Erosion by Wind


4.1 Introduction

Wind is a process which is much underestimated as an agent of erosion, to the extent that it is still considered to be a relatively insignificant process (Summerfield, 1989).

Wind erosion occurs when the air pressure of soil particles is sufficient to overcome the gravitational and cohesive forces resisting particle detachment. As surface roughness decreases, due for example, to overgrazing, this critical velocity for particle detachment can be obtained more easily. This explains the ready movement of particles in arid environments, where vegetation cover is sparse.

Any other process which reduces soil cohesion (e.g. drying of the soil or reduction in humus content) will also increase the susceptibility of the soil to wind erosion.

We should not forget about available potential energy within a geomorphic system; some areas are more prone to erosion by wind by virtue of their topographic characteristics (e.g. flat plains etc.) and therefore have limited environmental resistance to wind erosion, or limited capability to diffuse wind processes. This can promote soil erosion, independently of the other factors which relate solely to soil properties. This is easily seen in sand dune developmnt within the area of Quaternary Lake Chad and other Quaternary palaeolakes.

Detachment by wind erosion is essentially a two-stage event; (1) the static or fluid threshold applies to direct action of wind, causing detachment, and (2) dynamic or impact threshold, due to the bombarding effect of moving particles (i.e. saltation). Impact threshold are about 80 per cent of static threshold velocities. As with fluid flow, the critical velocity depends on the grain size of the particles, with finer and coarser particles requiring greater velocities for detachment. This explains the grain size range of loess deposits. There are, however important differences. The grains subject to eolian entrainment tend to be much smaller than those subject to entrainment by water.

Movement occurs in two ways; (1) in suspension (for smaller particles) and (2) by bedload transport by a series of jumps and creep [i.e. saltation - see 3.3 below].

4.2 The nature of wind movement

The motion of a air over a solid surface is characterised by the development of a boundary layer, wherin the flow is retarded by this surface. Within the boundary layer, there is a gradation from zero velocity of the fluid where it is in contact with the surface, to a velocity matching that of the external flow where it is no longer slowed by the frictional force at the surface (Fig. 3.). This velocity gradient produces shear within the boundary layer (Fig. 3. A).

When the velocity of the air beyond the boundary layer is steady, laminar flow may occur within the boundary layer. Laminar flow is most likely to occur only close to the solid surface where flow velocities are very low, or when the fluid is highly viscous. Under natural conditions involving atmospheric airflow, motion is not perfectly steady but gusty and the solid surface over which the airstream moves is generally irregular. Under such conditions, the flow develops turbulence. Turbulent flow is again controlled by the external forces which generate fluid motion, and again exhibits zero motion where the fluid is in contact with a solid surface. Notwithstanding the irregularities in this type of flow, average fluid velocities will reflect equilibrium velocity gradients just as in the case of laminar flow. However, with turbulent flow, the different mechanism involved in the exchange of momentum produces a different vertical distribution of flow velocity (Fig. 3.1B). In a turbulent boundary layer, there is more mixing of the flow. Thus the vertical gradient at the surface is steeper and shear stress is greater. Within a thick boundary layer, such as that occurring within the Earth's atmosphere above the land surface, turbulence in the airflow results from climatic conditions, changing meteorological conditions, and the topographic roughness of the surface. Within this boundary, however, subsidiary boundary layers may develop (e.g. over a sand sea, over dunes, over ripples, over grains).

4.3 Particle entrainment by wind

The shear velocity (or wind velocity at a particular height above a surface) required to initiate grain motion varies with the average size and sorting of the surface grain size population. Additional factors such as the density and shape of the grains and other surface roughness elements (e.g. vegetation) can be significant.

The threshold velocity of air necessary for detachment is governed by the equation;


where UT is the threshold velocity; A is a coefficient dependent on particle size; s is the specific gravity of the particles; d is the specific gravity of the air; D is the grain size diameter of the particles; and g is the gravitational constant (Bagnold, 1941).

Typically, the fluid threshold for roughly spherical quartz grains with a diameter of ~0.25 mm is achieved when the wind velocity at 0.01 m above the surface reaches 5.0 ms-1 (Bagnold, 1941). As wind velocity increases beyond the fluid threshold, grain motion becomes more pronounced.

Three distinct types of grain motion have been identified (Bagnold, 1941); creep, saltation and suspension.

Sand grains which move over a surface while staying in contact with other grains are said to creep. If they are lifted from the surface for a brief period, they are said to saltate (Fig. 3.2). In the case of a mixed grain size population, which is normal for most natural sediments, the larger grains tend to move by surface creep. However, individual grains may move by creep and saltation at different times. Bagnold (1941) estimated that 20-25 per cent of a graded sand population of mean diameter of 0.25 mm may move by surface creep.

