Evaluation of Wave-Ice Parameterization Models in WAVEWATCH III® Along the Coastal Area of the Sea of Okhotsk During Winter

Abstract

Ocean surface waves tend to be attenuated by interaction with sea ice. In this study, six sea ice models in the third-generation wave model WAVEWATCH III® (WW3) were used to estimate wave fields over the Sea of Okhotsk (SO). The significant wave height (Hs) and mean wave period (Tm) derived from the models were evaluated with open ocean and ice-covered conditions, using SO coastal area buoy observations. The models were validated for a period of 3 years, 2008–2010. Additionally, the impact of sea ice on wave fields was demonstrated by model experiments with and without sea ice. In the open ocean condition, the root-mean square error (RMSE) and correlation coefficient for hourly Hs are 0.3 m and 0.92, and for hourly Tm 0.97 s and 0.8. In contrast, for the ice-covered condition, the averaged RMSE and correlation coefficient from all models are 0.44 m (1.6 s) and 0.8 (0.6) for Hs (Tm), respectively. Therefore, except for the bias, the accuracy of model results for the ice-covered condition is lower than for the open water condition. However, there is a significant difference between the six sea ice models. For Hs, the empirical formula whereby attenuation depends on the frequency relatively agrees with the buoy observation. For Tm, the empirical formula that is a function of Hs is better than those of other simulations. In addition, the simulations with sea ice drastically improved the wave field bias in coastal areas compared to the simulations without sea ice. Moreover, sea ice changed the monthly Hs (Tm) by more than 1 m (3 s) in the northwestern part of the SO, which has a high ice concentration.