Evaluation of Moderate-Resolution Imaging Spectroradiometer (MODIS) Deep Blue Aerosol Products Using Ground-Based Measurements over Beijing

Abstract

Deep Blue (DB) is a new MODIS retrieval algorithm to retrieve aerosol properties over bright surfaces such as arid, semiarid and urban areas. It's expected to fill in data gaps over bright surface left by the standard algorithm based on Dark Target (DT) retrieval algorithm. DB has been employed in recent Collection 5.1 AOD product. This study presents a comparison between the DT approach and the more recent DB algorithm using international Aerosol Robotic Network (AERONET) data as reference, analyzes the improvement in DB AOD product over urban surface and discusses the influence of aerosol model variation on MODIS AOD retrieval. Results show that DB products perform better than DT products under clear condition, and a systematic AOD overestimation in DT products doesn't occur in DB products. DT and DB AOD products both appear strong seasonal pattern, with better performance in autumn and winter and worse performance in spring and summer, and the underestimated Aerosol Single Scattering Albedo (SSA) during spring and summer over Beijing in MODIS retrieval probably is the major factor inducing the AOD overestimation in these seasons. In comparison with DB AOD, DT AOD is more likely overestimated, unsuitable surface reflectance over Beijing may be the key factor, and it may play a more important role in MODIS DT AOD retrieval than unsuitable SSA.