Click for next page ( 30


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 29
~ Atmospheric Processes Physical and chemical processes in the atmosphere deter- mine the fates of emissions of precursor gases and hence the exposures of primary receptors to pollutants. In this chapter we review current understanding of these atmospheric processes in light of the need to characterize relationships between emissions and deposition. For convenience, we consider these atmospheric processes as occurring in a sequence of clearly defined steps (Figure 2.1). The separate processes are as follows: I. Transport and mixing II. Chemical reactions in the homogeneous gas phase III. IV. V. (dry reaction) Dry deposition Attachment Chemical reactions in the homogeneous aqueous phase (wet reaction) VI. Wet deposition Heterogeneous chemical processes may occur between gases and liquids adsorbed on solid surfaces, although these are generally considered to be less important in the development of acid deposition than the homogeneous processes. We therefore do not consider heterogeneous processes in this report. Each of the separate processes takes a certain amount of time; the sum of the processing times along any particular pathway is the source-receptor transport time for the pollutant along that pathway. The processing times are extremely variable, depending strongly on meteorological processes, ambient conditions, and the presence and concentrations of various chemical species. Many of the steps are reversible, so that itinerant 29

OCR for page 29
30 UNREACTED POLLUTANT O REACTED POLLUTANT OCONDENSED WATER f _ ~WET 15E ~ _m REACTION A1TA9HMENT~ ~-_ TRANSPORT ~ / AND MIX~ ~ _ art DRY - REACTIONS m- ". ~ DRY OPPOSITION PRECIPITATION; !. \'.'\ \ FIGURE 2.1 Atmosphenc pathways leading to acid deposition. pollutant molecules may undergo repeated cycling with corresponding lengthening of the effective processing times. TRANSPORT AND MIXING The transport of pollution by normal atmospheric advec- tion and mixing is a vitally important influence on deposition phenomena, and it works both directly and indirectly. Transport phenomena directly determine where the pollution goes before it is deposited and therefore affect the atmospheric residence time of pollutant materials. Dry deposition, for example, is often limited by the speed at which the atmosphere can vertically transport pollution to the proximity of the surface. Transport can indirectly affect pollutant deposition in a number of ways. Transport processes, for example, bring pollution into contact with storm systems, where precipitation scavenging occurs. Transport also can introduce pollutants into environments more (or less) conducive to transformation chemistry. This complex of interactions links transport with the other processes shown in Figure 2.1.

OCR for page 29
31 It has been the usual practice to divide atmospheric transport processes into two categories. The first, usually termed advection, pertains to the net motion of a parcel of air as it drifts with the mean wind. The second category, diffusion, pertains to the intermixing of the parcel with its surroundings. Historically the distinction between atmospheric advection and diffusion has not been totally clear. Quite often, for example, atmospheric transport models incorporate diffusionlike terms to account for time-averaging of meandering plumes, when in fact the physical processes described have little to do with actual intermixing of materials. Similar treatments often arise in transport models using grids to approximate the desired solutions numerically. Advection processes occurring on scales smaller than the grid spacing escape resolution by the system and thus are often lumped in terms of pseudo-diffusion processes (see Appendix B). Such approximations are often inescapable. They do, however, contribute significantly to the uncertainty in our ability to model atmospheric pollution, and they obscure the meaning of diffusion in such processes. It is therefore important to remember that advection and mixing are indeed distinct transport phenomena that can lead to different behavior of parcels of polluted air. The distances associated with pollution transport obviously depend strongly on how long the pollutant resides in the atmosphere and thus is available for action by the advection-diffusion process. In this context it is important to note that atmospheric residence times for typical power plant pollutants {sulfur com- pounds, for example) are rather uniformly distributed; some pollutant molecules are deposited from the atmosphere relatively quickly and thus at locations near the source, whereas others are deposited more slowly and thus much farther away. On the basis of the best current estimates, it is not unusual for the transport distance of a given pollutant molecule to be of the order of hundreds or even a thousand kilometers. It also is not unusual, however, for a molecule to be deposited close to the source. From this one can conclude that while long-range transport processes certainly are important, shorter-range phenomena are occurring as well. Another factor that must be taken into account in assessing transport is the height at which pollutants are released into the atmosphere. One approach to local air- quality problems has been to increase the height of

