Simplified and Fast Atmospheric Radiative Transfer model for satellite-based aerosol optical depth retrieval

Abstract

Satellite-based aerosol optical depth (AOD) retrieval over land remains a considerable challenge when high temporal and spatial resolutions are required. This paper presents the Simplified and Fast Atmospheric Radiative Transfer (SFART) model for direct calculation of satellite AOD via analytical equations rather than the lookup table (LUT) approach. The SFART model considers the impact of both Rayleigh and aerosol multiple scattering. A comprehensive comparison with the Simplified Method for Atmospheric Correction (SMAC) and single scattering approximation method is conducted. In validation with 6S, good atmospheric reflectance (Rayleigh + aerosol) accuracy is achieved by SFART, with approximately 69% of the data falling within the 5% estimated error (EE) envelope at both 440 and 640 nm. This is higher than the SMAC accuracy (within 5% EE: 42.83% and 39.91% at 440 and 640 nm, respectively) and is a tremendous improvement over the single scattering approximation method (within 5% EE: 15.67% and 20.5% at 440 and 640 nm, respectively). SFART is then applied for AOD retrieval to Himawari-8 satellite data over the North China Plain on both normal and hazy days. The retrieved AOD values are validated against collected AERONET data (Version 3, Level 2.0). Approximately 59% of the SFART AOD values fall within the EE bounds of ±(0.05 + 15%) with a root mean squared error of 0.22 (for 339 collocations). The promising results given by SFART indicate that this method can facilitate improved AOD calculation using empirical or real-time aerosol models in an efficient and flexible manner.