PERFORMANCE EVALUATION OF GLOBAL AND LOCAL WAVE MODELS FOR SIGNIFICANT WAVE HEIGHT FORECASTING IN SANTOS BAY, SP

Abstract

Climate change has intensified extreme events in Santos Bay, such as high waves and elevated tides. This study aims to evaluate four local models (SWAN v1, SWAN v2, Delft3D, and ADCIRC) and two global models (CMS-Wave and WAVEWATCH III) in simulating significant wave height at Ilha das Palmas, Santos, SP, Brazil, during 2023. Statistical metrics such as Bias, RMSE, Willmott Skill Score, and correlation were applied. The results show that all models perform well in predicting significant wave height in Santos Bay, especially the local scale models, with Willmott values above 0.88 and RMSE below 0.37 m. Based on the obtained results, it is concluded that the models have good capability to predict and simulate significant wave height variations in Santos Bay, with the Delft3D model standing out for its greater agreement with observed data, presenting lower errors and higher correlation.

Keywords: Significant wave height; numerical models; Santos Bay.