National Academies Press: OpenBook

Global Tropospheric Chemistry: A Plan for Action (1984)

Chapter: 5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES

« Previous: PART II ASSESSMENTS OF CURRENT UNDERSTANDING
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 55
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 56
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 57
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 58
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 59
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 60
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 61
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 62
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 63
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 64
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 65
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 66
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 67
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 68
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 69
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 70
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 71
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 72
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 73
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 74
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 75
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 76
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 77
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 78
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 79
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 80
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 81
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 82
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 83
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 84
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 85
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 86
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 87
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 88
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 89
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 90
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 91
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 92
Suggested Citation:"5 CRITICAL PROCESSES AFFECTING THE DISTRIBUTION OF CHEMICAL SPECIES." National Research Council. 1984. Global Tropospheric Chemistry: A Plan for Action. Washington, DC: The National Academies Press. doi: 10.17226/177.
×
Page 93

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Critical Processes Affecting the Distribution of Chemical Species BIOLOGICAL AND SURFACE SOURCES BY. R. CICERONE, C. C. DEEWICHE, R. HARRISS, AND R. DICKINSON THE IMPORTANCE OF BIOLOGICAL SOURCES In recent years we have seen increasingly clear mani- festations of the important, even dominant, roles of bio- logical processes as sources of atmospheric chemicals. While it has long been recognized that the sheer variety and adaptivity of the biosphere permit a wide range of phenomena, the quantitative size of biological influ- ences on the atmosphere (examples below) has amazed many atmospheric chemists and climate experts. More biologically oriented scientists have long recognized the great potential of atmosphere-biosphere exchange proc- esses; it has even been proposed that much of our chemi- cal and physical environment is under biological con- trol. For example, Lovelock has postulated the existence of an encompassing living feedback system through which the biosphere regulates the physical environment by and for itself as external stimuli change-the Gala hypothesis. There are both general and specific indications for the importance of biology in affecting or controlling the chemical composition ofthe atmosphere. Rough indica- tions may be obtained from examining the state of dis- equilibrium of the atmosphere's chemical concen- trations of gases that would exist if the earth's oceans and atmosphere were in perfect thermodynamic equilib- rium (TE) or with only external inputs of solar, galactic, and electrical energy. With calculations such as these, Lovelock and Margulis found concentrations of N2, 02, CH4, N2O, NH3, and CH3I many orders of magnitude lower than in the actual atmosphere. Large departures from pure TE or an analogous abiological system in the atmosphere's composition are one indication of the in- fluences of the biosphere. From a more empirical con- sideration, a comparison with Venus and Mars, planets with no life, suggests that there would be several orders of magnitude less O2 on earth without life. The importance of biological sources is indicated strongly by other empirical data and analyses. For ex- ample, the ~4C content of atmospheric CH4 was mea- sured to be about 80 percent of that in recent wood. These data, mostly from W. F. Libby, were gathered in 1949-1960, before nuclear explosive devices were tested widely in the atmosphere, and they show that at least 80 percent of atmospheric CH4 then was derived from re- cent organic material and not from old carbon, e.g., primordial CH4 or fossil fuels. Although the burning of biomass could also yield CH4 high in t4C, it is likely that ruminating animals and termites, rice paddies, and shallow inland waters are the principal contemporary CH4 sources. Also, from a geological and geochemical analysis the mostly biological origin of atmospheric O2 and the important biological role in controlling CO2 levels can be deduced. Other strong evidence exists for the importance of biological processes in the cycles of many elements 55

