave height and water level was estimated by joint probability method (JPM). Comparative result shows that the total water level estimated by FA is lower than it estimated by JPM under the same occurrence probability. Monte Carlo simulation method was used to simultaneously simulate large amount data of significant wave height and water level, in order to assess the uncertainty of the joint probability of wave height and water level. The coefficient of variation (COV) is used as the uncertainty index. We find the uncertainty of joint probability did reduce by increasing the simulation data. However, the amount of field data is required to improve the simulation accuracy. The simulation result shows that when the field data amount is more than 1000, the uncertainty of joint probability of wave height and water level is less than 10%. This is a useful information for further study when one has to collect field data.