Wind Erosion
There is a fundamental difference between wind and water erosion, in that material eroded by water is travelling exclusively downslope and downstream towards the sea, whereas material entrained by the wind can travel in all directions. In practice, most coarse material detached by the wind is re-deposited locally, perhaps blocking local drainage lines and drifting along fence lines etc, while fine material (silt/clay) and organic dust is lifted into the atmosphere, where it may travel a long way, and generally diffuses down-wind from eroded areas to surrounding vegetated areas. This dust is observed to cross the Atlantic from the Sahara, so that the material is essentially lost to the source area. Of course there are massive sand accumulations in dunes, but in most of the study sites of DESIRE, this is not the main issue of wind erosion.
Our approach has been to simulate the mechanics of disturbance of the soil surface, estimating the frequency of disturbance as an index of the frequency of removal of the fine materials and organics that provide most of the fertility of fragile semi-arid soils that are prone to wind erosion (Visser and Sterk, 2007). To do this, we first estimate the critical velocity, for disturbance at the level of the soil surface roughness (10mm) as a function of monthly soil saturation deficit and soil surface grain size.
where D is the soil saturation deficit (mm) and d is the soil grain size (mm).
This expression shows a strong increase in the critical velocity for soils as they approach saturation, and is highly sensitive to grain size.
The wind speed profile empirically extends the normal logarithmic profile down through the vegetation to the surface roughness height (figure 2.3) , even though this procedure is thought to underestimate the importance of periodic velocity bursts in a sparse canopy (King et al 2008: Kenney et al, 2008). The wind speed at instrument height (say 2m), corresponding to this critical near-surface velocity is then calculated as:
Where z0 is the roughness height, derived from the vegetation cover. The frequency of wind speeds exceeding this value is estimated by fitting a gamma distribution to the wind velocity distribution. Where there is not adequate information on wind velocity distribution, values are transferred from neighbouring sites. In this approach, there is no attempt to estimate the volume of material removed by the wind, but to estimate the frequency with which surface fines are mobilised, relating this frequency linearly to the loss of fertility of the soil.
Although the theoretical principles of this work on wind erosion were proposed in the context of the DeSurvey project (15%), it has been brought to a concrete conclusion within the DESIRE project, and has not been reported in full previously.