Publications

Su, X; Cao, MD; Wang, LC; Gui, X; Zhang, M; Huang, YH; Zhao, YJ (2023). Validation, inter-comparison, and usage recommendation of six latest VIIRS and MODIS aerosol products over the ocean and land on the global and regional scales. SCIENCE OF THE TOTAL ENVIRONMENT, 884, 163794.

Abstract
MODIS and VIIRS aerosol products have been used extensively by the scientific community. Products in operation in-clude MODIS Dark Target (DT), Deep Blue (DB), and Multi-Angle Implementation of Atmospheric Correction (MAIAC) and VIIRS DT, DB, and NOAA Environmental Data Record products. This study comprehensively validated and inter -compared aerosol optical depth (AOD) and angstrom ngstrom exponent (AE) over land and the ocean of these six products (seven different algorithms) on regional and global scales using AErosol RObotic NETwork (AERONET) and Maritime Aerosol Network (MAN) observations. In particular, we used AERONET inversions to classify AOD and AE biases into different scenarios (depending on absorption and particle size) to obtain retrieval error characteristics. The spatial pat-terns of the products and their differences were also analyzed. Collectively, although six satellite AODs are in good agreement with ground observations, VIIRS DB (land and ocean) and MODIS MAIAC (land only) AODs show better validation metrics globally and better performance in 8/10 world regions. Therefore, they are more recommended for usage. Although land AE retrievals are not capable of quantitative application at both instantaneous and monthly scales, their spatial patterns show qualitative potential. Ocean AE shows a relatively high correlation coefficient with ground measurements (>0.75), meeting the fraction of expected accuracy (> 0.70). Error characteristic analyses em-phasize the importance of aerosol particle size and absorption-scattering properties for land retrieval, indicating that improving the representation of aerosol types is necessary. This study is expected to facilitate the usage selection of operating VIIRS and MODIS products and their algorithm improvements.

DOI:
10.1016/j.scitotenv.2023.163794

ISSN:
1879-1026