When a surface particle is dislodged from the surface owing to the drag and lift forces exerted by the airflow or the momentum imparted by the impact of another moving grain, it travels on a characteristic path (Fig. 3.), which may reach 1-2 m above surface and extend several metres downwind. This excursion is a single saltation. About 75-80 per cent (by weight) of the grains moving across a surface during aeolian sand transport move by saltating. The typical saltation path of a particle results from the lift and acceleration generated by airflow. Initial lift may be generated by the reduced pressure immediately above sand grains protruding into the flow (Watson, 1969). Once dislodged from the surface, the grains move upwards at a steep angle. If the grains land on a surface of loose particles, it may impart its momentum to these grains causing them to creep or to be ejected from the surface and commence saltating. The transmission of momentum becomes an important factor as wind velocity decreases following grain transport. Particle movement continues even when shear velocity falls below that at the fluid threshold when grain motion was initiated. This is because the momentum imparted by saltation impacts can maintain sand movement at shear velocities below those at the fluid threshold. This shear velocity at which particle movement ceases is called the impact threshold velocity.

Grain motion in suspension occurs when grains which are lifted or ejected from the surface are subjected to sufficient lift and drage to overcome their tendency to fall back to the surface under their own weight. This occurs under conditions of highly turbulent airflow or when particles smaller than ~0.8 mm in diameter are mobilised. Suspended particle flow normally accounts for about 5 per cent of the material in motion. However, it represents the dominant mechanism for the generation of dust storms and the transport and deposition of aeolian silts (or loess).

3.4 Erosion processes associated with wind

Wind not only detachs and transports material but also erodes. Wind erosion takes place by (1) abrasion, (2) attrition, and (3) deflation processes. Abrasion is the wearing away of relatively solid rock or indurated material (e.g. soil, regolith), and results from the impact of wind-transported particles on a target. Like rainsplash, this involves the transfer of momentum (Greeley and Iversen, 1985). Generally speaking, the greater the particle size or impact velocity, the greater the abrasion. However, the actual abrasion process is more complex. Experimental studies by Marshall (1979) have shown that abrasion occurs as a sequential series of events, where, on impact, the particle tends to flatten slightly and to indent the impact surface. At critical limits, dependent on particle size or impact velocity (or a combination of both), a circular crack develops at the indentation site (Fig. 3.3), known as a Hertzian fracture, which has a radius of about 10%-30% greater than the contact radius (Greeley and Iversen, 1985). As particle size or impact velocity increases, the diameter of the fracture increases. Fractures produced by angular particles are different to those produced by rounded particles. Figure 3.4 shows the sequence of abrasion patterns resulting from the impact of angular particles.

Attrition is the set of processes which break, fragment or crush rock particles. Fragmentation can occur; (1) by breakage resulting from collisionof particles in transit in a fluid (i.e. air, water); (2) by fragmentation on impact cratering; and (3) by crushing and grinding of rocks and rock particles by mass movements and various other means (e.g. tectonic, glacial, etc.).

Deflation is the process of removal of loose particles by wind, regardless of the detachment process involved. This is a very localised process, in which particle removal by deflation is selective, and large (i.e. sand-sized or greater) particles tend to be transported on over short distances (i.e. a few metres). finer silt and clay-sized particles can be transported over very long distances, often over several hundreds of kilometres (Statham, 1977; Greeley and Iversen, 1985). Dust is the fine-grained component of erosion by wind. Dust can be used as a proxy index of erosion. This can be seen, for example, by increases in the frequency of dust storms in Mauritania (Fig. 3.5) (Middleton, 1991). This increase are correlated with changes in mean annual precipitation elsewhere in the Sahel (Fig. 3.6) and probably reflects a reduction in vegetation cover and greater particle entrainment by eolian processes (Goudie, 1983), despite no apparent surficial evidence of desertification in this area (Middleton, 1991).

Fig. 3.6 Annual variation in rainfall and dust storms at El Obeid, Sudan (After Middleton, 1991). Data on dust transport cannot be accepted with reservation (Middleton, 1991), particularly data gathered from meteorological observing stations, which are often based on visibility at eye level, with the result that it may be difficult to distinguish between dust in the atmosphere and fog or pollution. On the other hand, erratic observations at thes stations may underestimate the frequency of dust storms.

