Comparative study of typhoon wind hazard estimation in coastal region of China using different wind field parameter models

Abstract

Radius to maximum wind speed (Rmax) and Holland pressure profile parameter (B) are two key parameters for parametric typhoon or hurricane wind field modelling and hazard estimation. They are difficult to be observed or estimated for all historical storms and therefore always unavailable in most best track datasets. The statistical models were developed to correlate these two parameters with other available parameters based on limited real observation data and widely employed in typhoon hazard estimation. The applicability of these models to different ocean regions remains questionable. In this study, series of statistical models of Rmax and B are revisited and classified based on their regions of application, i.e., Gulf of Mexico and Atlantic Basin (GM&AB), South Pacific (SP), Bay of Bengal (BB) and Western North Pacific (WNP, 0°N-60°N, 100°E−180°E). Moreover, recursive models of Rmax and B in WNP are also introduced. The comparison of variation characteristics of different models for these two key parameters along with the change of typhoon surface central pressure deficit ΔPs are conducted. The stochastic typhoon tracking and intensity models are then coupled with an analytical wind field model to estimate the typhoon wind hazard of five coastal cities of China. The probability of exceedance and return period curves of wind speed are compared using different Rmax and B parameter models. The use of different Rmax and B models is shown to have significant effects on wind hazard curves. The extreme wind speeds estimated using parameter models from GM&AB and wind field model in this study are almost consistent with each other. Different models in WNP produce significant differences of wind hazard curve. The new findings in this study provide a forward step to the rational estimation of typhoon wind hazard in WNP by using the appropriate wind field parameter models.