Global error maps of aerosol optical properties: an error propagation analysis

Abstract. Among the numerous atmospheric constituents, aerosols play a unique role on climate, due to their scattering and absorbing capabilities, visibility degradation and their effect on incoming and outgoing radiation. The most important optical properties are the aerosol optical depth (AOD), the asymmetry parameter (g) and the single scattering albedo (SSA). Uncertainties in aerosol microphysics in global models, which in turn affect their optical properties, propagate to uncertainties on the effect of aerosols on climate. This study aims to estimate the uncertainty of AOD, g and SSA attributable to the aerosol representation in models, namely mixing state, aerosol size and aerosol associated water. As a reference, the monthly mean output of the general circulation model LMDz-INCA from the international comparison exercise AEROCOM B was used. For the optical properties calculations, aerosols were considered either externally mixed, homogeneously internally mixed or coated spheres. The radius was allowed to vary by ±20% (with 2% intervals) and the aerosol water content by ±50% (with 5% intervals) with respect to the reference model output. All of these possible combinations were assumed to be equally likely and the optical properties were calculated for each one of them. A probability density function (PDF) was constructed at each model grid point for AOD, g and SSA. From this PDF, the 1σ and 2σ uncertainties of the AOD, g and SSA were calculated and are available as global maps for each month. For the range of the cases studied, we derive a maximum 2σ uncertainty range in AOD of 70%, while for g and SSA the maxima reach 18% and 28% respectively. The mixing state was calculated to be important, with the aerosol absorption and SSA being the most affected properties when absorbing aerosols are present.