Dust can be derived from a variety of sources, including; (1) deflation of ergs (i.e. sand seas or sand sheets); (2) playa mud flats; (3) dune systems; (4) physical weathering of rocks and soils in hot climates (cf. Cooke and Warren, 1974); (5) physical weathering in cold climates (Embleton and King, 1976); and (6) human activities (e.g. traffic, overgrazing, burning etc.). During the Quaternary, the greater coverage of sand seas (Goudie, 1984) and vast areas covered by ice or affected by periglacial conditions resulted in more extensive erosion by wind than at present. In terms of dust movement, this is reflected in the vast accumulations of fine-grained wind-blown material (e.g. loess) which accumulated to great thicknesses in areas marginal to the ice sheets and in continental areas of Europe, Asia, and North America.

3.5 Accelerated weathering processes associated with wind action Table 3.1 Classification of weathering processes (After Cooke and Doornkamp, 1990).

1. Processes of disintegration

    (a) Crystallization processes
    Salt weathering (crystal growth, hydration, thermal expansion)
    Frost weathering

    (b) Temperature/pressure change processes
    Insolation weathering
    Sheeting, unloading

    (c) Weathering by wetting and drying
    Moisture swelling
    Alternate wetting and drying
    Water-layer weathering

    (d) Organic Processes
    Root wedging
    Colloidal plucking
    Lichen activity

2. Processes of Decomposition

    (a) Hydration and hydrolysis

    (b) Oxidation and reduction

    (c) Solution, carbonation, sulphation

    (d) Chelation

    (e) Biological chemical changes
    Micro-organism decay, bacteria, lichens

Weathering processes are a precursor to erosion, in that they can facilitate, either directly or indirectly, particle detachment. Weathering can be divided into two main categories of processes; (1) disintegration processes; and (2) decomposition processes (Table 3.1) which reflect the complexity of processes and agencies contributing to weathering (Fig. 3.7). Wind movement, for example, aids the operation of a variety of mechanical and chemical weathering processes, particularly under specific ground cover types and where there exist protruding surface obstacles (e.g. buildings).

This, however, is an oversimplification, in that many local factors will affect the efficacy of weathering processes, particularly chemical weathering, which also need to be considered (Fig. 3.9).

Most weathering processes depend on climatic conditions and, indeed, climatic parameters have been widely used to identify world-wide regions of chemical and physical weathering (Fig. 3.8), providing the foundation for much of the sub-discipline of climatic geomorphology (Stoddart, 1969). However, these relations (Fig. 3.8), when viewed in terms of Figures 3.7 and 3.9, are crude and over-simplified, particularly where they are related to present day climatic conditions.

Some attempts to refine these, for example, the use of N-values in southern Africa, with:


where Ej = potential evaporation for January (the warmest month) and Pa = annual precipitation (Weinert, 1965). Weinert's N values were established to define a weathering boundary for foundation and building materials, which matched the common behaviour of doleritic materials in southern Africa.

In addition to climate, many local variables may influence the effectiveness of weathering processes at a particular location, including (a) topographic location (including aspect and slope); (2) the drainage conditions; (3) nature of vegetation cover and animal life; and (4) the age of the surface (Birkeland, 1974; Gerrard, 1981). The effect of local variables is best seen in terms of effects on buildings., where aspect, exposure, microclimates and the availability of moisture accentuate the effectts of weathering, particularly at points of maximum exposure and where the building is in contact with the ground. Where there are large concentrations of buildings (and associated industrial activity, automobiles, etc.), these effects can be accelerated. Rapid increases in the size of urban area and their degree of industrialisation over the past two hundred years has, in addition, led to vast increases in the concentration of air-borne pollutants, particularly those associated with petrochemicals (Goudie, 1984). In addition to obvious effects on air quality, these can accelerate physical weathering and contribute specific types of chemical weathering which lead to rapid breakdown of building materials (Douglas, 1983).The most important of these include (1) salt weathering and (2) chemical weathering.

These potential effects are shown, for example, by weathering gradients in south east England, where weight loss from blocks of Portland stone, exposed for 10 years was almost double in urban areas relative to rural areas (Honeyborne and Price, 1977). Further work by Jayne and Cooke (1987) confirms London's urban-rural weathering gradients and shows that they are quite complex, reflecting in part the complexity of atmospheric pollution in a large urban area (Cooke and Doornkamp, 1990).

These studies suggest a gross amplification of weathering processes which can be positively correlated with greater concentrations of air-borne pollutants in urban environments (Douglas, 1983; Cooke and Doornkamp, 1990). Weathering is principally accelerated by modifying the composition of the boundary layer of the atmosphere with SO2 and NOx gases, which has the effect of actiivating a variety of processes, notably (1) solution; (2) sulphation; and (3) salt weathering. When these processes are combined with the complex edaphic architecture of urban areas and multinodal heat flows, this produces ideal conditions for local and regional intensification of weathering processes.

