Аннотация:Nitrogen oxides NOx=NO+NO2 are one of the key indicators of anthropogenic air pollution, since their main source is high-temperature fuel combustion at industrial and transport facilities. Along with surface measurements of NO and NO2 concentrations, observations of NO2 content in the lower troposphere by multi-angle differential absorbtion spectroscopy (MAX DOAS) are gaining ground. Previously, when solving the problem of estimating the vertical profile of an impurity from such measurements, Bayesian estimation methods were used. However, in most cases, information is not enough to correctly set the NO2 random distribution function, which leads to a scatter in the results obtained by different groups. To avoid this, this paper proposes another approach, which consists in narrowing the class of possible profiles based on the assumption of the number of possible extrema in the reconstructed NO2 vertical distribution.