Meso-micro-scale numerical simulation in marine wind environments considering air-sea coupling

The independent meso-scale weather research and forecasting (WRF) model neglects the modulation effects of sea surface deformation on the turbulent structure of the three-dimensional wind field. This sea surface deformation is induced by wave propagation and ocean current evolution. Consequently, the WRF model cannot output second-scale fluctuating wind speeds. To address this problem, this paper establishes a coupled meso-micro-scale multi-layer nested wind-wave-current numerical simulation method. This method is based on the WRF model, simulating waves nearshore (SWAN), regional ocean modeling system (ROMS) and large eddy simulation coupling framework (WRF-SWAN-ROMS-LES). Additionally, a coupled software platform, Marine_CFD, is developed to enable mesoscale forcing-driven meso-micro-scale coupling. Using this method, we conduct a comparative analysis of wind speed simulation performance among three models. These models include the meso-scale air-sea coupled atmospheric model, namely the coupled ocean-atmosphere-wave-sediment transport (COAWST) model, and the standalone atmospheric models, namely the WRF model and the WRF_SST model. The effectiveness of the micro-scale large eddy simulation in reproducing turbulent fluctuations is also verified. The results indicate that the wind-wave-current coupled model achieves higher accuracy in characterizing the evolutionary process of offshore wind speeds. At the height of 38 m, the Pearson correlation coefficient of the COAWST model increases to 0.789. This value is higher than those of the WRF model (0.733) and the WRF_SST model (0.735). This improvement demonstrates that the COAWST model possesses a distinct advantage in simulating the evolution of marine wind speeds. In the micro-scale simulation, the three-dimensional spatiotemporal fields of velocity, potential temperature, and pressure are extracted from the meso-scale atmospheric output. These fields are then employed to drive the micro-scale large eddy simulation. The resulting wind speed time series are overall situated in the middle region of the fluctuating wind speed curve generated by the large eddy simulation. Moreover, these time series can effectively retain the high-frequency fluctuating characteristics at the second scale.