Impacts of physical calibration of a spectral wave model and effects of using different temporal wind inputs

Abstract

Spectral wave modelling can reduce uncertainties in the estimation of wave energy resource assessment, converter design, extreme value analysis, etc. In spectral models, wave growth is represented with different approaches, resulting in wave resource assessments having large differences especially at high wave values. In this paper a modified version of the North Sea Wave Database is used to quantify the impact of wind temporal fidelity on the wind growth components. The Simulating WAves Nearshore (SWAN) model has been modified, with two different wind inputs used from the European Centre for Medium-Range Weather Forecasts (ECMWF). Results are compared with in-situ measurements an inter-comparison for 20 years (1980-1999). Differences are found on mean and maxima values of wave parameters, with little changes in directionality. However, higher temporal resolution of the wind does not mean always a better hindcast, in fact attention to the calibration of wind-wave growth interactions and whitecaps leads to similar results. Finally, the high fidelity hindcasts are compared, identifying limitations and opportunities for improvements in wave energy assessments.