Dust Production Mechanisms - Mineral Dust: A Key Player in the Earth System - page 108

seems relatively insensitive. In agreement with these measurements, Gillette and
Passi ( 1988 ) proposed a very simple expression to describe dust emission fluxes, F,
similar to the expression of the saltation flux but proportional to U * to the power
of 3-5. It includes an empirical proportionality coefficient, in which various wind
erosion factors are lumped together. This scheme is attractive due to its simplicity,
but the definition of the proportionality coefficient and the power of U * must be
empirically determined based on pre-existing knowledge of the soil and surface
properties. For a quantitative estimate, Marticorena and Bergametti ( 1995 )assumed
that, to first order, dust emission flux can be considered as a constant fraction of
the saltation flux depending on the amount of dust particles present within the
soil. The percentage of clay-size particles in the soil was used as a proxy for the
amount of potentially available dust in the soil. An empirical relationship between
the sandblasting efficiency and the soil clay content has thus been established by
fitting the sandblasting efficiency on the dataset from Gillette ( 1977 ) for soils having
clay contents ranging from 1.8 to 19 %. Average sandblasting efficiencies vary
from 10 4 to 10 6 cm 1 , highlighting the fact that only a tiny fraction of the
soil particles in motion near the surface is susceptible to be entrained in vertical
motion and long-range transport. This empirical approach captures two aspects of
the dust emission process: dust emission is proportional to saltation intensity, and
the sandblasting efficiency depends on soil properties. However, it oversimplifies
dust emission processes by assuming that the sandblasting efficiency only depends
on soil properties. A major limitation of this scheme is that it cannot provide size-
resolved estimates of dust emissions. Its main advantage is to provide reasonable
estimates of the dust mass fluxes as a function of soil type and U * (through the
saltation flux) when applied at the regional scale (e.g. Marticorena et al. 1997b ;
Laurent et al. 2005 , 2008 ).

5.5.2 Physically Based Models

Several approaches have been developed to describe dust emission processes though
bombardment by saltating grains. Shao et al. ( 1993 , 1996 ) considered the energy
balance of a saltating particle during its collision with the surface with the binding
strength of the dust particles, which was then estimated to be proportional to
U t . The process of dust emission is assumed as a linear combination of saltation
bombardment of various sand particle sizes. A specific problem with this scheme
is that the proportionality coefficients have been derived based on a small set of
idealised wind-tunnel experiments. Alfaro et al. ( 1997 , 1998 ) used a similar concept
to model sandblasting efficiency, but they estimated the binding energy of dust
particles based on wind-tunnel measurement performed with several soils from arid
or semi-arid regions. The measured dust size distribution and its evolution as a
function of U * were used as a constraint to determine the binding energy of each
mode composing the emitted dust size distribution.