Probabilistic relation between winds and waves during tropical cyclone processes

Combination of extreme winds and waves are essential for engineering design, which mostly occurs during tropical cyclone (TC) processes in China Seas. This study investigates the wind-wave relationship in the sea area adjacent to Hainan Island during the TC processes. A total of 807 TC events and corresponding storm waves from 1949 to 2022 were reconstructed and simulated in numerical wave model SWAN (Simulating WAves Nearshore). Statistical analyses reveal that, for a given wind speed at a specific location, probability distribution of significant wave heights tends to follow the generalized extreme value (GEV) distribution. Characteristic values of the GEV distribution are effectively parameterized as functions of wind speed through piecewise linear regression. At low wind speeds (below 5 m/s), the wave height at a specific location has weak correlation with local wind speed and is mainly influenced by water depth and geographical location. Generally, higher wave heights exist in open and deep-sea areas and smaller wave heights appear in regions sheltered by lands. For wind speeds higher than 5 m/s, the wave height shows a strong positive correlation with local wind speed. The rate of wave height change with local wind speed is small in shallower areas and increases with water depth, while in waters deeper than 100 m, the rate of wave height change tends to stabilize.