Abstract Four-dimensional data assimilation (FDDA) was applied in WRF-CMAQ model sensitivity tests to study the impact of observational and analysis nudging on model performance in simulating inversion layers and O3 concentration distributions within the Uintah Basin, Utah, U.S.A. in winter 2013. Observational nudging substantially improved WRF model performance in simulating surface wind fields, correcting a 10 °C warm surface temperature bias, correcting overestimation of the planetary boundary layer height (PBLH) and correcting underestimation of inversion strengths produced by regular WRF model physics without nudging. However, the combined effects of poor performance of WRF meteorological model physical parameterization schemes in simulating low clouds, and warm and moist biases in the temperature and moisture initialization and subsequent simulation fields, likely amplified the overestimation of warm clouds during inversion days when observational nudging was applied, impacting the resulting O3 photochemical formation in the chemistry model. To reduce the impact of a moist bias in the simulations on warm cloud formation, nudging with the analysis water mixing ratio above the planetary boundary layer (PBL) was applied. However, due to poor analysis vertical temperature profiles, applying analysis nudging also increased the errors in the modeled inversion layer vertical structure compared to observational nudging. Combining both observational and analysis nudging methods resulted in unrealistically extreme stratified stability that trapped pollutants at the lowest elevations at the center of the Uintah Basin and yielded the worst WRF performance in simulating inversion layer structure among the four sensitivity tests. The results of this study illustrate the importance of carefully considering the representativeness and quality of the observational and model analysis data sets when applying nudging techniques within stable PBLs, and the need to evaluate model results on a basin-wide scale.

中文翻译：

---

四维数据同化 (FDDA) 在模拟 Uintah 盆地 CMAQ 模拟的冬季臭氧浓度的逆温层结构和分布时对 WRF 性能的影响

摘要 2013 年冬季，在 WRF-CMAQ 模型敏感性测试中应用四维数据同化（FDDA），研究观测和分析轻推对模拟美国犹他州 Uintah 盆地逆温层和 O3 浓度分布的模型性能的影响。观测微调在模拟地表风场、校正 10 °C 暖地表温度偏差、校正对行星边界层高度 (PBLH) 的高估和校正由常规 WRF 模型物理学产生的反演强度的低估方面显着提高了 WRF 模型性能，而无需微调。然而，WRF 气象模型物理参数化方案在模拟低云时性能不佳的综合影响，温度和湿度初始化以及随后的模拟场中的温暖和潮湿偏差，可能会在应用观测轻推的逆温日放大暖云的高估，从而影响化学模型中由此产生的 O3 光化学形成。为了减少模拟中潮湿偏差对暖云形成的影响，应用了行星边界层 (PBL) 上方分析水混合比的微调。然而，由于垂直温度剖面分析较差，与观测轻推相比，应用分析轻推也增加了模拟逆温层垂直结构的误差。结合观测和分析轻推方法导致不切实际的极端分层稳定性，在 Uintah 盆地中心的最低海拔处捕获污染物，并在四个敏感性测试中模拟逆温层结构的 WRF 性能最差。本研究的结果说明了在稳定 PBL 中应用微调技术时仔细考虑观测和模型分析数据集的代表性和质量的重要性，以及在全流域范围内评估模型结果的必要性。