Mechanically Generated Aerosols

Particles produced by mechanical processes tend to be larger than those resulting from gas-to-particle conversion. In general particles larger than a micrometer are mechanically formed by processes such as the wind erosion of soils, the bursting of bubbles in seawater, the shedding of plant fragments, etc. Although the relationship between the size of an aerosol particle and length of time it remains suspended in the atmosphere is complex, larger particles generally fall out of suspension more quickly than smaller ones (Fig. 1); hence large mechanically generated particles tend to have comparatively short atmospheric residence times.

Even though most mechanically generated aerosols are removed from the atmosphere close to their sources, some coarse particles remain suspended in the atmosphere for weeks and can travel thousands of kilometers before finally being deposited. While they are in suspension, aerosol particles can react with gases, with hydrometeors, and with other particles. As illustrated below, such reactions link the cycles of various atmospheric constituents in complicated ways.

The strengths of the various aerosol sources can be evaluated in several ways, and one of the most straightforward is to consider the mass of material injected into the atmosphere. As mechanical sources tend to produce physically and aerodynamically large particles, the importance of these sources is most evident when mass fluxes or related characteristics, such as particle volume, are being considered. In contrast, when evaluating source strengths with respect to the numbers of particles produced, the contributions from the mechanical sources tend to be less important compared with those producing numerous small particles via gas-to-particle conversion.

2 MECHANICALLY GENERATED AEROSOLS 195 0.0001 0.001 0.01 0.1 1 10 100 1000 10000


Accumulation Mode |


-Fine Particles

j-Gas-to-Particle Conversion (Secondary)-;

■Mechanical (Primary)



■Mechanical (Primary)

Gravitational Settling--

Removal Processes

Coagulation -f-






Particle Radius, micrometers

Figure 1 Characteristics of aerosol particles and the processes by which they are removed from the atmosphere.

Mineral Aerosol

The physical and chemical weathering of Earth's continental crust results in the production of mineral aerosol particles, commonly called atmospheric dust; and this represents one of the largest sources on a mass basis for natural particulate material in the atmosphere. Chinese records of dust storms date back thousands of years, and plumes of mineral aerosol over the oceans have been observed by mariners since humans have taken to the sea. Modern technology has shown that dust plumes over the oceans are among the most dramatic features seen in satellite images of aerosol optical depth (Husar et al., 1997).

Worldwide, about a third of Earth's surface can be considered potential sources for dust, but the arid and semiarid lands in Africa and Asia are the largest sources (Fig. 2). Climate clearly affects the amount of atmospheric dust produced. In general more dust is generated as the land becomes drier, but in hyperarid areas deserts can become "blown out" and less important as sources. Drought cycles also are linked to the emissions of desert dust. For example, studies at Barbados, an island in the North Atlantic Ocean, have shown that atmospheric dust concentrations increased during the Sahelian drought of the early 1970s (Prospero and Nees, 1977). Dust loads in the atmosphere also can vary over longer periods of time as a consequence of large-scale changes in climate and circulation. In this context, studies by An et al. (1990) suggest that patterns in dust deposition to the Chinese loess plateau over thousands of years can be linked to variations in the strength of the Asian winter monsoon.

The exact amount of dust injected into the atmosphere remains uncertain, but recent estimates are of the order of ~1500Tg/yr (Andreae, 1995; Tegen et al.,


40" 80"


40" 80"






V >



i St t





n / V

. s





160° 120° 80° 40° 0° 40° 80° 120° 160° 180° Figure 2 Sources for mineral aerosol (atmospheric dust). {From Pêwe, 1981.)

160° 120° 80° 40° 0° 40° 80° 120° 160° 180° Figure 2 Sources for mineral aerosol (atmospheric dust). {From Pêwe, 1981.)

1996). with an uncertainty of perhaps a factor of 2. As human activities have altered the globai landscape, a portion of the atmospheric dust load can be considered anthropogenic. For example, modeling studies by Tegen et al. (1996) indicate that 50% ±20% of the global dust flux may come from disturbed soils. On the other hand, efforts made by humans to reclaim some desert lands (Parungo et al., 1994) may have reduced the strength of natural dust sources.

Minera! particles are formed by a variety of processes, including grinding, weathering, abrasion, etc, (Pye, 1987). Once the particles arc formed, the wind deflates and disperses them, but other factors, such as the sizes and shapes of the particles, the roughness of the particle bed, the cohesiveness of the particles, ihe presence of cementing agents, the extent of vegetative cover, and especially the amount of soil moisture influence the credibility of the soils. Studies of the dynamics of the dust generation process by Gillette et al., (1974} showed that sandblasting of the soils was the dominant mechanism for mineral aerosol production by wind erosion, and these and other authors have shown that the wind velocity also shapes ihe size distributions of the suspended dust particles.

The transport of desert dust affects the global cycles of nitrogen, phosphorus, sulfur and various trace elements (Prospero, 1981; Schlcsinger et a!., 1990). Some areas of Earth benefit from the transport and deposition of mineral dust; for example, the fertility of the Loess Plateau in central China results in large measure from the accumulation of nutrient-rich mineral particles transported through the atmosphere from deserts in northwestern China (Liu et al., 1985). Similarly, dust originating from the Sahara Desert is transported through the atmosphere to the Central Amazon Basin where it supplies critical trace elements (Swap et al., 1992). Other parts of the continents are stripped of nutrients by the combined actions of wind and water erosion, and the economic consequences of erosion are substantial. For example, cost estimates for lost agricultural productivity, damage to waterways and infrastructure, and public health problems due to erosion by wind and water run into the billions of dollars for the United States alone (Pimentel et al., 1995).

