Validating a data-driven methodology for 2D Wave Spectra Correction across Brazil’s complex wave climate

Accurate wave spectra forecasting is essential for marine operations along the diverse Brazilian coast, yet numerical wave models frequently exhibit systematic biases. This study evaluates the Wave Spectra Correction (WASCO) methodology, a data-driven technique combining Principal Component Analysis and Multiple Linear Regression to refine 2D wave spectra forecasts at Recife, Santos, and Rio Grande. Spectral clustering analysis maps regional complexity, revealing a transition from a stable northeastern trade wind regime to a complex bimodal southeastern zone and a seasonally alternating southern system. WASCO consistently reduces Mean Absolute Error across spectral energy, frequency, and directional distributions. The greatest performance gains occur where baseline models show the largest errors, including a 39.7 % reduction in 2D spectral energy density error at Recife. Evaluation across forecast lead times confirms that WASCO provides stable and reliable corrections up to 48 h, maintaining H s error reductions of 6 % to 20 % and directional spread improvements of 10 % to 32 %, beyond which atmospheric forcing errors constrain post-processing efficacy. Furthermore, the method efficiently corrects swell parameters and directional spread, confirming its utility as a spectral shape corrector for operational forecasting.