气溶胶单次散射反照率SSA（Single Scattering Albedo）的卫星定量遥感对气候评估和大气污染治理均具有重要意义。搭载于S5P（Sentinel-5 Precursor）上的对流层监测仪（TROPOMI）具有目前同类卫星传感器中最优的空间分辨率。本文基于S5P/TROPOMI数据开展了中国东部地区的SSA反演研究。首先利用中国东部地区AERONET（Aerosol Robotic Network）站点数据对OPAC（Optical Properties of Aerosols and Clouds）气溶胶模型进行约束改进，构建了更为合适的气溶胶类型，并使用地基激光雷达（Lidar）预设相应气溶胶类型的垂直结构。然后使用辐射传输模型SCIATRAN构建查找表LUT（Look-Up Table），将TROPOMI UVAI（Ultraviolet Absorbing Index）和MODIS AOD（Aerosol Optical Depth）数据联合输入反演气溶胶SSA数据。反演结果与地基站点数据对比，相关系数R2为0.61，均方根误差为0.05；和OMI SSA产品相比，总体趋势一致且具有空间连续性更好。基于TROPOMI的高分辨率SSA算法和数据将有助于中小尺度下气溶胶时空分布、光学特性等研究。
Quantitative retrieval of aerosol Single Scattering Albedo (SSA) from satellite remote sensing is important for climate assessment and air pollution control. This study developed a preliminary SSA retrieval algorithm based on S5p/TROPOMI and Aqua/MODIS in eastern China.The optical absorption of aerosol is highly sensitive in the near-ultraviolet bands, which is significantly correlated with the aerosol model, vertical profile, and corresponding aerosol loading. Considering the actual situation of aerosols in eastern China, the Optical Properties of Aerosols and Clouds aerosol model is constrained using AERONET data in eastern China to produce a more suitable aerosol type, and the corresponding vertical structure of different aerosol types is predefined using ground-based Lidar. Then, the radiative transfer model SCIATRAN is used for sensitivity analysis to further adjust the aerosol model and establish a series of look-up tables (LUTs) for different aerosol subtypes. The SSA can retrieve individual LUT at pixels with the collection of TROPOMI Ultraviolet Absorbing Index (UVAI) and MODIS AOD together. In the process, the ?ngstr?m Index values from MODIS and UVAI are used in combination for preliminary classification of aerosol type to improve the accuracy and efficiency.Compared with the ground-based observations, the coefficient of determination (R2) is 0.61, and the root mean square error is 0.05. Compared with OMI instantaneous inversion and monthly average SSA images, the distributions of TROPOMI SSA show a consistent overall trend and have better spatial continuity and larger inter-pixel variation. Further site-by-site analysis shows that the SSA and AOD are highly correlated with the type of aerosols where the site is located. The SSA in Shanghai with more sea salt aerosols is stable above 0.95, while the Beijing area is affected by multiple factors and the SSA varies greatly with time series from 0.85 to 0.98.In conclusion, the preliminary SSA retrieval algorithm based on TROPOMI in this study has good verification accuracy. Using MODIS aerosol products as input can effectively improve the accuracy of SSA inversion. The algorithm still has some uncertainties and needs to be further improved from the aspects of aerosol type and aerosol vertical profile. Nevertheless, the algorithm is helpful for the classification of aerosol types, aerosol microphysical, and optical properties of aerosol in small- and medium-scale regions.