Dust Production Mechanisms - Mineral Dust: A Key Player in the Earth System - page 109

These descriptions of the sandblasting processes are based on two main assump-
tions. The first assumption is that the impact velocity of the saltating particles
increases as a function of U * , such that the available kinetic energy increases as U * to
the power of two and as a function of the diameter of saltating particles. The second
assumption is that, like the erosion threshold, the dust binding energy increases
as the particle size decreases. As a result, these models predict a coarser dust
size distribution for weak U * and an increasing proportion of fine dust for higher
wind velocities. This behaviour is consistent with the difference in the dust size
distribution measured during intense dust emission events compared to background
conditions (d'Almeida and Schütz 1983 ; Gomes et al. 1990 ). Based on wind-tunnel
measurements, Alfaro et al. ( 1998 ) determined three typical modes for the size
distribution of emitted dust whose relative contribution depends on U * . For weak
events, only the coarse mode, having the lowest binding energy, is emitted. With
increasing U * , the relative proportion of the finest modes increases once the kinetic
energy of the saltating particles exceeds the binding energy associated with each
mode. This scheme predicts that the sandblasting efficiency depends on the soil size
distribution and on U * . The variation of the sandblasting efficiency with U * is very
sensitive close to the erosion threshold, but for high U * , it is relatively constant or
slightly decreasing for coarse soil size distribution (Alfaro and Gomes 2001 ). This
may explain the different behaviour of the experimentally determined sandblasting
efficiency (Gillette 1977 ). The capability of this scheme to reproduce mass dust flux
measurements has been tested by comparison with measurements obtained in Spain,
Niger (Gomes et al. 2003 ) and the USA (Nickling and Gillies 1989 ). As a result of
these studies, an adjustment of the binding energy initially determined from wind-
tunnel studies has been proposed (Alfaro et al. 2004 ).
Shao ( 2001 , 2004 ) proposed the most complete and physically explicit descrip-
tion of dust emission processes. His model takes into consideration the three
dust emission mechanisms: aerodynamic entrainment, saltation bombardment and
aggregate disintegration. It reflects the fact that dust emission is proportional to
saltation mass transport, but also depends on soil texture and soil “resistance”
expressed by the soil plastic pressure, p , characterising the flowing behaviour of the
soil subject to the particle impact. The bombardment efficiency, defined as the ratio
between the mass ejected by bombardment and the mass of impacting particles,
is estimated as a function of the soil bulk density and of the soil plastic pressure.
Both the range of bombardment efficiency and the soil plastic pressure have been
estimated from the data of Rice et al. ( 1995 , 1996 ). The amount of emitted dust is
assumed to be proportional to the amount of dust potentially available in the soil
and quantified by the soil texture. The size distribution of the emitted dust evolves
between two limits defined by the “undisturbed” soil size distribution and the “fully
disturbed” size distribution. This approach implies that during weak erosion events,
only emission of free dust is possible, while during strong erosion events, aggregated
dust can be released. The capability of this model to reproduce all the measurements
of mass dust emission fluxes available at this time (Gillette 1977 ; Nickling 1983 ;
Nickling and Gillies 1993 ; Nickling et al. 1999 ; Gomes et al. 2003 ; Rajot et al. 2003 )
has been evaluated by Shao ( 2004 ). They found that the model has the capacity to