Chong Li1, Jing Li1, Oleg Dubovik2
- Department of Atmospheric and Oceanic Sciences, School of Physics, Peking University
- Laboratoire d’Optique Atmosphérique, CNRS/Université Lille-1, 59655 Villeneuve d’Ascq, France
In recent years, satellite remote sensing has advanced the observations of aerosol properties on both regional and global scales. Although global validation proves that current sensors largely meet accuracy requirements of aerosol monitoring, large uncertainties still exist regionally. For example, over China, only ~50% of the aerosol optical depth (AOD) retrieved by the VIIRS satellite fall within the expected accuracy interval with an overall bias of 0.13. The primary sources of uncertainties in satellite aerosol retrieval include cloud screening, surface reflectance parameterization, aerosol model assumption and aerosol vertical profile assumptions. In this study, we first conducted a systematic investigation of the sources of uncertainty in aerosol retrieval through scattering and radiative transfer calculation. Results show that the vertical distribution of absorbing aerosols has significant impact on the AOD retrieval, leading to ~ 30% AOD error. By refining this assumption using lidar measurements, the accuracy of VIIRS retrieved AOD greatly improved. Especially during the winter season, the improvement can be as large as 80% (Figure 1a). We further developed a joint retrieval method for lidar and satellite sensors using the GRASP algorithm, which not only improved the accuracy of AOD retrieval, but also allowed the retrieval of more parameters such as aerosol extinction profile and surface properties. The AOD retrieved using AHI measurements onboard Himawari-8 satellite achieved excellent agreement with ground observation in China, with a correlation above 0.8 (Figure 1b). This study has provided a high accuracy and high spatial-temporal resolution AOD dataset over China for environmental and climate studies.