Abstract: The Yin–Yang-grid Unified Model for the Atmosphere (YUNMA) is a nonhydrostatic atmospheric model developed on a quasi-uniform Yin–Yang grid to avoid the pole problem. To improve the performance of the model in simulating atmospheric boundary-layer processes, a scale-aware Mellor–Yamada–Nakanishi–Niino (MYNN) boundary-layer parameterization scheme is implemented into YUNMA, replacing the original Medium-Range Forecast model (MRF) scheme. By incorporating prognostic turbulent kinetic energy and scale-aware parameterization of both local and nonlocal turbulence, the MYNN scheme improves the simulation of turbulent structures and the distribution of physical variables within the boundary layer. A numerical simulation of a sea fog event over the Yellow Sea on February 26, 2014, shows that the YUNMA model with the MYNN scheme reproduced the spatiotemporal evolution of sea fog accurately, with the Threat Score increased by 0.48. The root mean square error of dew point temperature decreased by 0.3°C–1.8°C below 700 hPa. The MYNN scheme enhanced low-level vertical turbulent transport and the vertical structure of temperature and moisture fields, leading to realistic representations of temperature decrease and moisture increase associated with turbulence in the Yellow Sea region. Overall, the MYNN scheme demonstrates a strong capability to represent subgrid-scale turbulent mixing and vertical exchanges. Its implementation evidently improves the performance of the YUNMA model in simulating boundary-layer processes and provides useful guidance for further atmospheric model development.