利用地球静止轨道卫星Himawari-8气溶胶光学厚度 (AOD) 产品能够估算空间覆盖范围广、时间分辨率高的近地表 PM_2.5 浓度。基于三维变分同化系统将AOD估算得到的 PM_2.5 资料同化进入WRF-Chem大气化学模式中，通过控制实验与同化实验的对比与分析，探讨了AOD估算得到的PM_2.5资料同化对 PM_2.5 污染模拟的改进作用。实验结果表明：(1) AOD估算得到的 PM_2.5 资料同化能够改进 PM_2.5 污染模拟效果；(2) PM_2.5 污染模拟改进效果存在时空差异。此外，与其他研究中使用AOD观测算子直接同化AOD的方法相比，该方法的操作更加简单。

The aerosol optical depth (AOD) product of Himawari-8 satellite can be used to estimate PM_2.5 concentration near the earth surface with wide spatial coverage and high temporal resolution. Based on the three-dimensional variational assimilation system, the PM_2.5 data estimated by AOD are assimilated into the WRF-Chem atmospheric chemistry model. Then through the comparative analysis of control experiment and assimilation experiment, the improvement of PM_2.5 pollution simulation by assimilation of PM_2.5 data estimated by AOD is discussed. The results show that the assimilation of PM_2.5 data estimated by AOD has a significant effect on the improvement of the background field, resulting in the significant improvement of PM_2.5 pollution prediction. And in other hand, the improvement of PM_2.5 pollution from AOD variational assimilation is different in time and space. In addition, compared with the other methods using AOD observation operators directly to assimilate AOD, this method has the advantage of easy operation.