/ Particles which V \contain main aerosol mass'

103 102 101 10° 101 102

Fig, 22

Importance of particle size for various fields of meteorology (Junge, 1963). (By courtesy of Academic

Press and the author)

103 102 101 10° 101 102

Fig, 22

Importance of particle size for various fields of meteorology (Junge, 1963). (By courtesy of Academic

Press and the author)

First, aerosol particles are important from the point of view of atmospheric electricity. A fraction of the air molecules is electrically charged (small ions), as a result cf ionizing radiation . Measurements show (Bricard and Pradel, 1966) that in 1 cm3 of surface air about 10 ion-pairs are formed each second. Eight out often are produced by the radioactivity of the air and soil, while the other two are produced by cosmic radiation. The elcctrical properties of the air are determined by the clectrical mobility (B) of ions formed:

where £ is the potential gradient, while v is the ion velocity. Small ions of opposite sign can combine with one another, or they can coagulate with existing aerosol panicles. The ion mobility decreases as a result of coagulation, since the mass of aerosol particles is much greater than that of air molecules. For this reason there is an inverse relation between the aerosol concentration and the electrical mobility of the air.

It is well known that particles suspended in a gas absorb and scatter radiation. These processes depend on the size, form and composition of the particles as well as on the wavelength of the radiation. If the particle size is much smaller than the wavelength, the scattering is described by the Rayleigh theory. However, if the two parameters are of comparable magnitude the theory developed by Mie must be applied (see Friedlander, 1977). Mie scattering is particularly important in the visible range. Since the wavelength of visible light coincides with the size of the large panicles, these play an important role in the control of the atmospheric visibility. Fogs and clouds are the only other major controlling faclors over most of the Earth: their particle size is larger, but they exert their influence by virtue of much higher mass concentrations5. Owing to the above extinction process, large particles also play a part in the regulation of the radiation and heat balance of the Earth-atmosphere system (see Chapter 6).

Another important property of aerosol particles is their role in the atmosphere as condensation nuclei for water vapour. Potentially, all particles may be condensation centres. However, under normal atmospheric conditions only a proportion of the particles take part in the formation of clouds. According to recent studies of Twomey (1971 and 1972) the majority of active condensation nuclei consists of ammonium sulfate and have a radius of the order of 10"2 /«m (see Subsection 5.3.1).

It follows from this discussion that the study of properties and effects of atmospheric aerosol particles exceeds the scope of air chemistry. Considering the complexity of the problem, we restrict our discussion in the following to the presentation of formation processes and principal physical and chemical properties of background aerosol in the troposphere and stratosphere.

5 In this book, except for the discussion of removal process in Chapter 5 we always deal with unsaturated air, that is. with "dry aerosol '.

42 Origin of atmospheric aerosol particles 4.11 Dispersal of particles of surface origin

The majority of atmospheric particulate matter arises from two basic processes6:

(a) dispersal of materials from the Earth's surface;

(b) chemical reaction and condensation of atmospheric gases and vapours.

The dispersal of surface materials produces particles in two major categories: sea salt and soil or mineral particles.

Sea salt particles can be formed by the direct dispersal of the ocean water from the foam of the waves. However, these particles are generally too large to remain airborne, even after evaporation of the water. A much greater number of particles is produced by the bursting of gas bubbles reaching the water surface. According to the laboratory work of Moore and Mason (1954) this process takes place in two stages. In the first stage, when the bubble arrives to the surface, small particles are ejected from the bursting water film. In the second stage, a thin jet is formed by the water flowing into the cavity remaining in the surface after the rupture. The particles formed in the second stage are less numerous and their size are in the giant range (Woodcock, 1953).

The sea salt particles produced in this way are composed mostly of sodium chloride, which reflects the composition of sea water. Among other substances, marine particulate matter also contains a large amount of sulfates (see Subsection 3.6.2). Furthermore, during their rise through the water, bubbles scavenge a lot of surface active organic materials which are partly injected into the air when the bubbles burst (see Subsection 3.3.3).

Woodcock (1953) as well as Moore and Mason (1954)demonstrated that the rate of bubble formation increases with increasing wind speed. In more recent work A. Meszaros and Vissy (1974) reported that over the oceans the correlation between the number of sea salt particles and the wind speed becomes gradually weaker as the particle size decreases. Thus, it is not excluded that the smallest sea salt particles (r <0.3 /an) originate from a type of bubbles the formation of which is independent of the wind speed.

