A Landsat-5 Atmospheric Correction Based on MODIS Atmosphere Products and 6S Model

The Landsat satellite series represents the longest record of global-scale medium spatial resolution earth observations, and the utility of Landsat data for long-term and/or large-area monitoring depends on accurate and quantitative atmospheric correction to produce a consistently corrected surface reflectance (SR) dataset. In this study, we developed a rapid, automated atmospheric correction procedure based on Moderate Resolution Imaging Spectrometer (MODIS) atmospheric characterization products and the 6S (Second Simulation of a Satellite Signal in the Solar Spectrum) radiative transfer code for Landsat-5 Thematic Mapper (TM) imagery. Three MODIS atmosphere products at the resolution of 0.05°, MOD04, MOD05, and MOD07 were used as input to the 6S radiative transfer model in order to compute the parameters required for atmospheric correction, which were then used to correct TM imagery per-pixel automatically. This method was tested using five multi-date Landsat TM images in Beijing, China, and the atmospheric correction precision was assessed using ground-measured reflectance. The result showed that the SR retrieved from Landsat TM is consistent with the ground measurements, with a mean R² of 0.773 and a mean root mean square error value of 0.045.