<p>This study addresses the complex dependencies among met-ocean variables across China’s Bohai Sea, Yellow Sea, East China Sea, and South China Sea under dynamic marine environments, evaluating the applicability of mixed copula models for joint distribution modeling in both shallow and deep-water regions. Leveraging three decades of European Centre for Medium-Range Weather Forecasts (ECMWF) ERA5 reanalysis data from 1994 to 2023, mixed lognormal and mixed three-parameter Weibull distributions are employed to enhance marginal fitting for met-ocean variables. Mixed copula models were constructed using five common copula functions—Gumbel, Student’s <i>t</i>, Gaussian, Clayton, and Frank—selected to capture the diverse tail dependence characteristics of bivariate met-ocean variable structures. Parameter estimation was performed via the expectation-maximization algorithm, and the optimized model was integrated with the inverse first-order reliability method to generate environmental contours corresponding to design return periods. Results indicate that the mixed marginal distribution model improves the fit to the marginal distributions of met-ocean variables; although bivariate distributions resist unique characterization due to substantial spatial and bathymetric variations in dependence structures, the mixed copula models effectively represent these complex interdependencies. The environmental contours derived from the established mixed copula joint distribution provide accurate design parameters for marine engineering applications, thereby addressing regional systematic limitations and methodological constraints. This approach offers a robust scientific basis for engineering safety assessments in China’s major sea areas.</p>