Application of the Ensemble Kalman Filter to a high-resolution wave forecasting model for wave height forecast in coastal areas

Two different approaches related to the implementation of the Ensemble Kalman Filter (EnKF) in a wave modelling system SWAN (Simulating Waves Nearshore) are explored. A new methodology is tested first, in which the observations in the EnKF algorithm are replaced by the previously obtained analysis. The forecast is performed with one time-step advance, showing a high improvement in the accuracy of the significant wave height predictions. This technique can be applied to the guidance of ship movements in coastal and harbour approach areas. As a next step, an application of the Kalman filter is designed to improve the wave hindcast in terms of the main statistical parameters. In fact, this represents the beginning of a more theoretical and extended study related to assimilating data in wave hindcasts, where the filtered data corresponds to the best estimative of the underlying state of the system.