Atmospheric correction of high resolution multi-spectral satellite images using a simplified method based on the 6S code

A set of high-resolution multi-spectral satellite images were collected and processed to estimate the sediment concentration in the sea-breaking zone. The value of Total Suspended Matter (TSM) is empirically related to the sea surface reflectance. However, in order to obtain accurate reflectance measurements from the satellite images, the effect of the atmosphere needs to be accounted for. A set of satellite images from the SPOT/HRVIR and Terra/ASTER sensors were collected for the study area, a section of the northwest coast of Portugal around Aveiro. The images were atmospherically corrected using a method combining 6S simulations and ground horizontal visibility measurements. The method uses a reference scenario based on the typical values for each image of the following parameters: solar zenith angle, viewing zenith angle, viewing azimuth angle, ground height, and ground horizontal visibility. Some of these parameters are nearly constant for the whole scene. Each of the relevant parameters is allowed to vary within a reasonable range around its reference value. The 6S radiative transfer code is used to generate bottom of the atmosphere reflectance values as a function of the top of the atmosphere recorded reflectances, for each of the parameters. Suitable atmospheric and aerosol models are selected according to the meteorological auxiliary data available. For each pixel in the image, an atmospheric corrected reflectance is obtained as a sum of the individual corrections due to each of the parameters. Sea sand locations were used as test sites for an evaluation of the atmospherically corrected reflectance accuracy.