Abstract An advanced online-coupled meteorology and chemistry model WRF-CAM5 has been applied to East Asia using triple-nested domains at different grid resolutions (i.e., 36-, 12-, and 4-km) to simulate a severe dust storm period in spring 2010. Analyses are performed to evaluate the model performance and investigate model sensitivity to different horizontal grid sizes and aerosol activation parameterizations and to examine aerosol-cloud interactions and their impacts on the air quality. A comprehensive model evaluation of the baseline simulations using the default Abdul-Razzak and Ghan (AG) aerosol activation scheme shows that the model can well predict major meteorological variables such as 2-m temperature (T2), water vapor mixing ratio (Q2), 10-m wind speed (WS10) and wind direction (WD10), and shortwave and longwave radiation across different resolutions with domain-average normalized mean biases typically within ±15%. The baseline simulations also show moderate biases for precipitation and moderate-to-large underpredictions for other major variables associated with aerosol-cloud interactions such as cloud droplet number concentration (CDNC), cloud optical thickness (COT), and cloud liquid water path (LWP) due to uncertainties or limitations in the aerosol-cloud treatments. The model performance is sensitive to grid resolutions, especially for surface meteorological variables such as T2, Q2, WS10, and WD10, with the performance generally improving at finer grid resolutions for those variables. Comparison of the sensitivity simulations with an alternative (i.e., the Fountoukis and Nenes (FN) series scheme) and the default (i.e., AG scheme) aerosol activation scheme shows that the former predicts larger values for cloud variables such as CDNC and COT across all grid resolutions and improves the overall domain-average model performance for many cloud/radiation variables and precipitation. Sensitivity simulations using the FN series scheme also have large impacts on radiations, T2, precipitation, and air quality (e.g., decreasing O3) through complex aerosol-radiation-cloud-chemistry feedbacks. The inclusion of adsorptive activation of dust particles in the FN series scheme has similar impacts on the meteorology and air quality but to lesser extent as compared to differences between the FN series and AG schemes. Compared to the overall differences between the FN series and AG schemes, impacts of adsorptive activation of dust particles can contribute significantly to the increase of total CDNC (∼45%) during dust storm events and indicate their importance in modulating regional climate over East Asia.

中文翻译：

---

WRF-CAM5 在东亚沙尘暴期间的精细应用：对网格分辨率和气溶胶激活参数化的敏感性

摘要 先进的在线耦合气象和化学模型 WRF-CAM5 已应用于东亚，使用不同网格分辨率（即 36、12 和 4 公里）的三重嵌套域来模拟2010 年春季。执行分析以评估模型性能并研究模型对不同水平网格大小和气溶胶激活参数化的敏感性，并检查气溶胶-云相互作用及其对空气质量的影响。使用默认的 Abdul-Razzak 和 Ghan (AG) 气溶胶激活方案对基线模拟的综合模型评估表明，该模型可以很好地预测主要气象变量，如 2 米温度 (T2)、水汽混合比 (Q2)、 10 米风速 (WS10) 和风向 (WD10)，以及不同分辨率的短波和长波辐射，域平均归一化平均偏差通常在 ±15% 以内。基线模拟还显示了对降水的适度偏差和对与气溶胶-云相互作用相关的其他主要变量的中到大的低估，如云滴数浓度 (CDNC)、云光学厚度 (COT) 和云液态水路径 (LWP) ) 由于气溶胶云处理的不确定性或限制。模型性能对网格分辨率很敏感，特别是对于 T2、Q2、WS10 和 WD10 等地表气象变量，这些变量的性能在更精细的网格分辨率下通常会提高。灵敏度模拟与替代方案（即 Fountoukis 和 Nenes (FN) 系列方案）和默认方案（即，AG 方案）气溶胶激活方案表明，前者在所有网格分辨率下预测了云变量（如 CDNC 和 COT）的更大值，并提高了许多云/辐射变量和降水的整体域平均模型性能。使用 FN 系列方案的灵敏度模拟还通过复杂的气溶胶-辐射-云-化学反馈对辐射、​​T2、降水和空气质量（例如，降低 O3）产生很大影响。在 FN 系列方案中包含粉尘颗粒的吸附激活对气象和空气质量有类似的影响，但与 FN 系列和 AG 方案之间的差异相比影响较小。相比FN系列和AG方案的整体差异，