Assessing the performances of parametric wind models in predicting storm surges in the Pearl River Estuary

Abstract

Portraying wind fields plays a crucial role in storm surge simulation, which includes the information of the storm track, central maximum wind speed, maximum wind speed radius, and pressure drop. In this study, five widely used parametric wind models and their performances in predicting storm surges have been estimated in the Pearl River Estuary. The synthetic wind field reveals that different wind models produce wind fields with diverse structures, each with its own focus and advantages for the storm's center or its periphery. By implementing these wind models into surge simulation, the wind and surge estimates show a strong dependence on model selection. Selection of parametric wind models may lead to a surge height error greater than 1.45m. The distribution of maximum wind speed and peak water level under various typhoons dominated by storm intensity is also investigated. As the storm intensity increases, the prediction deviations between these selected models are also enlarged and over 10% in Typhoon Hato. Overall, the model proposed by Emanuel and Rotunno (2011) yields accurate estimates among the selected parametric wind models. The uncertainty of the storm track parameters and drag coefficient are quantified based on Emanuel and Rotunno (2011) model. Small track changes can result in huge variations in surge heights with maximum values of 0.5–1m. The choice of drag coefficient also has a maximum difference of 5.4%, which contributes a significant uncertainty to the simulation. The multiple choice of parameters in wind models increases the risk of uncertainty propagation. This study is expected to provide a reference for the numerical modeling of wind field and storm surge.