The transport and deposition of mineral aerosol affects the cycles of a large number of trace elements in addition to those of N, P, and S. Bulk atmospheric dust particles generally have an elemental composition similar to that of average crustal rock (Rahn, 1976), and the composition of the ambient aerosol often is evaluated through "enrichment factors" (EFs) which are defined as

(X/Al)Aerosol (X/Al)Crast where X is any element of interest; Al is aluminum, a commonly used reference element; and the subscripts Aerosol and Crust refer to the aerosol sample of interest and the crustal reference material, respectively. Another commonly used reference element is Si, but Sc or a variety of other elements would serve the purpose equally well.

Weathered crustal material is the presumptive source for any element whose enrichment factor for a given sample approaches unity; those elements with EFs greater than ~5 have significant noncrustal sources. Direct comparisons of elemental ratios in aerosol samples versus crustal rock also show that the atmospheric loadings of mineral dust govern the concentrations of a large number of trace elements in the atmosphere (Table l). It is important to point out, however, that individual mineral dust particles can have a composition quite different from either the bulk dust or average crustal material (Anderson et al., 1996).

In a study of erodible soils, Schütz and Rahn (1982) showed the concentrations of most elements increased as particle sizes decreased to 20 to 50 (im radius, but the concentrations reached a plateau for particles less than 10 to 20 |im in radius. These authors predicted that some variability in the elemental composition of dust should occur near the desert source areas where a significant fraction of the particles would have radii > 10 |im. More than ~ 1000 km from the sources, however, the bulk of the particles would be < 10 |im in radius, and therefore these authors concluded that the elemental composition of dust transported long distances would be similar to that of the continental crust.

Dust particles in the atmosphere are far from inert, and reactions occurring on dust particles have significant implications for several important chemical cycles. Direct observations of individual particles showed sulfate coatings were present on >40% of the mineral dust particles collected over the North Atlantic at 25°N, and nitrate coatings were observed on >30% of the particles (Parungo et al., 1986). Further evidence for the uptake of gaseous sulfur species on dust from the Asia-Pacific region was obtained through statistical analyses of the elemental composition of aerosols (Winchester and Wang, 1990). Reactions between dust particles and gaseous nitrogen oxides have been reported from laboratory studies (Mamane and Gottlieb, 1992) and from analyses of ambient aerosols (Wu and Okada, 1994). The formation of nitrate on dust particles via heterogeneous reactions constitutes a sink

TABLE 1 Mass Ratios of Crustal Elements to Aluminum for High-Dust Events at Barbados, Bermuda, and Izana


TABLE 1 Mass Ratios of Crustal Elements to Aluminum for High-Dust Events at Barbados, Bermuda, and Izana






Average Crustal Rock"





6.8x 10~3


2.9x 10"1

3.3x 10-'




2.4x 10"4

2.4 xlO-4

3.Ox 10^4

1.2x 10"4


1.1 x 10"3

1.9x 10"3


4.4x 10~4


4.7x 1(T5

6.2x 10"5

7.1x 10"5

4.6x 10"5


2.2 x 10"5

2.3 xlO-5





6.1 x 10"1




5.2x 10"5

5.8x 10"5

8.1 x 10~5

7.2x 10"5








9.5x 10"3

1.2x 10~2

7.5x 10~3


l.lx 10°










1.2x 10-5

1.3x 10"5

1.4x 10-5

2.5 x 10-6



2.Ox 10"4

2.3x 10~4

1.4x 10"4


2.1 x 10~5

2.1 x 10"5




1.6x 10~5

1.5x 10 5

1.9x 10-5

8.Ox 10-6


1.6x 10-4

2.Ox 10-4

2.0 x 10"4

1.3x 10"4





7.5 xlO"4


4.2x 10"5

5.Ox 10~5

5.2x 10-5


"Taylor and McLennan (1985).

for nitrogen oxides, and reactions involving dust, N205, 03, and H02 radicals may affect the cycles of photochemical oxidants, leading to decreases in tropospheric ozone near dust sources (Dentener et al., 1996).

Mineralogical studies have shown that atmospheric dust consists of silicates (quartz and feldspars); clay minerals (e.g., kaolinite, smectite, illite, mica), carbonates (calcite and dolomite), and sulfur minerals (gypsum and anhydrite) (see Pye, 1987). Mineralogical analyses of aerosol particles by Zhou and Tazaki (1996) have provided independent lines of evidence for the chemical reactivity of atmospheric dust. Their analyses showed S-rich submicrometer particles frequently are found attached to mineral dust particles, and they inferred that H2 S04 reacted with calcite during transport to form gypsum.

The selective removal of dust particles as a function of particle size during transport probably has little effect on elemental composition, except perhaps for the rare earths (Sholkovitz et al., 1993), but size fractionation can lead to mineralogical differences among samples (Johnson, 1976). Glaccum and Prospero (1980) similarly suggested that the proportion of quartz particles relative to clay minerals should be low in dusts that have traveled long distances owing to the preferential fallout of the quartz particles, which tend to be aerodynamical ly large. Even so giant quartz particles ~50 |im radius have been found in the atmosphere over the central North Pacific, thousands of kilometers from their sources (Betzer et al., 1988). An unresolved paradox confronting atmospheric scientists is that the presence of such large particles so far from their sources is difficult, if not impossible, to explain based on our current understanding of transport dynamics.

Was this article helpful?

0 0

Post a comment