Atmospheric correction algorithms comparison for Lake Water based on Landsat8 images

Atmospheric correction is one of the important issues that restrict the application of remote sensing monitoring in inland lakes. Therefore, In this paper, we taking TaiHu Lake, the freshwater lake in China as study area, using the 6S, FLAASH, QUAC and ACOLITE four atmospheric correction algorithms of Landsat8 images and discussing the applicability of four algorithms in TaiHu Lake. The results show that all the four algorithms can be used under the premise of low precision, but compared with the other three algorithms, the average relative error of the 6S algorithm is 31.98%, this show that 6S algorithm is suitable for TaiHu Lake in the four algorithms; Compared single band error, the results show that FLAASH, QUAC and 6S algorithm are similar in the visible wavelength band, their errors at the wavelengths of 561nm and 655nm are in the range from 14.43% to 26.06%. However, the accuracy of the 443nm, the 482nm and 865nm band is lower than that of the ACOLITE algorithm, so the corresponding atmospheric correction method needs to be selected according to the required band to obtain higher accuracy; The relative error between the 561nm band and the 655nm band is the minimum; The ACOLITE algorithm has higher accuracy in the band ratio. It is suggested that the ACOLITE algorithm should be chosen when establishing the inversion algorithm using the band ratio, which can effectively reduce the atmospheric impact.