Global OMI Aerosol Single Scattering Albedo evaluation using ground-based AERONET

Aerosol measurements are carried out worldwide in order to reduce the uncertainties about the impact of aerosols on climate. Over the past two decades, different methods (ground- or satellite-based) for measuring aerosol properties have been developed, covering a variety of approaches with different temporal and spatial scales, which can be considered complementary. Aerosol optical properties are essential for assessing the effects of aerosols on radiation and climate. Aerosol single scattering albedo (SSA), along with optical depth and asymmetry parameter, is one of the three key optical properties that are necessary for radiation transfer and climate models. At the same time, SSA strongly depends on different aerosol types, thus enabling the identification of these different aerosol particles. However, despite the strong need for aerosol SSA products with global and climatological coverage, and the significant progress in retrieving SSA from satellite measurements, the satellite SSA retrievals are still subjected to uncertainties.

In this study, we perform an evaluation of the OMAERUVd (PGE Version V1.8.9.1) daily L3 (1° x 1° latitude-longitude) aerosol SSA data, which are based on the enhanced two-channel OMAERUV algorithm that essentially uses the ultraviolet radiance data from Aura/Ozone Monitoring Instrument (OMI), through comparisons against daily SSA products from 541 globally distributed Aerosol Robotic Network (AERONET) stations for a 15-year period (2005-2019). The comparison is performed between the available OMAERUVd SSA data at 354 nm, 388 nm, and 500 nm, and the AERONET SSA data at 440 nm (or 443 nm). The comparison is made on an annual and seasonal basis in order to reveal possible seasonally dependent patterns, as well as on a climatological and a year-to-year basis. The statistical metrics, such as Coefficient of Correlation (R) and Bias, are computed for individual AERONET stations as well as for all stations. The effect of availability of common OMI and AERONET data pairs on the comparison is assessed by making comparisons when at least 10, 50 and 100 common pairs are available.

The results show that about 50% (75%) of OMI-AERONET matchups agree within the absolute difference of ±0.03 (±0.05) for the 500 nm OMI SSA and the 440 nm (or 443 nm) AERONET SSA. The corresponding percentage for the 388 nm OMI SSA and the 440 nm (or 443 nm) AERONET SSA increases to 58% (81%), while the corresponding numbers for the 354 nm SSA OMI and the 440 nm (or 443 nm) AERONET are 43% (67%). It is found that in overall, OMI tends mainly to overestimate (underestimate) SSA for the 500 nm (354 nm) products in comparison to AERONET 440 nm (or 443 nm) with a total bias of 0.025 (-0.024), or 2.7% (2.6%) in relative percentage terms with respect to AERONET (mean AERONET value equal to 0.908), and an overall R value of 0.399 (0.386). At 388 nm, OMI tends to retrieve higher SSA over regions where biomass burning occurs, against lower SSA values elsewhere, with overall bias and R values equal to -0.002 (0.22%) and 0.395, respectively.