Dust Deposition - Mineral Dust: A Key Player in the Earth System - page 192

Fig. 8.6 Intercomparison of annual dry and wet deposition masses for mineral dust as simulated
by different global dust models. Experiment A shows simulations performed in the nominal
conditions for each model. Experiment B shows results for simulations performed by the same
models but forced with identical dust emissions, size distribution, and injection height (derived
from the AeroCom data base at http://nansen.ipsl.jussieu.fr/AEROCOM ) . Arrows and associated
numbers indicate the ratio between highest and lowest model values for each parameter of interest
in the annual dry deposition mass between the two extreme models are reduced
to a factor of 1.9 compared to Experiment A (3.7). This suggests that part of the
differences between models in Experiment A could be due to differences in the
dust size distributions. However, when looking at the annual wet deposition masses,
the situation is reverse: the maximum difference between the simulated annual wet
deposition masses is only a factor of 2.4 in Experiment A and almost tripled in
Experiment B (6.8).
These results from the AeroCom dust model intercomparison exercise clearly
suggest that the terms of the dust cycle are not sufficiently constrained to pro-
vide accurate estimates of the dust mass budget. As a general trend, the tested
models simulate the climatology of vertically integrated parameters (AOD and
Angström exponent) within a factor of two, whereas the total deposition and
surface concentrations are reproduced within a factor of 10 (Huneeus et al. 2011 ).
This points to large uncertainties in dust emissions, size distribution, and wet
and dry deposition. Nevertheless, it should be recalled that AOD and Angström
exponent are largely determined by the submicrometer fraction of aerosol and,
therefore, model differences in AOD do not reflect discrepancies in the supra