REVIEW OF RESUSPENSION MODELS 219 



vegetation, the higher it stands above the ground, the more it reduces wind velocity near 

 the ground, and the lower the erosion. 



Gillette and his collaborators (1972; 1973; 1974; 1976) have been studying the 

 vertical flux of dust from agricultural soils under the influence of the winds. In a study of 

 particle size distribution at 1.5 and 6.0 m along with the horizontal velocity, it was 

 concluded that the upper limit of the ratio of the settling velocity to the friction velocity 

 for aerosols having the potential for long-range transport is approximately 0,2 or sUghtly 

 greater (Gillette and BUfford, 1974). Also, by studying the sizes of the aerosols produced 

 at different values of the friction velocity, it was concluded that the dominant injection 

 mechanism of soil aggregates less than 10 iim in diameter is not direct aerodynamic 

 entrainment (Gillette and Blifford, 1974). It was noted that the size distribution of the 

 vertical flux [expressed as dN/d (log r)] was proportional to r^ for particle sizes greater 

 than 2 pim. A similar particle size distribution was noted for the parent soil when the size 

 distribution was measured with liquid Freon dispersal. Since the dielectric constant of 

 liquid Freon is close to that of air, it tends to preserve the aggregate state of the soil. 

 However, if the aggregates are broken up, the number of particles less than 10 /im greatly 

 increases (Gillette and Blifford, 1974), showing once again the importance of aggregation 

 of the smaller particles in the soil. 



In a study of the vertical flux above several soils, Gillette (1974) noted that the 

 production of a flux of particles less than 10 Mm in size increased more rapidly with 

 friction velocity on a soU that had a higher percentage of silt and clay than it did on a soil 

 with a relatively low percentage of these materials. On both soils extrapolation of the 

 curves to the point of intersection indicates a threshold friction velocity of about 0.18 

 m/sec, about the same value as that for the impact threshold discussed earUer. For the 

 soils containing 3,5% clay (< 1 /jm) and either 0,5 or 1 .0% silt (<25 jum), the vertical flux 

 of particles < 1 jum increased as the ratio of the friction velocity to the threshold friction 

 velocity raised to the 5.14 power. For the soil containing 3,5% clay and 1% silt, the 

 vertical flux increased as the 9.62 power of the ratio of friction velocity to the threshold 

 friction velocity. This provides a model for the vertical flux of ,r 



Fa = Const. (u*/u*threshold) - (6) 



For the two soils evaluated, the constant at the point of intersection was about 

 6x 10^* cm^ see"' cm~^, or, assuming an aggregate density of 2.5 g/cm^ , about 

 1.5 X 10~^ g see"' m~^. An important conclusion from these studies was that the 

 increase in the vertical flux at powers of the friction velocity much greater than the 

 observed cube for the horizontal flux was due to the breakup of the aggregates by 

 sandblasting and release of the smaller particles to suspension. 



In a further study of the vertical flux resulting from eight different soils (Gillette, 

 1976), a regular pattern of increase with friction velocity appeared to exist with 

 considerable spread. A curve that I fitted by eye to the bulk of the data indicated an 

 average increase of vertical flux as the 5th power of the friction velocity with a constant 

 at the assumed threshold friction velocity as given by the earlier study of two soils. A 

 comparison of the size distribution of a parent sandy soil with the size distribution of the 

 horizontal flux to a height of 1 .3 cm indicated that the size distribution of the horizontal 

 flux is very similar to the size distribution of the parent soU aggregates. The aerosol at 1 

 m in height showed a mode for the particles greater than 20 /im centered around 50 jum. 

 As height increased, the particles less than 20 [im became an increasingly larger 



