FIGURE 8.19 Concentrations of major anions and cations in coastal stratus clouds at La Jolla Peak, California, on July 6, 1993 (adapted from Collett et ai, 1994).
centrations, e.g., by 02 and 03, these variations in composition with particle size can be important. Thus oxidation of S(IV) by 02 and 03 would be expected to be inhibited in the more acidic droplets.
In the vast majority of studies, the chemical composition of fogs and clouds is measured on bulk samples, i.e., collected without discrimination by size, and the relative importance of various oxidation mechanisms assessed based on this bulk composition. As discussed by Collett and co-workers (Collett et al., 1994; Bator and Collett, 1997), the use of the average cloudwater composition without taking into account the differences in chemistry in different drop sizes leads to an underestimation of the rate of S(IV) oxidation if the pH-dependent oxidation mechanisms contribute significantly. For example, model predictions by Gurciullo and Pandis (1997) suggest that the sulfate production could be underestimated by a factor of ~30 in the early stages of cloud formation. The size of the effect depends not only on the time but also especially on the concentrations of the oxidants H202 and 03. The reason is that at higher pH values, 03 can contribute significantly to the oxidation of S(IV) and this will therefore occur preferentially in the larger drops having lower concentrations of H+. On the other hand, if the H202 concentration is large relative to 03, the oxidation by the peroxide overwhelms the ozone reaction; in this case, there is little effect of particle size since the H202 oxidation is not highly dependent on pH.
Using bulk average composition to assess the chemistry of fogs and clouds can be misleading in other respects as well. For example, Pandis and Seinfeld (1991, f992) show that the bulk mixture formed from drops with different pH values that are each in equilibrium with the gas phase does not itself conform to Henry's law, with the bulk mixture being supersaturated with respect to species such as weak acids and
NH3. Winiwarter et al. (1992) show that deviations in the opposite direction can result for highly soluble species in a closed system.
In short, the use of average bulk aqueous-phase composition to assess the importance of various oxidation pathways can lead to erroneous results if the composition of drops varies significantly with particle size.
i. Fog-Smog-Fog Cycles
Studies of the chemical composition of fogs have shown that they can be quite acidic and contain high concentrations of other cations and anions as well. For example, pH values in the range of 2-4 are not uncommon, and values as low as 1.69 have been observed (e.g., see Munger et al., 1983; Hoffmann and Jacob, 1984; Jacob et al., 1985; Fuzzi et al., 1984; Muir et al., 1986; Hering et al., f987; and Erel et al., 1993).
The reason that the aqueous-phase concentrations in fogs can be so high is related in part to the liquid water content (LWC), which is a major difference between clouds and fogs. The liquid water content for fogs is typically of the order of O.f g m-3 of air, whereas that for clouds is about an order of magnitude higher. This small LWC in fogs corresponds to increased solute concentrations.
A second reason for the higher concentrations in fogwater is the relatively rapid condensation and evaporation of water vapor onto preexisting particles as the fog forms and then dissipates. Figure 8.20 is a schematic of this process. Aerosol particles, which typically contain sulfate and nitrate as well as a variety of other species, act as condensation nuclei for the condensation of water as the relative humidity increases during fog formation. As more water condenses, the particle components become increasingly diluted. Simultaneously, the solution can take up gases such as S02, HN03, and NH3, and chemical reactions such as the oxidation of S(IV) to sulfate can occur in the droplet. Both the initial sulfate in the particle on which water condensed and sulfate formed by oxidation of S(tV) in the fog aqueous phase can contribute significantly to the total sulfate deposition. For example, Bergin et al. (1996) estimate that in a typical Arctic radiation fog, preexisting sulfate aerosol accounts for about 60% of the sulfate deposition and oxidation of S(IV) in the aqueous phase the remaining 40%.
Because these fog drops are much larger than the initial particles, gravitational settling is more important and acts to remove some of the particles (Pandis and Seinfeld, 1990). When the temperature rises, water begins to evaporate out of the fog droplet, concentrating the solutes. It is this process that likely led to the
Concentration by evaporation
Concentration by evaporation
Growth by water condensation on existing nuclei
Growth by water condensation on existing nuclei
Dissolution of gaseous components e.g., NOx, S02, NH3
Chemical conversions and ^
deposition to surface Temperature Relative humidity
FIGURE 8.20 Schematic of role of atmospheric aerosols in fog formation and evaporation (adapted from Munger et al., 1983).
extremely low observed pH value of 1.69. This mechanism is consistent with fog-smog-fog cycles that are commonly observed in urban areas (Munger et al., 1983; Pandis et al., 1990). [However, it is interesting to note that in some cities with relatively good air quality, reduced fog formation has been observed and attributed to increased temperatures via the heat island effect (Sachweh and Koepke, 1995).]
Because such fogs can be so concentrated, there has been a concern regarding their effects on both human health and ecosystems. For example, Hoffmann (f984) has estimated the sulfate ion concentrations in the London fog during the 1952 smog episode. While the pollutants that were present and their concentrations are not well known, 4000 excess deaths occurred in a 5-day period (Chapter l.B). Hoffmann and co-workers estimate that the sulfate concentration was approximately 11—46 milliequivalents per liter (mequiv L~') during this air pollution episode. This is of the same order of magnitude as those measured in acid fogs in other areas. Hence the health effects of acids and associated species in fogs continue to be of interest.
Adsorption of SOz on the surfaces of solids, followed by its oxidation on the surface, may provide a third route for the formation of sulfuric acid. The solid surfaces may be suspended in the gas phase or in atmospheric droplets, as discussed in detail by Chang and Novakov (1983).
It is known that S02 in air is oxidized on both graphite and soot particles, with water vapor enhancing the sulfate formation on the surfaces (Hulett et al., 1972; Novakov et al., 1974; Halstead et al., 1990). Whether this surface oxidation is enhanced by the presence of oxidants other than air (e.g., 03 or N02) is not clear. For example, under certain conditions some researchers report an enhancement of the rate of uptake of S02 due to the presence of the copollutants 03 and N02 (Cofer et al., 1981; Britton and Clarke, 1980); however, others have not observed such an effect (Baldwin, f982; Halstead et al., f990), particularly at low gas concentrations more closely approximating those in the atmosphere (Cofer et al., 1984).
Carbon surfaces of various types have been the subject of most studies. However, other types of surfaces, including alumina, fly ash, dust, MgO, V2Os, Fe203, and Mn02, have also been shown to oxidize S02 and/or remove it from the gas phase (Hulett et al., 1972; Judeikis et al., f978; Liberti et al., 1978; Barbaray et al., 1977, 1978; Halstead et al., 1990). As expected, the rate of removal depends on the nature of the particular surface, the presence of copollutants such as N02, and, as in the case of carbonaceous surfaces, the relative humidity. The increase with increasing water vapor suggests that oxidation of the S02 may occur in a thin film of water on the surface of the solid.
Carbonaceous particles suspended in aqueous solutions can also act as sites for efficient S02 oxidation. For example, Novakov and co-workers have studied the kinetics of such processes with respect to the concentrations of 02, S(IV), and carbon particles as well as the pH and temperature dependencies (Brodzinsky et al., 1980; Chang et al., 1981; Benner et al., 1982).
Extrapolation of their results to atmospheric conditions led them to suggest that such heterogeneous reactions could be important in the aqueous-phase oxidation of
Was this article helpful?