Accurate wind modeling is important for wind resources assessment and wind power forecasting. To improve the WRF model configuration for the offshore wind modeling over the Baltic Sea, this study performed a sensitivity study of the WRF model to multiple model configurations, including domain setup, grid resolution, sea surface temperature, land surface data, and atmosphere-wave coupling. The simulated offshore wind was evaluated against LiDAR observations under different wind directions, atmospheric stabilities, and sea status. Generally, the simulated wind profiles matched observations, despite systematic underestimations. Strengthening the forcing from the reanalysis data through reducing the number of nested domains played the largest role in improving wind modeling. Atmosphere-wave coupling further improved the simulated wind, especially under the growing and mature sea conditions. Increasing the vertical resolution, and updating the sea surface temperature and the land surface information only had a slight impact, mainly visible during very stable conditions. Increasing the horizontal resolution also only had a slight impact, most visible during unstable conditions. Our study can help to improve the wind resources assessment and wind power forecasting over the Baltic Sea.

准确的风建模对于风资源评估和风功率预测非常重要。为了改进波罗的海海上风建模的 WRF 模型配置，本研究对 WRF 模型对多个模型配置进行了敏感性研究，包括域设置、网格分辨率、海面温度、地表数据和大气波耦合。在不同风向、大气稳定性和海况下，根据 LiDAR 观测对模拟的海上风进行了评估。通常，模拟的风廓线与观测结果相匹配，尽管存在系统性低估。通过减少嵌套域的数量来加强再分析数据的强迫在改进风建模方面发挥了最大的作用。大气波耦合进一步改进了模拟风，尤其是在生长成熟的海况下。提高垂直分辨率、更新海面温度和陆面信息仅产生轻微影响，主要在非常稳定的条件下可见。增加水平分辨率也只会产生轻微影响，在不稳定条件下最为明显。我们的研究有助于改进波罗的海的风资源评估和风功率预测。