No abstract

Storm surge simulation models require certain parameters for evaluating the worst possible event that could occur at a site with respect to a certain return period. The most signi cant probable maximum tropical cyclone parameters are pressure de ciency at the centre (ΔP) and maximum wind speed (V max ) during the cyclone. In this study the probable maximum tropical cyclone parameters that would yield the maximum probable storm surge along the Indian coasts of Arabian Sea and Bay of Bengal are estimated. Datasets are created based on various existing data archives for each basin. The datasets are subjected to extreme value analysis for determining the ΔP and V max parameters. The data is tted to various probability distributions (Gumbel, Frechet, Weibull and Log-normal) whose parameters (scale, shape, and location parameters) are estimated using graphical (least square t) and numerical (order statistics approach) methods. A mean recurrence interval of 1,000 and 10,000 years is considered for strategic structures. The best t distribution and its parameters are obtained based on goodness of t criteria. The resulting ΔP and V max is compared with theoretical maximum cyclone parameter values of each basin and revised till an optimal set of values are reached. The analysis shows that ΔP and V max values for Arabian Sea and Bay of Bengal are best represented by Weibull distribution. The estimated parameters are useful input to a storm surge model to determine the design basis ood level for the strategic coastal sites.

Storm surge simulation models require certain parameters for evaluating the worst possible event that could occur at a site with respect to a certain return period. The most significant probable maximum tropical cyclone parameters are pressure deficiency at the centre (ΔP) and maximum wind speed (Vmax) during the cyclone. In this study the probable maximum tropical cyclone parameters that would yield the maximum probable storm surge along the Indian coasts of Arabian Sea and Bay of Bengal are estimated. Datasets are created based on various existing data archives for each basin. The datasets are subjected to extreme value analysis for determining the ΔP and Vmax parameters. The data is fitted to various probability distributions (Gumbel, Frechet, Weibull and Log-normal) whose parameters (scale, shape, and location parameters) are estimated using graphical (least square fit) and numerical (order statistics approach) methods. A mean recurrence interval of 1,000 and 10,000 years is considered for strategic structures. The best fit distribution and its parameters are obtained based on goodness of fit criteria. The resulting ΔP and Vmax is compared with theoretical maximum cyclone parameter values of each basin and revised till an optimal set of values are reached. The analysis shows that ΔP and Vmax values for Arabian Sea and Bay of Bengal are best represented by Weibull distribution. The estimated parameters are useful input to a storm surge model to determine the design basis flood level for the strategic coastal sites.