Accuracy assessments of aerosol optical properties retrieved from Aerosol Robotic Network (AERONET) Sun and sky radiance measurements

Sensitivity studies are conducted regarding aerosol optical property retrieval from radiances measured by ground‐based Sun‐sky scanning radiometers of the Aerosol Robotic Network (AERONET). These studies focus on testing a new inversion concept for simultaneously retrieving aerosol size distribution, complex refractive index, and single‐scattering albedo from spectral measurements of direct and diffuse radiation. The perturbations of the inversion resulting from random errors, instrumental offsets, and known uncertainties in the atmospheric radiation model are analyzed. Sun or sky channel miscalibration, inaccurate azimuth angle pointing during sky radiance measurements, and inaccuracy in accounting for surface reflectance are considered as error sources. The effects of these errors on the characterization of three typical and optically distinct aerosols with bimodal size distributions (weakly absorbing water‐soluble aerosol, absorbing biomass‐burning aerosol, and desert dust) are considered. The aerosol particles are assumed in the retrieval to be polydispersed homogeneous spheres with the same complex refractive index. Therefore we also examined how inversions with such an assumption bias the retrievals in the case of nonspherical dust aerosols and in the case of externally or internally mixed spherical particles with different refractive indices. The analysis shows successful retrieval of all aerosol characteristics (size distribution, complex refractive index, and single‐scattering albedo), provided the inversion includes the data combination of spectral optical depth together with sky radiances in the full solar almucantar (with angular coverage of scattering angles up to 100° or more). The retrieval accuracy is acceptable for most remote sensing applications even in the presence of rather strong systematic or random uncertainties in the measurements. The major limitations relate to the characterization of low optical depth situations for all aerosol types, where high relative errors may occur in the direct radiation measurements of aerosol optical depth. Also, the results of tests indicate that a decrease of angular coverage of scattering (scattering angles of 75° or less) in the sky radiance results in the loss of practical information about refractive index. Accurate azimuth angle pointing is critical for the characterization of dust. Scattering by nonspherical dust particles requires special analysis, whereby approximation of the aerosol by spheres allows us to derive single‐scattering albedo by inverting spectral optical depth together with sky radiances in the full solar almucantar. Inverting sky radiances measured in the first 40° scattering angle only, where nonspherical effects are minor, results in accurate retrievals of aerosol size distributions of nonspherical particles.