Abstract. The atmospheric aerosol properties were obtained within the framework of the AERONET program at the Moscow State University Meteorological Observatory (Moscow MSU MO) over the 2001-2014 period. The quality data control has revealed the necessity of additional cloud screening and NO 2 correction. The application of additional cloud screening according to hourly visual cloud observations provides a decrease in monthly average aerosol optical thickness (AOT) at 500 nm of up to 0.03 compared with the standard data set. We also show that the additional NO 2 correction of the AERONET version 2 data is needed in large megalopolis, like Moscow, with 12 million residents and NO x emission rates of about 100 kt yr −1 . According to the developed method, we estimated monthly mean NO 2 content, which provides an additional decrease of 0.01 for AOT at 340 nm, and of about 0.015 -for AOT at 380 and 440 nm. The ratios of NO 2 optical thickness to AOT at 380 and 440 nm are about 5-6 % in summer and reach 15-20 % in winter when both factors have similar effects on UV irradiance. Seasonal cycle of AOT at 500 nm is characterized by a noticeable summer and spring maxima, and a minimum in winter conditions, changing from 0.08 in December and January up to 0.3 in August. The application of the additional cloud screening removes a local AOT maximum in February. Statistically significant negative trends in annual AOT for UV and mid-visible spectral range have been obtained both for average and 50 % quantile values. The pronounced negative changes were observed in most months with the rate of about −1-5 % yr −1 and could be attributed to the negative trends in emissions (E) of different aerosol precursors of about 135 Gg yr −2 in E SO x , 54 Gg yr −2 in E NMVOC , and slight negative changes in NO x over the European part of Russia. No significant influence of natural factors on temporal AOT variations has been revealed.