OCR for page 29
32 stacks in accordance with the notion that the higher the point of release the less the pollutants would affect the surrounding area. This approach to pollution control was applied to large new plants where taller stacks were con- structed. At older plants, relatively short stacks were replaced with taller ones. A number of factors, such as meteorology and terrain, influence how the height of an individual stack affects dispersal of a given pollutant, so it is difficult to evaluate the effectiveness of tall stacks for dispersal in general. Recently Koerber (1982) studied a set of 62 coal-fired power plants in the Ohio River Valley. He developed a measure of the potential for long-range transport that involved physical stack height, plume rise, and mixing height. Figure 2.2 shows the temporal trend in Koerber's parameter between 1950 and 1980. The implication of this and other work is that stack heights must be taken into account when assessing source-receptor relationships involving long-range transport. Because of cumulative uncertainties, the trajectories and times associated with long-range transport are much more difficult to estimate than their shorter-range counterparts. Early, very crude attempts to simulate long-range phenomena simply employed local wind roses and straight trajectories from the sources in question. The obvious deficiencies associated with this approach prompted further efforts to develop curved-trajectory simulations, which were driven by conceptualized, time- evolving wind fields. The curved-trajectory approaches, while representing a major advancement over their straight-line predecessors, suffered from two major disadvantages. The first of these was that the data from which the wind fields were derived were usually extremely sparse in both time and space--a problem that becomes particularly severe under complex meteorological conditions involving fronts and storm systems. Although a variety of sophisticated interpolation techniques has been advanced subsequently to offset this problem, the poor coverage of meteoro- logical data in both space and time remains particularly troublesome. The second major problem associated with these types of trajectory approaches is caused by mass motions of air vertically and vertical wind shear, i.e., the dependence of wind speed and direction on altitude. Early trajectory simulations, based on constant-altitude wind fields, soon were replaced by layer-averaged or constant-pressure

OCR for page 29
33 o 0.3~ . _ ~ O 0.2 Q) A en ~ ~ ~ 0.1 .= a) ~ 0 Weighted by SO2 emission /_' _ 1 _ Weighted by I/ MW rating l,/ 117 _; If 1950 1955 1960 1965 1970 1975 1980 Year FIGURE 2.2 Trend in long-range transport potential for 62 sources in the Ohio River Valley. SOURCE: Koerber (1982~. surface versions to overcome this disadvantage par- tially. On the basis of thermodynamic arguments, it is expected that vertical motions of air parcels should adhere rather closely to constant-entropy surfaces in the atmosphere, and from this a few ~isentropic~ trajectory simulations have evolved as well. However, vertical motions caused by the heat released or absorbed during cloud formation are not taken into account by either method. Although curved-trajectory simulations can produce rather reliable results in simple meteorological situa- tions, they are fraught with uncertainty when conditions become complex, such as near frontal systems. Some idea of this uncertainty may be gained from Figure 2.3, which shows the results of two different calculations of a trajectory under the same conditions in the vicinity of a frontal storm. One calculation (solid curves) uses the assumption of isentropic transport, while the other (dashed curves) employs isobaric transport. After 24 hours, the calculated positions of the two hypothetical air parcels are several hundred kilometers apart (also see Chapter 3 on the treatment of transport and mixing in models as well as Appendix B).

