Shui, YKieviet, JohanErfort, GCheng, ZMa, YChen, P2026-02-202026-02-202026-101873-52580029-8018https://doi.org/10.1016/j.oceaneng.2025.121742http://hdl.handle.net/10204/14699Offshore wind turbines are vital for sustainable energy but face deployment challenges due to extreme envi ronmental conditions. Accurate environmental modeling is crucial for optimal design, yet comprehensive data for the South China Sea and South African Sea remains scarce. This study addresses these gaps by establishing joint probability distributions of wind speed (Uw), significant wave height (Hs), and peak wave period (Tp) using ERA5 reanalysis data (2004–2023). Validation against 20-year buoy data from Slangkop (South Africa) shows good agreement, with extreme value discrepancies below 10 %. The framework integrates Weibull and lognormal marginal distributions validated via probability plot paper and employs a binning strategy to stabilize parameter estimation. Conditional dependencies are modeled through power-law relationships (Uw, Hs) and lognormal distribution (Tp), while the Rosenblatt transformation derives 50-year environmental contour surfaces for probabilistic extreme condition prediction. Results reveal distinct regional contrasts: the South China Sea exhibits higher wind speeds with moderate waves, whereas the South African Sea experiences similar winds but more severe waves. These findings emphasize the need for region-specific turbine designs—prioritizing motion stability in the South China Sea and resonance mitigation in the South African Sea. This study suggests continuous data integration to enhance reliability under extreme conditions.AbstractenEnvironmental conditionsLong-term ocean assessmentOffshore wind turbinesSouth China seaSouth African seaArticleN/A