We develop a new methodology for the deterministic forecasting of directional ocean surface waves based on nonlinear frequency corrections. These frequency corrections can be pre-computed based on measured energy density spectra and, therefore, come at no additional computational cost compared to linear theory. The nonlinear forecasting methodology is tested on highly nonlinear synthetically generated seas with a variety of values of average steepness and directional spreading and is shown to consistently outperform a linear forecast.
