Regional assessment of typhoon wind, wave, and current multi-hazards for offshore structure design using SWAN + ADCIRC: Sensitivity to reanalysis wind forcings and extreme value distributions

This study develops a comprehensive framework for regional assessments of typhoon-induced wind, wave, and current multi-hazards required for offshore structure design using SWAN+ADCIRC coupled numerical models. A primary focus is on sensitivity analysis of estimated high-quantile extremes to the reanalysis wind forcing datasets and extreme value distributions, which remain underexplored. Results reveal that the choice of reanalysis dataset has a limited impact on design-level extremes. However, the selection of extreme value models significantly affects estimated hazards. Generally, the obtained hazard maps for 50-year return period values using the generalized extreme value distribution, generalized Pareto distribution, and optimal distribution (based on the Kolmogorov–Smirnov criterion) are comparable, with uncertainties within 10%, whereas Gumbel distribution results in a larger deviation of up to 20%. Spatial analysis shows that the current extremes exhibit greater spatial variability due to localized seabed topography effects. Additionally, a logistic model-based multivariate extreme value distribution is proposed for modelling typhoon wind, wave, and current tri-variates, demonstrating that the obtained environmental contours are highly site-dependent and vary significantly due to the use of different marginal distributions. These findings provide valuable insights for robustly and reliably assessing the design-level hazards and the characterization and modelling of regional typhoon multi-hazards for offshore structure design. • Developed and verified a framework for assessing regional typhoon multi-hazards based on the ADCIRC+ SWAN coupled numerical models • Revealed the limited sensitivity of using different reanalysis wind forcing datasets on the estimated typhoon hazards • Showed the crucial impact of different extreme value distributions on the estimated typhoon hazards • Showed the moderate-to-high correlation strengths among typhoon wind-wave-current extremes • Provided (joint) typhoon wind, wave, and current hazard maps for the Pearl River Delta offshore region