Efficiency estimation of four different atmospheric correction algorithms in a sediment- loaded tropic lake for Landsat 8 OLI sensor

Atmospheric correction algorithms allow the reduction of atmospheric components influence in the acquisition of earth’s surface reflectance properties. Quantifying this reduction is critical for the remote sensing science. Different regions have different natural or man-made composition and respond differently for each algorithm. Highly turbid inland waters are significantly sensitive to atmospheric correction algorithms, and these waters must be evaluated with care. This paper investigated the impact of turbidity levels in a sediment-laden tropical lake through the correlation analysis between image and ground-based Remote Sensing reflectance (Rrs) for the Landsat 8 OLI sensor. Four atmospheric algorithms were tested for Rrs estimation: DOS (Dark Object Substraction), FLAASH (Fast Line-of-sight Atmospheric Analysis of Spectral Hypercubes), 6S (Second Simulation of the Satellite Signal in the Solar Spectrum) and QUAC (QUick Atmospheric Correction). A regression linear model was applied to the data, and results showed two distinguishable features within the samples. The first feature set presents a near-one higher-angle slope value with R 2 values ranging from 49-64%, and a near-zero lower-angle slope with R 2 ranging from 49-71%. Despite the small data set used in this work, it is reasonable to assume the results demonstrate that, during the atmospheric correction process, the Rrs correlation undergoes a slope direction change. This is most likely due to the influence of higher turbidity levels.