56 whose volatile forms pass through and affect the atmo- sphere. In the nitrogen cycle, biological and industrial fixation of N2 from the atmosphere leads to subsequent return of NH3, N2O, N2, and possibly NO to the atmo- sphere. The importance of biogenic sulfur compounds to the atmosphere and to the global sulfur cycle was recognized long ago. Principal compounds of interest are dimethyl sulfide, dimethyl disulfide, H2S, COS, and possibly others. Methylated materials of several other elements also appear to constitute globally impor- tant biological sources. These include CH3Cl, CH3I, CH3Br, and several methyl-metal compounds, e.g., those of mercury, arsenic, and probably others. While biomass burning is a source of CH3Cl, the oceans ap- pear to be a more important source. The physical Cl--I- exchange reaction in seawater postulated in the early 1970s may not be compatible with recent measure- ments indicating the simultaneous existence of CH3I and CH3C1 in seawater. In any case, CH3C1 is the only significant natural source of chlorine atoms to the strato- sphere, and its origin, probably biogenic, requires study. Over land, much of the water transferred from the surface passes through vegetation. Regardless of whether the biosphere acts as an inte- grated, almost purposeful, system in its release and reg- ulation of atmospheric gases, or whether individual spe- cies and blames act independently with inadvertent results, the biologically released materials are important in atmospheric chemistry and climate. Available data show that biogenic sulfur compounds can make appre- ciable contributions to sulfate deposition measured in certain regions, and probably to dry SO: deposition; it is important to define these contributions and regions more-clearly. Similarly, field survey data and studies on individual plants show that' the natural emissions of veg- etation represent potentially significant hydrocarbon sources for the atmosphere. Often these are isoprene and terpenes (comprised largely of isoprene-like units). These biogenic hydrocarbons react photochemically and along with NOX gases can lead to photochemical production of O3. There is also evidence that the direct emission of natural hydrocarbons and the burning of biomass can generate significant amounts of organic acids in rainfall; tropical forest areas are especially inter- esting in this regard. Processes and raw materials that produce and/or con- trol O3 concentrations in the background troposphere are not only central to tropospheric chemistry but are also of importance to climate. In the relatively clean, unpolluted troposphere, O3 iS a major source of reactive free radicals as well as a key reactant itself. Its climatic role arises from its 9. 6-pm band absorption (in the atmo- spheric wavelength window); in the troposphere, this PART II ASSESSMENTS OF CURRENT UNDERSTANDING band is considerably pressure-broadened. In the photo- chemical control oftropospheric 03, biogenic gases such as CH4, CO, many hydrocarbons, N2O, and possibly NH3 are important players, N2O through its strato- spheric production of odd-nitrogen oxides that flow downward into the upper troposphere, and NH3 as a possible NOX source and as an NOX sink. From a climatic point of view, the potential warming effects of several biogenic and anthropogenic trace gases are startling. Upward trends in the concentrations of tropospheric O3, CH4, N2O, CH3CCl3, CCl2F2, and CC13F are of particular concern, although other species also need attention. The effects of these gases will add to those of CO2. For O3, the extent of influence of biogenic input, summarized above, is a key question. For CH4 and N2O, biological sources are known to dominate the global budgets. For CH4 and CH3CCl3, whose primary sinks are tropospheric OH reactions, any understand- ing and ability to predict future trends will require knowledge of the natural and human-controlled sources and of global patterns of OH and their controlling proc- esses. For the fluorocarbons CC12F2 and CCl3F, it will be important to understand how tropospheric chemistry will respond to the stratospheric changes they cause, e.g., more ultraviolet radiation reaching the tropopause and higher O3 concentrations in the lower stratosphere and upper troposphere. Finally, there is another large role for biological proc- esses in atmospheric chemistry, that of surface receptors for depositing gases and particles. There is evidence that vegetated continental areas are major sinks for O3, HNO3, SO2, and possibly NH3, for example. Further, the effectiveness of these surface sinks is largely con- trolled by dynamic responses of the involved plants. In certain regions and seasons, the deposition of atmo- spheric gases and particles delivers important nutrients and, at other places and times, pollutants and toxins for plant life and soils. There has been an increasing awareness by the scien- tific community of the critical role played by biosphere- atmosphere interactions in biogeochemical cycles in gen- eral and in tropospheric chemical cycles in particular. A number of recent documents discuss these interactions and their implications in considerable detail. Examples include the National Research Council reportAtmosphere- Biosph~re Interactiorl. Toward a Better Urlderstarlding of the Eco- logical Cor~sequences of Fossil Fuel Combustion (~1981~; the NASA Technical Memorandum 85629, Global Biology Research Progress Program Plan (1983~; and several publi- cations by the Scientific Committee on Problems of the Environment (SCOPE), including Some Perspectives of the May'or Biogeochern~'cal Cycles (1981), and The May'or Biogeo- chernical Cycles and TheirInteractions (1983~.