OCR for page 29
34 1""' ,,, mu; \~\, ,' ~ ~ ~N ,} , ~ ~ ~ FIGURE 2.3 Calculated plume trajectories in the vicinity of frontal systems for 24 hours after release. Solid plumes were calculated on the assumption of isentropic vertical motion; dashed plumes were calculated on the assumption of isobaric motion. Shading indicates area of precipitation. SOURCE: Adapted from Davis and Wendell (1977~. The difficulties associated with wind shears and vertical motions could be largely overcome if the vertical motions were indeed known. Several weather prediction models produce such information, albeit prognostic in nature, rather coarse in scale, and dedicated more to upper regions of the troposphere than to the planetary boundary layer. In addition, a few mesoscale dynamic models currently exist that can supply initial estimates, at least, of information of this type. The application of such techniques for pollution trajectory simulation has been comparatively limited owing to the complexity of the modeling process and the expense of the simulations. A summary of several of the trajectory simulation techniques discussed here appears in Chapter 3 (Table 3.2). They have been applied to form composite regional pollution models. A major factor contributing to the uncertainty in long-range trajectory simulations stems directly from our current inability to measure long-range transport. Several balloon studies have been attempted, but they have been less than satisfactory because of technical difficulties, the balloons' supposed inability to track

OCR for page 29
35 vertical motions exactly, and statistical problems associated with tagging a stochastic system with too few units. Chemical tracers have not been particularly successful to date owing to detection difficulties over large distance scales. Tracer techniques are evolving rapidly, however, and it is not unreasonable to expect some highly significant results to emerge from experiments with tracers in the next 5 years. For example, during the summer of 1983 six releases of tracers are to be made from Ohio and Ontario under the Cross Appalachian Tracer Experiment (CAPTEX). A wide arc of measurement sites will be set up over 600 km downwind of the releases. This experiment will be the first step in a long-range tracer program. An important feature of the long-range transport of air pollutants is that the plumes from individual sources may become so dilute and so thoroughly mixed far downwind of major source areas that the attribution of specific parcels of polluted air to specific sources is imprac- tical. In these cases, the sources contribute pollutants to air masses that may be considered to be entrained in synoptic-scale meteorological systems. The classic example of mixing occurs in large stagnant air masses that occur most frequently in summer in the eastern United States (see Chapter 4). The motion of air masses on the synoptic scale may be important for understanding acid deposition in areas remote from major source regions. The average flows across North America are shown in Figure 2.4, which illustrates that the region in which acid deposition is currently thought to be an environ- mental problem is also a region of intense interaction between tropical marine and arctic air masses. CHEMICAL TRANSFORMATION During transport through the atmosphere, SO2, NOX, hydrocarbons, and their oxidation products participate in complex chemical reactions that transform the primary pollutants into sulfates and nitrates. The transformation processes are important because, as we discuss later, deposition of the primary pollutants and that of their transformation products are governed by different processes. There are many chemical pathways through which SO2 and NOX in the atmosphere can be transformed "oxidized) into sulfate and nitrate compounds, including homogeneous

OCR for page 29
36 Hi\ r~ , .S.trono~` _ ~ w~ . ~- WN \ \ ~ ~..,w'~' ~,\\~ ,~ / \ \. \ ~ ~ ~< v )Arctic a~rstrearri~ \ ~~ ~ r N.~..\ \N a' - I ~ ~ \ ~ westerli-~l ~ ~ ~ ~ W~ :,Wopim a Mr~/~,'.'~ 20~. ~ TOW' . _ . -con FIGURE 2.4 Surface flows across North America, illustrating the area of complex entrainment and mixing of air masses in the eastern portion of the continent. SOURCE: Bryson and Hare (1974~. processes that take place droplets or heterogeneous the surfaces of particles indicates the pathways by transformed into gaseous in the gas phase and in liquid processes that take place on or droplets. Figure 2.5 which SO2 and NOx are and aqueous-phase acids. Field studies indicate that the relative importance of gas- and liquid-phase reactions depends on meteorological conditions, such as the presence of clouds, relative humidity, intensity of solar radiation, and the presence and concentrations of other pollutants. A comprehensive review of homogeneous gas- and solution-phase atmospheric chemistry associated with acid deposition is presented in Appendix A. The appendix