The relation between the bubble size and the number of airborne particles produced upon bursting was studied by Day (1963) during his laboratory investigation. He pointed out that the number of particles increases with increasing bubble size. A bubble with a size of several millimeters forms some hundreds of particles when it bursts. On the basis of his atmospheric observations made in Hawaii, Blanchard (1969) speculated that the intensity of sea salt particle formation

6 Such processes as particlecoagulation may also produce new large particles, but not new particulate matter.

is between 25-100 cm 2 s " 1 at the surface of the ocean. This range is in a good agreement with the laboratory results of Moore and Mason (1954).

Airborne sea salt particles are transported to higher levels and over the continents by atmospheric motions. Because of the relationship between relative humidity and particle size (see Section 4.5), low relative humidity promotes the transport of sea salt particles.

The other category of particles arise from the solid surface of the Earth. This dispersal is obviously due to the effect of wind on rocks and soils. A well-known and highly visible example of this process is the formation of dust clouds and storms. However, the quantitative explanation of this particle production mechanisms is not easy, except when some external mechanical force agitates the surface (vehicles, animals, people etc.). The main reason for the difficulties in the explanation is the decrease of the wind speed with decreasing height above the surface, usually extrapolating to zero wind speed at the surface. It is believed that turbulent flow is necessary (see Twomey, 1977) for the detachment of grains. According to the most acceptable estimates the global strength of this source is (100-500) x 106 t yr"1 (SMIC, 1971 )7.

An important proportion of mineral particles produced by wind erosion is insoluble in water. The particles are composed of silicates (Junge, 1963). Their radius is generally greater than 0.1-0.5 ¿¿m. Some particles are removed from the air in the vicinity of sources, while another fraction is transported at great distances. This has been proved by the analyses of snow and ice in Greenland, which reveals potassium and calcium in concentrations that cannot be interpreted by the effect of maritime particles (Junge, 1963). Furthermore, particulate matter collected over the Atlantic Ocean contains a significant quantity of Saharan du«t under some conditions (Junge and Jaenicke, 1971); in fact, such dust particles were collected and identified even over the West Indies (Prospero, 1968; Blifford, 1970).

4.2.2 Formation of atmospheric aerosol particles by chemical reaction and condensation

Particles formed by the dispersal of surface materials generally have radii larger than about 0.1 /urn. This means that Aitken-size particles must be produced by another mechanism, namely by condensation of vapours, preceded in many cases by gaseous chemical reactions. These reactions are generally initiated by photochemical processes.

Thus, a large set of data obtained by McWilliams (1969) in clean air (W. Ireland) by means of expansion chambers showed that the concentration of Aitken particles is lower in the winter than in the summertime. Furthermore, McWilliams' observations also demonstrated that more aerosol particles can be detected during daylight than at night. This finding was confirmed by the investigations of Vohra et

7 SMIC: Study of Man's Impact on Climate.

al. (1970) and A. Mészáros and Vissy (1974) according to which, in a clean maritime environment, the number of Aitken particles is at a maximum during the afternoon. It was also shown by atmospheric measurements (e.g. Lopez et al., 1973) that after sunrise the aerosol concentration increases, which also points in the direction that particles with radii smaller than 0.1 ftm are produced by photochemical reactions. Recently, Hogan and Bernard (1978) have reported that over Antarctica there is a steady increase in the concentration after astronomical sunrise. Moreover, in Antarctic winter very small concentrations can be measured.

The formation of aerosol particles from gaseous components is appropriately investigated under laboratory conditions. In so-called aerosol chambers an artificial atmosphere is created to which small quantities of appropriate trace gases (e.g. SO,, NO,, H,0, NH, and organics) is added. It is also possible to use ambient air purified from particulate matter. The chamber may be illuminated to initiate photochemical processes, and the behaviour of particles formed is studied by the methods outlined in Subsection 4.1.2, e.g. by electrical mobility analyzers (Whitby ci aL 1972).

