Advanced Algorithm for Aerosol Retrieval from Sentinel-2 Multispectral Instrument Data

Aerosols have been proven to be associated with the Coronavirus Disease (COVID-19). Satellite-derived aerosol optical depth (AOD) at a medium or high resolution e.g. 20 m may provide a new perspective for public health study, especially controlling the spread of COVID-19. Based on the Sentinel-2 MSI data, this paper uses the structure function method and the 6SV model to perform aerosol retrieval, and obtains the AOD retrieval result with a resolution of 20m in the study area. The results show that the distribution of surface features has important influence and correlation on the retrieval results, and the pixel interval, aerosol model and atmospheric model have obvious influence on aerosol retrieval. Moreover, the AOD inversion value is relatively higher and the error is larger, so the more data need to be collected later to improve and verify the algorithm.