OCR for page 29
37 o ~n I ~n ~ ._ ._ , .O o ' 3 CD r~ o I O Ce, I _ ~ ' ._ ~ ~_ ~ o - o ~ ' O - x Q - o o =5 o X ~ ~ ~ O I I O ase4d sea o z I .= .O . _ z 4e - cn /! , - O _ I O _ _ C a, ._ ~ X o X o Q ~ C ~o O ~n _ lD ~ _ ~o ~ _ . _ .- X O I O C: O O _ I ~ O I I" ~r Z I - Z Z O ~ ~ - Z 4-~L ~o~ - 4 - 4 - 3 ~ ~ s . ~n- ._ ~._ ~ V} C ~ E ,c E ~ ._ ._ ._ 4 - c c c 'c 0 0 0 E ,c E `~, E ~ E ~ <: ~ s~ c~ o s C ~ _ c ~ 0 x 0 ~ ~ 4 - c ,o 4 - ca o o 0. N ~ I c ~ - S 11 _ o ._ . _ ~ ~ _ ~ C t0 e ~ C ~ _ o~a ~u!e~ 'slaldolp pnolo ase~d snoanb~ I ct ._ c' ce c' _. _ c~ ~0 O ~ ~ ct c - o'~ OCR for page 29
38 includes detailed descriptions of alternative oxidation pathways and analyses of reaction rates. General descriptions and conclusions, drawn from this material, are presented below. Homogeneous Gas-Phase Reactions SO; and NOX have been observed in the atmosphere to be oxidized through homogeneous gas-phase reactions at rates of a few and 20 to 30 percent/in, respectively (Step II in Figure 2.1). The observed rates cannot be explained by direct oxidation by atmospheric oxygen, reactions that occur too slowly for typical concentrations of pollutants and, in the case of SO2, in the absence of catalysts. Similarly, although there are direct pathways to the formation of sulfuric and nitric acids beginning with absorption of solar radiation by SO2 and NC2, respec- tively, these processes also appear to be unimportant under typical conditions in the troposphere. According to current understanding, most of the As-phase chemistry in the lower atmosphere that results in oxidation of S02, nitric oxide (NO), and nitrogen dioxide (NC>) entails reactions with a variety of highly reactive intermediate s -excited molecules, atoms, and free radicals (neutral fragments of stable molecules)--that are generated in reactions initiated by the absorption of solar radiation by trace gases. The most important of the intermediates for gas phase oxidation appears to be the hydroxy radical, HO. The hydroxy radical can be formed in the troposphere by a number of reactions. ~ ~~ ~ A common process begins wltn dissociation of NCk by absorption of sunlight, which forms a highly reactive oxygen atom that combines quickly with a diatomic oxygen molecule to form the triatomic oxygen molecule, ozone (O3). Ozone may be photodis- sociated, yielding an electronically excited diatomic molecule of oxygen and an electronically excited oxygen atom, O(1D), which reacts readily with a water molecule to form HO. The hydroxy radical, unlike many radicals that are fragments of complex molecules containing carbon, does not react readily with molecular oxygen; HO survives in the atmosphere to react with most impurity gases, such as hydrocarbons, aldehydes, NO, NO2, SO2, and carbon monoxide (CO). Reactions between HO and several impurity gases produce additional classes of reactive transient species, which, in turn, react with

