Reevaluation of Mineral aerosol radiative forcings suggests a better agreement with satellite and AERONET data
Modelling studies and satellite retrievals do not agree on the amplitude and/or sign of the direct radiative perturbation from dust. Modelling studies have systematically overpredicted mineral dust absorption compared to estimates based upon satellite retrievals. In this paper we first point out the source of this discrepancy, which originates from the shortwave refractive index of dust used in models. The imaginary part of the refractive index retrieved from AERONET over the range 300 to 700 nm is 3 to 6 times smaller than that used previously to model dust. We attempt to constrain these refractive indices using a mineralogical database and varying the abundances of iron oxides (the main absorber in the visible). We first consider the optically active mineral constituents of dust and compute the refractive indices from internal and external mixtures of minerals with relative amounts encountered in parent soils. We then compute the radiative perturbation due to mineral aerosols for internally and externally mixed minerals for 3 different hematite contents, 0.9%, 1.5% and 2.7% by volume. These constant amounts of hematite allow bracketing the influence of dust aerosol when it is respectively an inefficient, standard and a very efficient absorber. These values represent low, central and high content of iron oxides in dust determined from the mineralogical database. Linke et al. (2006) determined independently that iron-oxides represent 1.0 to 2.5% by volume using x-Ray fluorescence on 4 different samples collected over Morocco and Egypt. Based upon values of the refractive index retrieved from AERONET, we show that the best agreement between 440 and 1020 nm occurs for mineral dust internally mixed with 1.5% volume weighted hematite. This representation of mineral dust allows us to compute, using a general circulation model, a new global estimate of mineral dust perturbation between -0.47 and -0.24 Wm-2 at the top of the atmosphere, and between -0.81 and -1.13 Wm-2 at the surface for both shortwave and longwave wavelengths. The anthropogenic dust fraction is thought to account for between 10 and 50% of the total dust load present in the atmosphere. We estimate a top of the atmosphere forcing between -0.03 and -0.25 Wm-2, which is markedly different that the IPCC range of -0.6 to +0.4 Wm-2 (IPCC, 2001). The 24-h average atmospheric heating by mineral dust during summer over the tropical Atlantic region (15° N-25° N; 45° W-15° W) is in the range +22 to +32 Wm-2 τ-1 which compares well with the 30±4 Wm-2 τ-1 measured by Li et al. (2004) over that same region. The refractive indices from Patterson et al. (1977) and from Volz (1973) overestimate by a factor of 2 the energy absorbed in the column during summer over the same region. This discrepancy is due to too large absorption in the visible but we could not determine if this is linked to the sample studied by Patterson et al. (1997) or to the method used in determining the refractive index.