Retrieval of aerosol optical depth from satellite observations: Accuracy assessment, limitations, and usage recommendations over South Asia

The evaluation of satellite-based aerosol optical depth (AOD) is essential for several applications such as air quality and climate studies. The AOD retrieved from satellite data is influenced by three major sources of uncertainties, that is, cloud detection, estimation of surface reflectance, and selection of aerosol models. This study suggested a framework for AOD evaluation using several statistical metrics by comparing satellite-based AOD with ground-based AOD measurements. AOD retrieved using the Dark Target (DT), Deep Blue (DB), and Multi-Angle Implementation of Atmospheric Correction (MAIAC) algorithms applied to data from the Moderate Resolution Imaging Spectroradiometer (MODIS), flying on the Terra and Aqua satellites, are evaluated over South Asia (SA). The results show that the MAIAC AOD outperforms DT and DB AOD in spatial coverage and retrieval accuracy. DT shows high retrieval accuracy over dark surfaces while DB underestimates the AOD with around 30% of the results below the expected error (EE=0.05 ± 0.2AOD). The regional evaluation of the retrieved AOD, following a standard procedure, is recommended before the data are used in any application.