The Errors-in-Variables approach for the validation of the WAM wave model in the Aegean Sea

Abstract

In previous studies, wave fi elds from the 3rd-generation wave model WAM-Cycle 4 have been validated by using in situ buoy measurements in the Aegean Sea within the framework of the POSEIDON project; however, limitations of the data sets, concerning mainly the short distance of the buoys from the shore and the short length of the data, render those validation studies incomplete. In this work, signifi cant wave height forecasts obtained from WAM-Cycle 4 wave model are validated by means of TOPEX/Poseidon (T/P) data in specifi c offshore locations in the central part of the North Aegean Sea. The linear structural relationship between the two data sets has been modelled by implementing the Error- In-Variables approach, assuming that both T/P data and WAM results are subjected to errors. The underestimation of signifi cant wave height from WAM, which has been concluded from the comparison with buoys at near-shore points, is also observed from the WAM-T/P comparison at offshore locations, thus being considered of general validity for the Aegean Sea. In addition, a correction relation for the WAM model results, based on the linear structural relationship, is proposed and applied.