CRITICAL PROCESSES THE NATURE OF BIOLOGICAL SOURCES Those atmospheric constituents to which biological sources are major contributors (nitrogen, oxygen, CO2, N2O, compounds of sulfur, halogens, and others) are products ofthe energy reactions of one or another organ- ism. The process of photosynthesis and its reversal (res- piration) are the most familiar expression ofthis fact, but the close similarity of processes yielding the other con- stituents is not always appreciated. An examination of the interrelationships between these biological oxida- tions and reductions from the viewpoint of their driving forces and modulating influences is in order. First, although no one has yet offered a completely convincing explanation of the driving force behind the phenomenon called "life, " there is much evidence of its potency. Virtually any chemical reaction that can take place in an aqueous system, that yields energy in excess of 40 kilojoules (kl) per mole for two-electron transfer, and for which the required reactants have been available in reasonable abundance over the evolutionary period of the earth has been exploited. Aside from photosynthe- sis, no physical source of energy has been so utilized. The importance of this concept from the standpoint of the atmospheric chemist is that it implies that the present composition of the atmosphere is closely coupled to the biological system, and that any alteration by physical phenomena or human activity may be countered by a biological response giving the appearance of "resist- ance" to the change. As a single example from many possible, the annual fixation of CO2 in photosynthesis (and the equivalent release of CO2 in oxidative reac- tions) is about 10 percent of the atmospheric carbon pool. This speaks for an exceedingly tightly coupled cyclic exchange. Any increase in atmospheric CO2 levels would be expected to result in a countering increase in photosyn- thesis with the fixation of more CO2. This is known to happen, but because there are so many other factors influencing photosynthetic production, the relationship . , . Is not linear. Of the elements from biological sources appearing in the atmosphere, the ones most actively cycled are those having several oxidation states within the range of stabil- ity of water, and for which at least one reasonably stable gaseous form exists. Under the reducing conditions of anoxic environments (reducing because some biologi- cally oxidizable compound is present and the influx of atmospheric oxygen is limited by a diffusion barrier or other means), compounds of these elements serve as "electron acceptors" for the biological oxidation of other, more reduced, compounds. In the process, gas 57 ecus compounds are released, some of them reaching the atmosphere. Typical reactions are as follows: 1. Denitrification: a. [HCHO] + 0.8 NO3 + 0.8 H+-CO2 + 1.4H2O + 0.4N2 or b. [HCHO] + NO3 + H+-CO2 + 1.5 H2O + 0.5 N2O 2. Sulfate reduction: tHCHO] + 0.5 SO4 + H+-CO2 + H2O + 0.5 H2S 3. Hydrogen production: tHCHO] + H2O ~ CO2 + 2 H2 4. Methane production: tHCHO]-0.5 CO2 + 0.5 CH4 In all the above, tHCHO] represents carbohydrates, although the organic substrate can be any of a variety of materials. The reactions are simplified in their represen- tation, with none of the intermediates shown. The point is that each reaction yields one or more volatile com- pounds to the atmosphere. A cursory treatment of the subject, as given here, is sufficient to demonstrate the variety of reactions possi- ble and to suggest some of the implications discussed below. The operation of the various biomes contributing to the whole of this biological process is dynamic, influ- enced by all ofthe parameters that drive it. Most ofthese ecosystems (indeed, all from an absolute standpoint) are limited by one or more of their constituents. Available organic substrate, as we have seen, is a major limitation, but in most systems, one or more mineral elements required for life are also limiting. Compounds of nitro- gen and sulfur are among the more notorious products of modern industry, and, although one tends to classify these emissions as "pollutants," they probably also are "fertilizers" for some species and in some locations. Unquestionably, human activities have altered the bio- sphere, but it is difficult to evaluate that alteration as quantitatively as desired. Concentration of atmospheric constituents on an ar- eal basis is part of the question. Immediately downwind of a point source, the concentration of an element or compound can be lethal for some organisms. On the basis of the considerations offered above, one can as- sume that any alteration of the concentration of a partic- ular element can only result in an alteration of the associ