OCR for page 29
39 atmospheric constituents to form additional reactive species. For example, reactions of HO with CO and hydrocarbons produce peroxy radicals; peroxy radicals react rapidly with NO to form NO2 and alkoxy, acyloxy, and other HO radicals. The net result of all of these interactions is a large number of chemical pathways for oxidation of SO2 and NOx to sulfuric acid (H2SO4) and nitric acid (HNO3), respectively, many of which depend initially on the formation of HO. A sequence of these reactions can be constructed in which a single HO radical may oxidize CO, hydrocarbon, or aldehyde, followed by oxidation of NO to NO2 accompanied by production of additional HO radicals. Repeated cycling of the sequence results in continued oxidation of NO to NO2 and relatively constant concen- trations of HO. There are a number of gaseous-phase chemical reactions between SO2 and reactive transient species that may lead to formation of H2SO4; these reactions, along with currently accepted values for the reaction rates, are listed in Appendix A. While many of the rate con- stants are known only with an uncertainty of 50 percent, it appears as if the most important reaction is that between SO2 and HO, yielding HOSO2. TV 1clence i s goon that this reaction ultimately leads to the generation of sulfuric acid, and a number of pathways for this subse- quent reaction have been explored. Which of these pathways is most important is still unknown, but it is likely that the oxidation of SO2 by HO is a chain- propagating reaction. The principal agents for oxidizing NO to NO2 are ozone and peroxy radicals, whereas NO2 is oxidized to HNO3 by a well-characterized reaction with HO (Appendix A). According to current understanding, then, the rates at which sulfuric and nitric acids are formed in homogeneous gas-phase reactions depend on ambient concentrations of the hydroxy radical. Direct measurement of HO in the atmosphere is difficult, but both theoretical and experimental estimates are available from which to estimate rates of conversion from SO2 and NO2 to H2SO4 and HNO3, respectively. Using the rate constants listed in Appendix A, we find that for high concentrations of HO--characteristic of polluted summer sunny skies--SC will be converted to H2SO4 by reaction with HO at a daily averaged rate of about 0.7 percent/in (16.4 percent per 24-h period), whereas NOx

OCR for page 29
44 both nitric and sulfuric acids were formed rapidly in the cloud, although the oxidizing agent remained unidentified because of weaknesses in the analytical methods. Data from another experiment are now available showing an appreciable rate of conversion of SO2 to H2SO4 at night in clouds over coastal waters, indicating an oxidation process other than reaction with the hydroxy radical, which is present in significant concentrations only in daytime. The importance in atmospheric chemistry of aqueous- phase processes taking place in clouds is illustrated theoretically in Figure 2.7, which gives the results of calculations that combine homogeneous gas-phase chemistry with the current picture of aqueous reactions (Environ- mental Research & Technology, Inc., and MEP, Inc. 1982). The figure shows the progress of oxidation in clear air (beginning with NO and NO2 concentrations of 10 ppb, concentration of reactive hydrocarbon vapors of 200 ppb, SO2 concentration of 5 ppb, and SO4 concentration of 2 ug/m3) and the effects of introducing a cloud with 1 g/m3 of liquid water at 1400 h. In theory the inser- tion of cloud water causes dramatic decreases in atmo- spheric concentrations of H2O2, HNO3, SO2, and SO4. The behavior of NO, NO2, O3 and peroxyacetylnitrate (PAN) was not strongly influenced by the presence of cloud water. The example demonstrates that clouds have the potential to dominate chemical interactions involving water-soluble or water-scavengable constituents. Field experiments are required to determine if this dramatic effect actually occurs in the atmosphere. DEPOSITION Dry Deposition The term dry deposition is used to denote a variety of processes by which pollutant gases and aerosol particles reach the Earth's surface, including the surfaces of both living and inanimate objects on the ground (vegetation and buildings, for example). The processes depend on concentrations of the pollutants and small-scale meteoro- logical effects near the surface as well as on the characteristics of the receiving surface. Superficially, dry deposition seems to be almost trivially simple in comparison with other aspects of the relationships between emissions and deposition. Dry

OCR for page 29
45 Q - - Q 120 _ - z O 80 _ of UJ of 8 40 _ 12 _ 81-~02 NO \ \ 4 - \ \ 01 II ~ l\ , 810 12 14 16 TIME (hours) o3/ O ~1 1 1 1 8 10 12 14 16 TIME (hours) 12 _ o lo: 8 _ 4 _ of UJ of o O Q Ca) V - Q 6 _ Q O ~ _ / I HNO3/ / 1 I,~ I 111 10 12 14 TIME (hours) D 16 8 10 12 14 TIME (hours) Z=4 2 O O Z O 8 10 12 14 16 O 8 10 12 14 16 TIME (hours) TIME (hours) FIGURE 2.7 Theoretical calculations of gas and aerosol concentrations as a function of time for gas-phase reactions only (solid line) and with the introduction of cloud water (dashed line) at 1400 hours. SOURCE: Environmental Research & Technology, Inc., and MEP, Inc. (1982~. deposition takes place at the Earth's surface and thus is inactive in the volume of the atmosphere in which chemical transformation and processes leading to wet deposition occur. In fact, however, dry deposition is incompletely understood. Uncertainties in dry deposition may be an important source of error in today's regional modeling efforts.

