An Empirical Wind-Wave Model for Hurricane-Forced Wind Waves in the Caribbean Sea

Under excitation by hurricanes, previously quiescent wave fields and their properties can rapidly evolve as they absorb energy from the turbulent atmosphere. Significant wave height (H-s) is one of the primary variables scrutinized during extreme events. Thus, using Hurricane Dorian (2019) a case study, an empirical wind-wave model for forecasting hurricane-induced wind seas is proposed for the Caribbean Sea (CS). Results are compared with the current South China Sea (SCS) state-of-the-art using 10 other hurricanes that passed through the CS from 2011-2019. Results illustrate that the proposed CS model can achieve similar correlation coefficients between observed and simulated H-s as the SCS model, but crucially, the CS model produces approximately half the root mean square error generated by the SCS model. Mean average percentage errors were also significantly lower using the CS model for H-s predictions compared to the SCS model. These results illustrate a comprehensive improvement in regional H-s predictions under hurricane conditions. However, due to the single predictor of wind speed for either model, they both fail under swell conditions and modulation of the wave field by surface currents, leaving room for future improvement by the inclusion of additional variables. Plain Language Summary Hurricanes can cause extreme waves that can change rapidly where the significant wave height is a property that is especially important. Using wind and wave information from Hurricane Dorian (2019), this paper develops an equation to forecast hurricane-induced waves in the Caribbean Sea (CS) and compares it with another equation developed in the South China Sea (SCS). Results show that the CS model comprehensively outperforms the SCS in significant wave height predictions at minimal computational costs and time. However, because both models are dependent on only a single variable, when the wave field is no longer purely forced by the wind and is affected by ocean swell or surface currents, room for improvement by the addition of other variables remains.