Employing relaxed smoothness constraints on imaginary part of refractive index in AERONET aerosol retrieval algorithm

Abstract. In the Aerosol Robotic Network (AERONET) aerosol retrieval algorithm, smoothness constraints on the imaginary part of the refractive index (IPRI) provide control of retrieved spectral dependence of aerosol absorption by preventing the inversion code from fitting the noise in optical measurements and thus avoiding unrealistic oscillations of retrievals with wavelength. The history of implementation of the IPRI smoothness constraints in the AERONET aerosol retrieval algorithm is discussed. It is shown that the latest version of the IPRI smoothness constraints, termed standard (STD) and employed by Version 3 (V3) of aerosol retrieval algorithm, should be modified to account for strong variability of light absorption by brown carbon (BrC) containing aerosols in UV through mid-visible parts of the solar spectrum. In V3 strong spectral constraints were imposed at high values of the Angstrom Exponent (AE; 440–870 nm) since black carbon (BC) was assumed to be the primary absorber, while the constraints became increasingly relaxed as AE deceased to allow for wavelength dependence of absorption for dust aerosols. The new version of the IPRI smoothness constraints assigns different weights to different pairs of wavelengths which are the same for all values of the Angstrom Exponent. For example, in the case of four wavelength input, the weights assigned to short wavelength pairs (440–675, 675–870 nm) are small (10^-6) so that smoothness constraints do not suppress natural spectral variability of the IPRI. At longer wavelengths (870–1020 nm), however, the weight is ten times higher to provide additional constraints on the IPRI retrievals of aerosols with high AE due to low sensitivity to aerosol absorption for longer channels at relatively low aerosol optical depths for these fine mode dominated aerosols. The effect of applying the new version of the IPRI smoothness constraints, termed relaxed (REL), on retrievals of single scattering albedo (SSA) is analysed for case studies of different aerosol types: BC and BrC containing fine mode aerosols, mineral dust coarse mode aerosols and urban industrial fine mode aerosol. It is shown that for BrC containing aerosols employing the REL smoothness constraints resulted in significant reduction, compared to STD, in retrieved SSA and spectral residual errors at the short wavelengths. For example, biomass burning smoke cases show a reduction in SSA and spectral residual at 380 nm is ~0.033 and ~17 % respectively for the Rexburg site and ~0.04 and ~ 12.7 % for the Rimrock site, both AERONET sites in Idaho, USA. For a site with very high levels of BC containing aerosols (Mongu , Zambia) the effect of modification in the IPRI smoothness constraints is minor. For mineral dust aerosols at small AE values (Mezaira site, UAE) the spectral constraint on IPRI was already relaxed in V3 therefore the new REL constraint results in minimal change. In the case of weakly absorbing urban industrial aerosols at the GSFC site, there are significant changes in retrieved SSA using the REL assumption, especially reductions at longer wavelengths: ~ 0.016 and ~0.02 at 875 and 1020 nm respectively for 440 nm AOD ~0.3. Comparison of aerosol parameters retrieved by inversion using STD and REL assumptions is presented. Analysis is done for retrievals utilizing the four standard AERONET wavelengths obtained at four AERONET sites: Rexburg, Mongu, Mezaira, and GSFC. The largest difference is found for the IPRI retrievals for BrC containing biomass burning (Rexburg) and urban industrial (GSFC) aerosols in which cases employing the STD assumption in the AERONET inversion was over constraining the spectral dependence of the IPRI. The modification of smoothness constraints on the IPRI has a minor effect on retrievals of other aerosol parameters such as the real part of refractive index and parameters of the aerosol size distribution. Both SSA retrieved using STD and REL assumptions were compared to SSA derived from in situ measurements collected during the DRAGON-MD field campaign in 2011. The DISCOVER-AQ column-integrated in situ aircraft SSA data for 550 nm were compared to AERONET retrievals at 440 nm and 675 nm which were interpolated to 500 nm and showed a closer agreement between in situ SSA and SSA retrieved from inversions employing the REL assumption than between in situ SSA and SSA retrieved using STD constraints. The implementation of the relaxed smoothness constraints on the imaginary part of the refractive in the next version of the AERONET inversion algorithm will produce significant impacts at some sites with changes up to ±0.033 and ±0.017 in short wavelength channels (380 nm and 440 nm) for some biomass burning smoke cases with significant BrC content and possibly up to ±0.015 in mid-visible channels (500 nm and 675 nm) to near IR channels (870 nm to 1020 nm) for some urban industrial aerosol types while still mostly within the uncertainty of the AERONET SSA retrievals, depending on AOD level, Angstrom Exponent and wavelength. For mineral dust aerosols the impact will be insignificant, while for biomass burning aerosol dominated by BC the changes will be relatively small.