OCR for page 29
46 There are several reasons for the current uncertainties in understanding dry deposition (Appendix C). The first is that dry-deposition rates are extremely difficult to measure. Although a number of possible techniques exist (Hicks et al. 1981) and considerable effort has been devoted to developing appropriate methods for measuring fluxes to surfaces, the base of high-quality data is still distressingly small. Furthermore, the more reliable data that do exist tend to have been obtained under experimentally convenient conditions (for example, high pollutant concentrations, uniform terrain) and thus reflect only a small subset of the potentially important environmental conditions. A second reason for uncertainty in dry-deposition rates is a consequence of the complexity of the physical processes in the atmosphere. As indicated in Figure 2.8, several mechanisms convey pollutants to the surface, and it is often not clear which processes dominate under when conditions. Especially important in this regard are the near-surface mechanisms for aerosol particles, such as inertial impact, phoresis, and electrical effects. Uncer- tainties in this area are currently substantial, especially for deposition to surfaces of vegetation. The third reason arises from uncertainties In the characteristics of the substrate on which materials are deposited. Contrary to the superficial view that dry deposition is ourelv a surface phenomenon, phenomena both at and in the substrate can Play a role in determining the deposition flux. It is well known, for example, that stomata! openings on leaf surfaces influence the deposi- tion of eases such as SON and ozone. Soils and building materials have been shown to "saturate" with depositing gases. Re-emission of sulfur compounds from plant surfaces has been detected. All of these results render the concept of a simple boundary condition approach to dry deposition somewhat questionable; the corresponding uncertainties are again large. These difficulties combine to give a number of widely varying estimates for the temporal and spatial scales of dry deposition of specific pollutants. As a rule of thumb, for sulfur and nitrogen compounds at least, dry deposition is taken on the average to be about as effec- tive as wet deposition in pollutant removal. About one third of sulfur emissions is transported out of the continent. Thus roughly one third of northeastern emissions is assumed to be dry-deposited on the North American continent (see the section in Chapter 3 on material balance).

OCR for page 29
47 4, o [ ' --- C) ._ a) o Cat o _ ._ ~ ~ o _ U) ~ C ~i. E ~ . 1- _ _' Ct5 .O o ~ ._ V, C' ~ ~ An. U.! - _ ~ a, C'-- o o \ Cow ~ - o ~ o in . _ ~ ~ 4 - Cal o ~C ~ ~ CO CO _ ._ ~ C ~ Cal ~ o ~ o ~ o me ~ 30~JUnS 3~00V lH913H . 1 I c 1 it) ix, N z ,,0 Hi: J o :^ U' 3 Cal Go

OCR for page 29
48 There are also studies, however, that obtain a scale length for dry deposition in excess of 104 km for some species (Slinn 1983), strongly suggesting interaction with global circulation patterns. This work is in concordance with observations of deposition in Greenland and the Arctic, as well as the general haze buildup in the northern hemisphere. Until the extent of such long-range transport is more thoroughly understood, the modeling of dry deposition is likely to remain highly uncertain. Wet Deposition The term wet deposition encompasses all processes by which atmospheric pollutants are transported to the Earth's surface in one of the many forms of precipita- tion: rain, snow, or fog, for example. Wet deposition therefore involves attachment of pollutants to atmo- spheric water and includes chemical reactions in the aqueous phase as well as the precipitation process itself. Aqueous phase chemical processes (step V in Figure 2.1) have been discussed previously; here we address only the physical processes by which pollutants first become attached to water droplets and then are deposited in wet form (also see Appendix C). A rough indication of the significance of wet deposi- tion on a continental scale can be obtained from a map of annual precipitation in the United States (Figure 2.9). From the distribution, one would expect that wet depo- sition would be an important contribution to total deposition in the East and in the Pacific Northwest. In regions with frequent precipitation, wet deposition also becomes relatively more important than dry deposition far away from sources, where SO2 is depleted and sulfate particles are a significant fraction of the atmospheri sulfur burden. This also appears to be the case in remote areas of southeastern Canada. Attachment Processes CThe physical processes by which pollutants become attached to droplets and other falling hydrometeors such as ice crystals (step IV in Figure 2.1) have been the subject of extensive research, and a number of technical

