Dust Production Mechanisms - Mineral Dust: A Key Player in the Earth System - page 112

Alfaro and Gomes ( 2001 )orShao( 2001 , 2004 ) results from the assumption that
the impact velocity of saltating particles is proportional to U * , in disagreement
with some wind-tunnel results. Assuming a constant impact velocity of the saltating
particles, only a limited change in the kinetic energy of the saltation particles can
be expected as a function of U * , due to the enlargement of the horizontal flux
size distribution. Further investigation of the characteristics of individual particle
trajectories and of the size distribution of the saltation flux as a function of U * could
provide additional constrains for the description of the sandblasting process.
In essence, no definitive conclusion can be given on the capability of the
different dust emission models to reproduce the size-resolved measurements.
Whatever conceptual description is used (sandblasting, fragmentation), the link
between soil properties and dust size distribution is not sufficiently understood
and parameterised. All models include several input parameters, mainly related to
soil characteristics, which have not necessarily been experimentally determined,
such that they can be adjusted to fit the observations. Additional measurements of
size-resolved dust emission fluxes over large size spectra and for a wide range of
dynamical conditions with an extensive description of soil properties are necessary
to improve sandblasting models.

5.6 Conclusion

The understanding and modelling of dust emission processes have benefitted greatly
from a large body of theoretical and experimental studies on wind erosion processes
and, in particular, on saltation processes. The knowledge of these processes provides
some guidance in understanding the behaviour of dust sources and supporting
classifications of dust sources based on their geomorphological and pedological
characteristics (Callot et al. 2000 ; Bullard et al. 2011 ). Typically, from a physical
point of view, smooth surfaces with dry, sandy soils have the lowest erosion
thresholds and thus the highest probability to be frequent dust emission sources.
This explains why dust emission is frequent in sandy deserts such as Taklamakan
(Laurent et al. 2006 ; Bullard et al. 2011 ). On long timescales, such properties
may lead to soil depletion in dust-sized particles unless external sediment supply
is provided to the system. On the contrary, dust emission from rough surfaces,
characterised by high erosion thresholds, occurs less frequently since it requires
higher winds, but may be very intense. As an example, Laurent et al. ( 2006 ) found
that the annual dust emission from northeastern Asian deserts can be increased
by a factor of 2 due to a few unusually intense dust events originating from
the Gobi Desert. In terms of dust emission intensity, certain soil properties are
favourable. Because saltation is a prerequisite for intense dust emission, soils with
undisturbed particles in the optimum size range in terms of erosion threshold are
good candidates. But the intensity of dust emission is also influenced by the amount
of dust particles within the soil, as quantified by its fully dispersed size distribution
or its texture. This association of sand-sized ( 100 m) and dust-sized ( 1 m)