<p>A new ocean-atmosphere-wave regional coupled model named Windwave version 1.0 for simulating and predicting winds and waves has been developed for the Northwest Pacific Ocean.<span class="Apple-converted-space"> </span>In particular, the global-to-regional nesting technique is adopted for the ocean component to alleviate the bias due to the inconsistency in the lateral boundary. This paper is devoted to describing the coupling details of Windwave and the initialization scheme and assessing its basic performance, especially in predicting surface winds and significant wave heights (SWH) on the weather timescale. The control experiment set contains 31 experiments for August 2020, with seven typhoons passing through the Northwest Pacific. Each experiment starts at 0:00 am UTC of each day and runs for three days. Experiment results show that the new coupled model performs well in predicting surface winds, SWH, surface air temperature, and sea surface temperature on the weather timescale. In particular, the Root Mean Square Errors (RMSEs) of surface winds at 10 m height over the Northwest Pacific of the control experiment are 1.82 m s<sup>-1</sup>, 2.22 m s<sup>-1</sup>, and 2.59 m s<sup>-1</sup>, respectively, at lead times of 24 h, 48 h, and 72 h. Meanwhile, the RMSEs of SWH at lead times of 24 h, 48 h, and 72 h are 0.39 m, 0.43 m, and 0.51 m. In addition, we have explored the impacts of the different sea surface aerodynamic roughness parameterization schemes on predicting surface winds and SWH. In total, five different sea surface aerodynamic roughness parameterization schemes are adopted, corresponding to one control set and four sensitivity sets of experiments. Under normal conditions, the sea surface aerodynamic roughness parameterization scheme considering the effects of wind-wave direction tends to perform better for winds and waves, while that depending on wave age and SWH tends to perform worse. Under extreme wind and wave conditions, the schemes considering the effects of wind-wave direction and that considering wave age and peak wave length have better performance. These findings can provide new insights for developing a more advanced sea surface aerodynamic roughness parameterization scheme.</p>

Accurate precipitation forecast, especially heavy precipitation forecast, is challenging in weather forecasts. In particular, the threat score (TS), with higher values meaning more accurate precipitation forecasts, at the threshold of heavy precipitation is much lower than that of light precipitation (Case et al., 2011;Huang & Luo, 2017). Heavy precipitation always brings about flash floods, urban waterlogging, heavy snow, and the growth of infectious diseases, posing a severe threat to the lives and properties of humans (

<p>Ocean waves, especially extreme waves, are vital for air-sea interaction and shipping. However, current wave models still have significant biases, especially under extreme wind conditions. Based on a numerical wave model and a deep learning model, we accurately predict the significant wave height (SWH) of the Northwest Pacific Ocean. For each day in 2017-2021, we conducted a 3-day hindcast experiment using WAVEWATCH3 (WW3) to obtain the SWH forecasts at lead times of 24, 48, and 72hr, forced by GFS real-time forecast surface winds. The deep learning-based bias correction method is BU-Net by adding batch normalization layers to a U-Net, which could improve the accuracy. Due to the use of BU-Net, the mean Root Mean Squared Errors (RMSEs) of the SWH forecast from WW3 at lead times of 24, 48, and 72hr are reduced from 0.35m to 0.21m, 0.39m to 0.24m, and 0.43m to 0.30m, corresponding to drop percentages of 40%, 38%, and 30%, respectively. During typhoon passages, the drop percentages of RMSEs reach 45%, 42%, and 35% for three lead times. Therefore, combining numerical models and deep learning algorithms is very promising in ocean wave forecasting.</p>