Abstract
A highly efficient unidirectional deterministic wave prediction model is applied to measurements in weakly nonlinear sea states. The model is applied to both long-crested wave groups and irregular wave fields in a wave basin and, additionally, to irregular ocean waves measured offshore of Sandpatch, Albany, Western Australia. Single fixed-point and drifting-buoy measurements are used as input into the model which first linearizes the signal, then smooths and extends the ends of the record, before advancing it in space and time, and finally reconstructing and re-inserting the estimated bound harmonics. The linearization/re-insertion is based on narrowband approximations of the sum and difference terms and is a very computationally cheap time-domain process. Reflection analysis is necessary for basin data, as a second step, prior to propagation of linear incident waves. It is shown that the nonlinear model gives a better agreement between predicted wave fields and physical measurements downwave than the linear prediction model. Error calculations inside the forecast prediction zone indicate that the prediction error from the nonlinear model is slightly smaller than that for linear prediction