2014 , Volume 19, № 3, p.48-56

When solving problems for atmospheric correction of satellite images of the Earth surface in the visible and UV spectral ranges, it is important to consider such factors as multiple scattering, absorption of radiation, adjacency effect, and re-reflection. One of the approaches to solve these problems is connected with the solution of the stationary radiation transfer equation. It has no general analytical solution. Therefore, some of the above-listed factors are disregarded in widely used algorithms of atmospheric correction. In the present paper, an algorithm for correction of the aerospace images due the distorting effect of the atmosphere is suggested. It takes into account all main processes important for the formation of images (based on the solution of the radiative transfer equation by the Monte Carlo method) and involves a number of computer procedures which considerably reduce the computational time (by a factor of 6-10). Particular example for processing of the actual image of a fragment of the western coast of Africa with coordinates 27.5-27.90 N and 13.4-12.90 W (measurements were performed on August 19, 2011 at 11h 25 min) was considered. The image was recorded by MODIS in the third spectral channel (0.469 m). The correction for the distorting effect of the atmosphere was carried out using the suggested algorithm along with the algorithm of uniform correction and the MOD09 algorithm developed by NASA. A comparison demonstrates that in the considered situation, the application of the algorithm of uniform correction for the distorting effect of the atmosphere is sufficient, and the correlation coefficient between the reflection coefficients of the Earth surface computed by the suggested algorithm and the MOD09 algorithm makes 0.984. Results of the reconstruction of the reflection coefficients using the suggested algorithm are on average by 0.023 higher than the results obtained with the help of the MOD09 algorithm. This can be explained by the difference in the optical models of the atmosphere employed to solve the radiation transfer equation. On this basis, it is possible to assert that the suggested algorithm yields reliable results and can be used to correct aerospace images for the distorting effect of the atmosphere in the visible range of wavelengths.