OCR for page 29

OCR for page 29
50 reviews of current knowledge in this area are available (see, for example, Slinn 1983). The most important attachment process under most inrcloud conditions is undoubtedly nucleation. Nuclea- tion is a kinetic process in which water molecules condense from the vapor phase onto a suitable surface. Dust and pollutant aerosol particles provide such surfaces in the air. The result is a cloud of droplets (or ice crystals) containing the pollutant. The droplets may grow by the same process (condensation) or may lose water by evaporation. The tendency of water vapor to condense on aerosol particles depends on the characteristics of the particles and the degree of saturation of the air with water vapor. As a consequence the aerosol and associated cloud par- ticles compete for water molecules. Some particles will capture water with high efficiency and grow substantially in size. Others will acquire only small amounts of water, whereas still others will remain essentially "dry" elements. In addition, some particles may be effective for nucleation of ice crystals, whereas others will be active only for the formation of liquid water. The nucleating capability of a particular aerosol particle is determined by its size, its morphological characteristics, and its chemical composition. Acid-forming particles, by their very nature, are chemically competitive for water vapor and thus tend to participate actively as condensa- tion nuclei for liquid water. This attribute enhances their propensity to become scavenged early in storms and has a significant effect on the nature of the acid- precipitation formation process. Diffusional attachment, as its name implies, results from diffusion of the pollutant molecule or particle through the air to the surface of a water droplet. The process may be effective in the case of both suspended cloud elements and falling hydrometeors. It depends chiefly on the magnitude of the molecular (or Brownian) diffusivity of the pollutant; because diffusivity is inversely related to particle size, this mechanism is less important for larger particles. For practical purposes, diffusional attachment can be ignored for particles with radii of more than a few tenths of a micrometer. The motion of a molecule or particle to the surface of a water droplet by diffusion depends on the gradient in the concentration of the pollutant in the vicinity of the surface. Thus, if the cloud or precipitation droplet can

OCR for page 29
51 accommodate the influx of pollutant readily (for example, the pollutant is highly soluble in water), it will effectively depopulate the adjacent air, thus making a steep concentration gradient and encouraging further diffusion to the droplet. particle "bounce off" or low gas solubility) the droplet cannot accommodate the pollutant, further diffusion to the droplet will be discouraged. If the cloud or precipitation droplet supplies the pollutant to the local air through an outgassing mechanism, the concentration gradient will be reversed and diffusion will carry the pollutant away from the droplet. In general, diffusional attachment processes are sufficiently well understood to allow their mathematical description with reasonable accuracy. Inertial attachment arises by virtue of the facts that pollution particles and scavenging droplets are constantly in motion and that both have finite volume and mass. The most important example of inertial attachment is the impaction of aerosols by falling hydrometeors. In this case, the hydrometeor falls under the influence of gravity, sweeping out a volume in space. Collisions occur between the falling hydrometeors and some aerosol particles, resulting in attachment. The effectiveness of impaction depends on the size of both the aerosol particle and the hydrometeor; mathe- matical formulas exist to estimate the magnitudes of these processes. Impaction generally becomes unimportant for aerosols less than a few micrometers in size. In this context it is interesting to note that a two-stage capture mechanism can exist, in which a small aerosol first grows through nucleation to form a larger droplet that is then captured by inertial attachment. This two-stage process, called accretion, is an essential factor in the generation of precipitation in clouds and has been postulated as an important mechanism in scavenging pollutants below clouds. A second example of inertial attachment is turbulent collision. In this case, the particles and scavenging elements, subjected to a turbulent field, collide because of dissimilar dynamic responses to velocity fluctuations ~ "~ 1 ~__1 _: _ If for some reason (such as - ~ ..~ _~ ~_ ~. This scavenging mechanism is thought to be of secondary importance and has received compara- tively little attention in the literature, although some recent theoretical analyses have suggested it to be significant for droplets and particles of specific sizes.

