NO2 absorption correction for enhanced AOD retrieval

Aerosol Optical Depth (AOD) retrieval from sunphotometric measurements is sensitive to the concentration of atmospheric gases (e.g. NO₂), particularly in UV and lower visible spectral range. Current algorithms used in aerosol networks  either use  climatological NO₂ to estimate the corresponding absorption or it is totally ignored . NO₂ in the atmosphere is characterized by high spatial and temporal variations, especially in urban areas. Thus, climatological values are rarely representative of the actual NO₂ concentration, introducing non-negligible errors in AOD retrievals at specific spectral regions.

We propose a correction approach, using synchronous data from different networks/instruments. AOD is retrieved by sunphotometers (CIMEL and PREDE-POM) in AERONET and SKYNET networks. NO₂ total column is calculated by direct sun measurements of PANDORA spectroradiometers, part of PANDONIA network. Data from three stations, with colocation of these instruments are used in presented study to apply the correction and evaluate the new datasets. Two stations in Rome, Italy (Sapienza University at City Center and CNR-ISAC at Tor Vergata in suburban area) and one in Athens, Greece (National Observatory of Athens at city center). More specifically the NO₂ correction is applied on AOD at four bandwidths (340, 380, 400 and 440 nm). Propagation of the correction to the calculated Ångström Exponent is also estimated.

Highest mean relative differences are found at 440nm which are up to 1.7% for AERONET data and 5.3% at 400 nm for SKYNET (which’s algorithm does not consider NO₂). Highest absolute AOD difference found was 0.037 at 440nm. For Ångström Exponent 440-870 absolute maximum difference found was 0.31.Finally, cases of days with high NO₂ variability and the corresponding effect on AOD calculations will be presented.