Correction of the predicted wave characteristics using regression methods – a case study for the Iberian coastal environment

The west Iberian coast is affected by various storms developed in the North Atlantic Ocean. For this reason, an accurate prediction of the sea state conditions is very important to manage the protection of the harbours and population living in the coastal cities. In recent years computing power has increased significantly, which has allowed the improvement of the numerical models. The wave models have undergone the same evolution, which makes them an important tool for the most accurate prediction of the wave conditions. In this study, the wave predictions on the west Iberian coast are performed with the SWAN model. Simulations are performed in winter 2013-2014 when various storms affected all the European coasts. It is well known that in the simulation of extreme wave events some errors may occur. Altimeter measurements are used to correct the errors that appeared in the simulation results. By using these measurements together with regression methods, it was possible to improve the significant wave height fields simulated with the SWAN model in each point of the computational domain of the target area. This improvement is indicated by the statistical parameters calculated by comparing the simulated significant heights with buoy measurements.