We propose a general scheme for reconstructing vertical profiles of the scattering coefficient at the wavelength of λ = 0.52 μm within the altitude range up to 5 km. The external factors (season, type of air mass, and time of a day) as well as the data of measurements (near-ground values of the scattering coefficient, optical thickness, temperature, and relative humidity of air) are used as input parameters. The scheme assumes the account for correlation between the values of scattering coefficient at different altitudes, as well as the dependence of the height of mixing layer on the heating of the lower atmospheric layers. The errors of reconstructing the vertical profiles σd(H) and σω(H) are analyzed for different ways of taking into account the external factors and input parameters. It is shown that the use of such an approach provides a decrease in the error even at this stage approximately by 30% for winter and by 3 to 4 times for summer in comparison with the rms error of this parameter in the corresponding data arrays.