An important result from aerosol chamber studies was the discovery of the indirect photochemical process. Thus, Bricard et al. (1968) found that intense aerosol particle production can be observed in the chamber in the dark if ambient filtered air is sampled from a sunlit atmosphere. It is speculated that in the atmosphere some gaseous substance is excited by sunlight and is not collected by the filter used to obtain air which is free of aerosol particles. In the chamber these photochemically excited molecules initiate secondary thermal reactions leading to the formation of some supersaturated vapour (e.g. H,S04) which subsequently condenses (see also Subsection 3.6.3).

Fig. 23

Aerosol formation in art, irradiated chamber (Friedlander, 1978). N : particle number; V: volume of particles; A: total particle surface. (By courtesy of Atmospheric Eiuironmem)

Fig. 23

Aerosol formation in art, irradiated chamber (Friedlander, 1978). N : particle number; V: volume of particles; A: total particle surface. (By courtesy of Atmospheric Eiuironmem)

It was also demonstrated by aerosol chamber investigations that the behaviour of particles formed by condensation varies as a function of time. Figure 23 reproduces schematically the change in particle number (N), particle surface (A) and particle volume (V) according to Friedlander (1978). The curves are based on irradiation chamber experiments by Husarand Whitby (1973). It can be seen that three separate domains are identified. In domain I the formation of new particles is the dominant process. In this stage the number, surface and volume of particles all steadily increase. With increase in particle number (domain II) coagulation becomes more and more important (see Subsection 4.1.1). When the N curve has a maximum the coagulation loss just balances the particle formation rate. In domain III coagulation and condensation of the vapour on existing particles are the dominant processes. The number concentration decreases in this time interval, while particle volume further increases. The value of the surface area remains approximately constant.

A good atmospheric example of the above particle formation is the production of sulfate particles from gaseous precursors (see Subsection 3.6.3). It is further believed that the smaller organic particles discussed in Subsection 3.3.3 are also formed by gas-to-particle conversion. However, a large quantity of aerosol particles can form during the cooling of vapours with low saturation pressure, which are produced by combustion processes. This aerosol formation is obvious in urban and industrial environments. However, natural forest, brush and grass fires also provide an important atmospheric aerosol particle source (see Cadle, 1973).

The formation rate of small Aitken-size aerosol particles in the atmosphere was estimated by Lopez et al. (1974) on the basis of their aircraft measurements carried out over S. W. France. They argued that in an air column with base area of 1 cm2 3 x 104 particles are formed each second. A quarter of this quantity is due to human activity. In a more recent paper Bigg and Turvey (1978) speculate that the natural particle production rate is only 170 cm ~2 s"1, two orders of magnitude smaller than the above estimate. Bigg and Turvey establish this rate by using the results of their observations, carried out over Australia, together with an acceptable residence time of 3 x 10s s. It follows from this figure that the total Australian source strength is about 10"s-'. This may be compared with the total particle flux of 4 x 1019 s"1 produced by only one industrial area (Perth) which exceeds the global natural emission of the continent! The present author feels that the natural production rate proposed by Lopez et al. (1974) is too high, while Bigg and Turvey's value is too low, at least for the European continent. Thus, according to Selezneva (1966), who made a large number of aircraft measurements over the European part of Soviet Union, the particle number is 6 x 10Hcm 2 in a tropospheric air column. Using a residence time of 3 x 105 s, we calculate an acceptable production rate of 2 x 103 cm-2 s~ '.

4.2.3 Other formation mechanisms

Besides the major aerosol formation mechanisms discussed in the previous section other processes also produce atmospheric particles. The strength of these sources can be neglected on global scale. However, the effect of particles formed in these ways may be important under special conditions. For this reason these sources are briefly enumerated in this subsection.

First of all, volcanic activity must be mentioned; it introduces both gases (see Section 2.3 and Subsection 3.6.2) and particles into the atmosphere. The particles play an important temporary role in the control of atmospheric optical properties and radiation balance. Thus, after the eruption of Krakatoa in 1883 unusual darkness was observed over Batavia and the height of the volcanic cloud reached the altitude of nearly 30 km (18 miles). After the violent eruption of the Agung volcano in 1963 the optical effect of ash particles was identified at several points of the Earth and a temperature increase of 2 C was measured in the stratosphere (see Cadle, 1973)due to the radiation absorption of particles. While an important part of volcanic particulate matterconsists of dispersed lava, sulfuric acid also was detected in volcanic fume (Cadle, 1973).

