Evaluating the effects of columnar NO₂ on the accuracy of aerosol optical properties retrievals

Abstract. We aim to evaluate the NO2 absorption effect in aerosol properties derived from sun-sky radiometers as well as the possible retrieval algorithm improvements by using more accurate characterization of NO2 optical depth. For this purpose, we employ multiannual (2017–2022) records of Aerosol Optical Depth (AOD), Ångström Exponent (AE) and Single Scattering Albedo (SSA) collected by sun photometers at an urban and a suburban site in the Rome area (Italy) in the framework of both the AERONET and SKYNET networks. The uncertainties introduced in the retrievals by the NO2 absorption are investigated using high-frequency observations of total NO2 derived from co-located Pandora spectroradiometer systems as well as space-borne NO2 products from the Tropospheric Monitoring Instrument (TROPOMI). The correction is useful for lower AODs (< 0.3), where the majority of observations is found, especially under high NO2 pollution events. The analysis does not reveal any significant impact of the NO2 correction on the derived aerosol temporal trends for the very limited data sets used in this study. However, the effect is expected to become more evident for trends derived from larger data sets as well as in the case of an important NO2 trend. In addition, the comparisons of the NO2-modified ground-based AOD data with satellite retrievals from the Deep Blue (DB) algorithm of the NASA Moderate Resolution Imaging Spectroradiometer (MODIS) resulted in a slight improvement in the agreement of about 0.003 and 0.006 for AERONET and SKYNET, respectively. Finally, the uncertainty in assumptions of NO2 seem to have a non-negligible impact on the retrieved values of SSA at 440 nm leading to an average positive bias of 0.02 (2.5 %) in both locations for high NO2 loadings (> 0.9 DU).