This study aims to enhance the accuracy of the Weather Research and Forecasting model coupled with Chemistry (WRF-Chem) in forecasting Asian dust storms (ADSs) by using the micro-Genetic Algorithm (μGA). We developed an optimization system---the WRF-Chem-μGA system---to seek the optimal combination of the planetary boundary layer (PBL) and land surface parameterization schemes, which are crucial for numerical forecast of dust storms. The optimization was conducted concerning meteorological and air quality variables, i.e., aerosol optical depth, PBL height, 2 m temperature, 2 m relative humidity, and 10 m wind speed, simultaneously for three ADS cases over the optimization domain, including South Korea. Among a total of 32 available combinations of physical parameterization scheme options (8 from PBL and 4 from land surface schemes), the optimized set through the WRF-Chem-μGA system consists of the Asymmetrical Convective Model version 2 (ACM2) for the PBL scheme and the Noah land surface model with Multiple Parameterization options (Noah-MP) for the land surface scheme. The optimized set showed an improvement ratio of up to 22.5 % in terms of the normalized RMSE for all meteorological and air quality variables, compared to various non-optimized sets of physical parameterization schemes for two additional ADS cases. The optimal set proposed in this study can be used comprehensively in numerical forecasts of various meteorological and air quality problems in the East Asian region, using the WRF-Chem model.

中文翻译：

---

通过物理参数化方案的组合优化，利用 WRF-Chem 改进亚洲沙尘暴预测

本研究旨在通过使用微遗传算法（μGA）提高天气研究和预报模型与化学（WRF-Chem）相结合预测亚洲沙尘暴（ADS）的准确性。我们开发了一个优化系统——WRF-Chem-μGA系统——寻求行星边界层（PBL）和地表参数化方案的最佳组合，这对于沙尘暴的数值预报至关重要。针对气象和空气质量变量，即气溶胶光学深度、PBL 高度、2 m 温度、2 m 相对湿度和 10 m 风速，同时对包括韩国在内的优化域内的三个 ADS 案例进行了优化。在物理参数化方案选项的总共 32 种可用组合中（8 种来自 PBL，4 种来自陆地表面方案），通过 WRF-Chem-μGA 系统优化的组合包括 PBL 方案的非对称对流模型版本 2 (ACM2)以及用于陆地表面方案的具有多个参数化选项的诺亚陆地表面模型（Noah-MP）。与另外两个 ADS 案例的各种非优化物理参数化方案组相比，优化组在所有气象和空气质量变量的归一化 RMSE 方面显示出高达 22.5% 的改进率。本研究提出的最优集可以利用WRF-Chem模型综合用于东亚地区各种气象和空气质量问题的数值预报。