Another special class of particles is meteoritic dust of cosmic origin. Smaller meteoritic particles (r < 1 ¿/m) can reach the lower layers of the atmosphere without significant modifications. However, larger meteorites falling through the atmosphere partly or totally evaporate due to frictional heating. In the troposphere, spherical droplets from the condensation of the resulting vapour can be identified (e.g. Wirthand Prodi, 1972). Pettersson estimates (see Cadle, 1973) that 14 x 106tof meteoritic materials are collected annually by the atmosphere of our planet.

It is believed by some workers that meteoritic particles may play an important role in the formation of precipitation, since they serve as ice forming nuclei in clouds of appropriate negative temperatures. Since ice crystal formation may initiate precipitation formation in mixed clouds (i.e. clouds containing liquid as well as solid phase), Bowen argues (see Fletcher, 1962) that the distribution of precipitation is controlled to some degree by meteor showers reaching the atmosphere. It is to be noted, however, that this theory is far from generally accepted by cloud physicists (Fletcher, 1962).

Finally, many viruses, bacteria, pollens and spores can be found in the lower atmosphere. The size of viruses and bacteria is small, while the pollens and spores are in the giant size range. According to A. Meszaros (1977), on an average 20 of the giant particles in clean continental air are composed of pollen and spores during the appropriate seasons. The biological importance of these airborne materials is obvious.

4.2.4 Comparison of the strength of different aerosol sources

The strength of various natural and anthropogenic aerosol sources is tabulated in Table 17 (SMIC, 1971). The precision of estimates is represented by the intervals given. For further details, the interested reader is referred to the original work.

The most important fact emerging from the data given is the ratio of the intensity of anthropogenic sources to the strength of natural production mechanisms. It can be seen that on a mass basis emission by natural sources exceeds anthropogenic production by a factor of 4-5. The other essential thing shown by the table is the importance of gas-to-particle conversion in the formation of aerosol particles. The fraction of particles formed by gaseous reactions is particularly significant in the case of man-made emissions.

Table 17

Formation rate of atmospheric aerosol particles with radii smaller than 20 fim (SMIC, 1971)

Natural aerosol sources

Soil and rock debris* 100-500

Forest fires and slash-burning* 3-150

Sea salt (300)

Volcanic debris 25-150 Particles formed from gases

Sulfate from H2S 130-200

Ammonium salts from NHj 80-270

Nitrate from NO, 60-430

Hydrocarbons from plants 75-200

Subtotal 773-2200

Man-made aerosol sources

Particles (direct emission) 10- 90 Particles formed from gases

Sulfate from S02 130-200

Nitrate from NO, 30- 35

Hydrocarbons 15- 90

Subtotal 185-415

Total 958-2615

Note: Values are expressed in 10" t yr '. Asterisk denotes unknown amounts of indirect man-made contributions

43 Concentration and size distribution of atmospheric aerosol particles

43.1 Concentration and vertical distribution of Aitken particles

We can characterize the particle concentration of an aerosol in two different ways. Firstly, the number concentration can be employed which is the number of the particles in a unit gas volume. Secondly, like the concentration of trace gases, the particle mass in a unit volume (mass concentration) can be given as we have seen in Chapter 3. In this subsection we will always use the number concentration of all particles which is practically equivalent to the number of Aitken size particles (see

Subsection 4.1.2). All values discussed in the following were obtained using expansion chambers.

The early results of particle concentration measurements carried out under different conditions were compiled by Landsberg (see Junge, 1963). Some ofhisdata are reproduced in Table 18. The figures tabulated make it evident that the majority of particles are of continental origin. It is also obvious that man's activity plays an important role in particle production. Furthermore, one can see from Table 18 that the number concentration of atmospheric aerosol particles decreases with increasing height.

Table 18

Average concentration of Aitken particles under various conditions according to Landsberg (see Junge, 1963)

Numbei ol Average concentration

Location observations

Cities 2500

Towns 4700

Country 3500

Sea shore 7700 Mountains

500-1000 m 870

1000- 2000 m 1000

2000 m 190

Islands 480

Oceans 600

Was this article helpful?

0 0

Post a comment