Salt weathering (Goudie et al., 1970) occurs because wind movement, coupled with low precipitation, promotes high rates of evaporation, leading to the accumulation of a variety of salts on or near the surface or rock or soil (Cooke and Doornkamp, 1974). Salt weathering occurs (1) associated with pressures accompanying crystallisation; and (2) by hydration (where salt minerals expand when water is added to their crystal structure) (Goudie et al., 1970; Goudie and Wilkinson, 1977).

3.6 Acceleration or concentration?

Given that weathering is essentially a process which allows rock breakdown, erosion processes (either by water or wind) are also likely to be accelerated in these environments, simply because more materials are available for transport. There are, however, some outstanding problems with recognition of accelerated erosion in these environments; (1) are processes accelerated or simply concentrated, due largely to the nature of the artifical boundary conditions which prevail in urban areas; and (2) is it arbitrary to define norma versus accelerated when the principal differences reflect the nature of artificial (e.g. building stone) versus natural (e.g. soil) ground cover, in that the latter will produce a more dynamic and effective protection to erosion processes, partiicularly when vegetated.

3.7 References

Bagnold, R.A. 1941. The Physics of Blown Sand and Desert Dunes. Chapman and Hall, London.

Beckedahl, H.R., Bowyer-Bower, T.A.S., Dardis, G.F. and Hanvey, P.M. 1988. Geomorphic effects of soil erosion. In: Moon, B.P. and Dardis, G.F. (eds), The Geomorphology of Southern Africa, Southern Book Co.,Johannesburg, 249-277.

Birkeland, P. 1974. Pedology, Weathering and Geomorphological Research. Oxford University Press, New York.

Cooke, R.U. and Doornkamp, J.C. 1974. Geomorphology in Environmental Management.

Cooke, R.U. and Doornkamp, J.C. 1990. Geomorphology in Environmental Management. (2nd Edition).

Cooke, R.U. and Warren, A. 1974. Geo-morphology of Deserts. Batsford Educational, London.

Cryer, R. 1986. Atmospheric solute inputs. In: Trudgill, S.T. (ed.), Solute Processes, 15-84. Wiley, Chichester.

Dardis, G.F., Beckedahl, H.R., Bowyer-Bower, T.A.S. and Hanvey, P.M. 1988. Soil erosion forms in southern Africa. In:Dardis and Moon, B.P. (eds), Geomorphological Studies in Southern Africa, A.A. Balkema, Rotterdam, 187-214.

Douglas, I. 1983. The Urban Environment. Arnold, London.

Embleton, C. and King, C.A.M. 1975. Periglacial Geomorphology. Arnold, London.

Gerrard, A.J. 1981. Soils and Landforms. Allen & Unwin, London.

Goudie, A.S. 1981. The Human Impact. Blackwell, London.

Goudie, A.S. 1983. Dust storms in space and time. Progress in Physical Geography, 7, 502-530.

Goudie, A.S. 1984. The Nature of the Environment. Blackwell, Oxford.

Goudie, A.S. and Wilkinson J.C. 1977. The Warm Desert Environment. Cambridge University Press, Cambridge.

Goudie, A.S., Cooke, R.U. and Evans, I. 1970. Experimental investigation of rock weathering by salts. Area, 42-48.

Heathcote, R. 1983. The Arid Lands: Their Use and Abuse. Longman, London.

Honeyborne, D.B. and Price, P.B. 1977. Air pollution and the decay of limestones. Building Research Establishment, Note 117/77.

Hudson, N. 1981. Soil Conservation. Batsford Academic and Educational, London.

Hudson, N. 1992. Land Husbandry. Batsford Academic and Educational, London.

Jayne, S. and Cooke, R.U. 1987. Stone weathering in south-eastt England. Atmospheric Environment, 21, 1601-1622.

Middleton, N. 1991. Desertification. Oxford University Press, Oxford.

Morgan, R.P.C. 1986. Soil Erosion and Soil Conservation. Longman, London.

Rowntree, K.M. 1988. Equilibrium concepts, vegetation change and soil erosion in semi-arid areas: Some considerations for the Karoo. In:Dardis, G.F. and Moon, B.P. (eds), Geomorphological Studies in Southern Africa, A.A. Balkema, Rotterdam, 175-186.

Spooner, B. 1987. The paradoxes of desert-ification. Desertification Control Bulletin, 15, 40-45

Statham, I. 1977. Earth Surface Sediment Transport. Oxford University Press, Oxford.

Stiles, D. 1987. Camel vs. cattle pastoralism: stopping desert spread. Desertification Control Bulletin, 14, 15-22.

Summerfield, M.A. 1989. Global geomorphology. Longman, London.


Copyright 2005-2009  Sandman LLC.  All rights reserved.