58 ated biological population. The significance of this alteration is more difficult to interpret. The excess of CO2 injected into the atmosphere over the rate at which it can be sequestered by various carbon sinks is a good example of the complexities of this sort of question. Concern over the possible effects that in- creased atmospheric CO2 levels may have on climate is tempered by the uncertainty regarding the conse- quences ofthese effects. It is not known what would have been "normal" secular trends in the absence of this excess CO2. Intuitively, it is often assumed that any change is undesirable, but firmer information is re- quired for planning purposes. An interesting feature of the atmospheric carbon cy- cle is the role of CH4. Available data indicate that there has been a significant increase in the concentration of CH4 in the atmosphere since the industrial revolution, and i4C data imply that 80 percent or more of atmo- spheric CH4 was recently in living matter. The concen- tration increase of CO2 is better documented. The fact that the increase in atmospheric concentration of CO2 is less than would be expected on the basis of known proc- esses for the removal of carbon from the atmosphere has led to some debate. The carbon of atmospheric CH4 contains less ~4C than that of atmospheric CO2. Biologi- cal sources are largely "modern," and as noted above, they are large. The quantity of fossil carbon from natu- rat venting and fossil fuel burning does not appear to be sufficient to explain the deficit of i4C in atmospheric CH4. There is a discrimination against i4C in the CH4 formation reaction, but this can explain only part of the deficiency in TIC. Recent suggestions that large quantities of CH4 are coming from magmatic sources could explain some of this "fossil" CH4 but not all. Thus there are debates within the debates, all of them emphasizing the need for . ,% . more mtormatlon. The biological production of CH4is a marginal thing from an energetic standpoint. The rather elegant expo- sure of details of the process by H. A. Barker in 1941 has proved to be even more complex. What was thought to be the conversion of ethanol to acetate, with a concomi- tant reduction of bicarbonate ion to CH4, has turned out to be a coupling of reactions by two interdependent organisms, one oxidizing ethanol to acetate with the production of hydrogen, the other forming CH4 from hydrogen and the bicarbonate. Although there are two separate organisms involved, the removal of hydrogen by the methanogen utilizing hydrogen is required to provide the energy gradient for life support ofthe hydro- gen producer. Because of the close constraints placed on energy yield (and therefore growth) by the concentration of hydrogen gas in reducing systems, anything that reduces hydro PART II ASSESSMENTS OF CURRENT UNDERSTANDING gen concentration will accelerate oxidation of available organic substrate. On the other hand, hydrogen con- sumption (and CH4 production) depends on CO2 con- centration. Thus an increase in atmospheric CO2 levels will correspondingly affect diffusion rates from CO2 sources (anoxic environments) and could stimulate CH4 production. This, in turn, could accelerate the oxidation of available carbon compounds (some ofthem fossil) and result in an increased production of CH4, some of it deficient in i4C relative to atmospheric CO2. These en- ergy relationships are shown below. Fermentation of ethanol to acetate and hydrogen: CH3CH2OH + H2O-CH3COO- + H+ + 2 H2. At pH 7 and with other reactant concentrations stan- dard, this reaction yields only 5.3 kJ of energy, insuffi- cient to support life. With the partial pressure of hydro- gen at 1.0 x 10-3 atm, the energy yield is about 39.4 kJ, adequate forlife support if properly coupled to synthetic reactions. Oxidation of hydrogen with CO2 as an electron ac- ceptor: 4H2 + CO2 2H2O + CH4. The standard free energy for this reaction as written is about-140 kJ. By talking the 1 x 10-3 atm concentra- tion of hydrogen suggested by the former reaction, a partial pressure for CO2 of 3.2 x 10-4 aim, and a partial pressure for CH4 of 1.6 x 10-6 atm, the energy yield becomes 85.0 kJ. As written above, there are four molecules of hydro- gen involved. The exact pathway of the reaction is not known, so it is not possible to identify the point at which energy is extracted. For each mole of water produced (a two-electron process), there is a yield of 42.5 kJ of en- ergy, and it is difficult to visualize any other energy- coupling reactions. Because there is no room in the en- ergy figures for increase in the CH4CO2 ratio, the generalization probably is permitted that atmospheric CO2 concentrations should influence biological CH4 production. The close parallel in their rates of increase in the atmosphere may well be related to this interdepen- dence of biological processes. The increase in CH4 in the atmosphere would then be explained at least in part by the mobilization of organic matter in anoxic zones, some of which is "fossil" on the comparatively short time scale (thousands of years) of carbon half-life. The data to test the significance of processes such as this are lacking, and the extent to which a concentration feedback such as this can explain present inconsistencies in the data is not known. The example does serve to demonstrate the complex interaction of biological and other factors in establishing atmospheric composition, and the challenge to unravel the processes at play.