OCR for page 29
52 10 10-1 in - LL us 10 2 LL a: 10-3 _ 10-4 10-3 \ ~G reenf ield gap \ I it Diffusional \ - _ Attachment `` _ ~~' Inertial Attachment - 1o-2 lo-l 1.0 10 RADIUS OF AEROSOL PARTICLE (,um) FIGURE 2.10 Theoretical scavenging efficiency of a falling raindrop of diameter 0.31 mm as a function of aerosol particle size. SOURCE: Adapted from Pruppacher and Klett (1978~. Although the diffusional and inertial attachment processes are efficient for capturing very fine and very coarse particles, respectively, neither mechanism is effective for particles in the range of 0.1 to 5 Am. The resulting minimum in capture efficiency as a function of particle size, shown schematically in Figure 2.10, is known as the Greenfield gap. Depending on circumstances, there are several additional attachment mechanisms (including accretion via the two-stage nucleation-impaction mechanism mentioned earlier) that can operate in the Greenfield gap. The processes include turbulent deposition, electrical attraction, and phoretic effects (see Appendix C for details). As indicated by the dashed lines in Figure 2.10, these mechanisms can significantly relieve the Greenfield effect under appropriate circumstances (Appendix C).

OCR for page 29
53 From this discussion, it should be evident that the aggregate of possible attachment processes comprises a complex system that is difficult to characterize mathe- matically. This complexity, combined with the processes of formation and delivery of precipitation that occur both consecutively and concurrently, provides a major source of uncertainty in current models of regional pollution transport and deposition. REFERENCES Bryson, R.A., and F.K. Hare. 1974. Climates of North America. World Survey of Climatology, Vol. 11. New York: Elsevier Scientific Publishing Company. Davis, W.E., and L.L. Wendell. 1977. Some Effects of Isentropic Vertical Motion Simulation in a Regional Scale Quasi-Lagrangian Air Quality Model. BNWL-2100 PT 3. Richland, Wash.: Battelle Pacific Northwest Laboratories. Environmental Research & Technology, Inc., and MEP, Inc. 1982. Models for Long Range and Mesoscale Transport and Deposition of Atmospheric Pollutants. Phase I: Modeling System Design. Report SYMAP-101. Toronto, Ontario: Ontario Ministry of Environment. GCA Corporation. 1981. Acid Rain Information Book. Final Report. DOE/EP-0018. Prepared for the U.S. Department of Energy. Springfield, Va.: National Technical Information Service. Hicks, B.B., M.L. Wesely, and J.L. Durham. 1981. Critique of methods to measure dry deposition; concise summary of workshop. Presented at the 1981 National Meeting of the American Chemical Society, Atlanta. Ann Arbor: Ann Arbor Scientific Publications. Koerber, W.M. 1982. Trends in SO2 emissions and associated release height for Ohio River Valley Power Plants. In Proceedings of the 75th Annual Meeting of the Air Pollution Control Association, paper 82-10.5, New Orleans. Pittsburgh: Air Pollution Control Association. Martin, L.R. 1983. Kinetic studies of sulfite oxidation in aqueous solutions. In Acid Precipitation: NO, and NO2 Oxidation Mechanisms: Considerations. Ann Arbor, Mich.: Scientific Publications. In press. Atmospheric Ann Arbor SO2,

OCR for page 29
54 Pruppacher, H.R., and J.D. Klett. 1978e Microphysics of Clouds and Precipitation. Boston: D. Reidel Publishing Company. Sling, WeGeNe 1983e Precipitation scavenging. In Atmospheric Sciences and Power Production. D. Randerson, ad. Washington, D.C.: UeSe Department of EnergY,-In press.