多尺度优化技术在沿海风场同化预报系统中的应用

针对预报系统同化资料的时空分布特征，设计并开展了三组针对不同尺度信息优化的资料同化试验，并使用沿海地区两个测风塔边界层内风场观测数据对模式预报结果进行检验。测风塔观测表明陆地与海上近地层的风场特征截然不同。各组同化试验均能够再现陆地和海上观测风场的主要特征，但海上测风塔的风场预报误差高于陆地测风塔。各组同化试验的预报结果存在较大差异，其中结合新动量控制变量和大尺度约束的试验能够最好地模拟出观测风场的风向和风速分布。进一步表明，沿海地区的近地层风场模拟仍然存在较大的不确定性，需要进一步优化海洋边界层的参数化方案。

Application of multi-scale optimization technology in coastal wind field assimilation forecast system

According to the spatial and temporal characteristics of the observations used in this study, three data assimilation experiments were designed and carried out to optimize multi-scale information. The wind measurements in the boundary layer from two masts near the coast were used to verify the model forecasts. The observations show that the characteristics of wind field are distinct between the onshore and offshore masts. The model forecasts were able to reproduce the main feature of the observed wind field both onshore and offshore. However, the forecasts at the offshore mast were inferior to that of the onshore. The forecasts of different experiments vary in different cases with the experiment using new momentum control variables and large-scale constraint performing the best in general. Results also show that there is still great uncertainty in the forecast of nearshore wind and it is necessary to improve the parameterization scheme of marine boundary layer.