CRITICAL PROCESSES GLOBALLY IMPORTANT BIOMES Given the discussions above on the apparent impor- tance of biological sources and on the nature of the bio- logical processes that release materials to the atmo- sphere, it is now necessary to review some of the characteristics of world biomes and to formulate criteria for evaluating their potential importance. Before exam- ining data on various distinct biomes, we present the following criteria that permit identification of biomes that require research relative to their potential role as significant sources of atmospheric chemical species: 1. Biomes coveringlarge geographic areas; 2. Biomes with high gross primary productivity rates; 3. Biomes with fast cycling rates for nutrients; 4. Biomes with anoxic sites; 5. Biomes with fast rates of change of the local popu- ~at~on compared to the time scale for natural succession (30 to 70 years); 6. Biomes where processes can trigger irreversible changes (e.g., desertification or climatic change); 7. Biomes of special importance to human life (e. g., agricultural areas); 8. Biomes with unique characteristics (due, for ex- ample, to toxicological, meteorological, or successional considerations); 9. Biomes or processes that are poorly understood and that satisfy some of the criteria above. We will refer to these criteria frequently in the discussion that follows. Schemes to classify world biomes vary somewhat de- pending on the purpose of the classification, need for detail, and other reasons. Many of the available data and compilations have grown from research on the global carbon cycle and from needs of individual re- searchers to extrapolate data from isolated, in situ mea- surements into regional or global estimates. A further complication results from the distinction between gross and net primary productivity of the biomes. The former is the rate of photosynthetic carbon fixation; the latter is this rate minus the rate of respiration, i.e., the rate of carbon storage. Table 5.1 presents one compilation of data on terrestrial biomes, their sizes, net primary pro- ductivities, and phytomasses. This particular compila- tion accounts for nearly all terrestrial surfaces, or about 30 percent ofthe total global surface. For our purposes at present, the most important entries in Table 5.1 are the geographical areas covered by the individual biomes. To identify biomes of special interest in atmospheric chemistry, Table 5.2 lists about twenty specific, although informally classified, biomes and one process, biomass burning, as the rows of a matrix. The columns of this ~. 59 matrix are biogenic gases, individual species such as CH4, and groups of species such as methylated metals (CH3M) and organohalides (RX). A measure of the scientific interest in the emissions of the listed biogenic gases from the listed biomes is assigned, considering the criteria outlined above and the available data. An "X" indicates reason to expect significant emissions, and a circled "X" indicates strong reason (or directly applica- ble, available data) to expect a particularly significant biome-emission relationship. In the remainder of this section, we focus attention on several biomes and proc- esses that are potentially significant as sources for tro- pospheric chemical species. T'~n~lra and Other Northern Environments The "tundra" biome and the boreal forest at a lower latitude cover about 14 percent of the land area of the globe, most of it in the northern hemisphere. Total pho- tosynthetic productivity of this area, although less than many environments on an area basis, still is large (an estimated 10 percent of all land area, based on the fig- ures of Whitaker). Underlain by permafrost, much of it poorly drained, this area could be a large contributor to the reduced compounds delivered to the atmosphere by the biosphere (CH4, reduced sulfur compounds, and the products of denitrification). Because of its secondary economic interest and its inaccessibility, this area has not been studied intensively, and its significance in the budget of atmospheric constit- uents is poorly known. A research program to obtain needed information on fluxes from tundra regions should be flexible, starting with exploratory studies. The results of these prelimi- nary investigations will then guide further program de- velopment. Because of the two-phased nature of the tundra research, the exploratory studies should be initi- ated as early as possible on a modest scale, with the extent and nature of future studies left flexible. Aside from the physical (logistic) problems involved in investi- gations of this environment, there are geopolitical con- straints. Initial studies can be performed in North America and in the Scandinavian countries, but be- cause of the large area involved on the Eurasian conti- nent, cooperative participation by Soviet scientists should be sought. Because these frequently water-logged environments are expected to yield significant quantities of reduced carbon, nitrogen, sulfur, and in some cases, halides, the species of interest will be the products of denitrification and sulfate reduction: CH4, NH3, halides (in the vicin- ity of the ocean), various hydrocarbons, and other re- duced carbon species in forested areas. Sulfur may be a

60 limiting nutrient in many of these areas, so the sulfur components may be low or absent except within the range of ocean spray delivery. Analysis of precipitation may be desirable in later phases, but because of costs and logistic problems, precipitation sampling should not be attempted during the first 2 years except where facilities of cooperating institutions provide the opportunity. PART II ASSESSMENTS OF CURRENT UNDERSTANDING Initial investigations at three sites are proposed. Site selection is based upon a compromise of factors includ- ing the likelihood that these sites will yield information representative of significantly large areas, factors of cost and logistics, and the probable availability of cooperat- ing individuals and institutions. The suggested sites in- clude: TABLE 5. 1 Surface Areas, Net Primary Productivity, and Phytomass of Terrestrial Ecosystems of the Biospherea Ecosystem Type 1. Forests Tropical humid Tropical seasonal Mangrove Temperate evergreen/coniferous Temperate deciduous/mixed Boreal coniferous (closed) Boreal coniferous (open) Forest plantations 2. Temperate woodlands (various) 3. Chaparral, maquis, brushland 4. Savanna Low tree/shrub savanna Grass-dominated savanna Dry savanna thorn forest Dry thorn shrubs 5. Temperate grasslands Temperate moist grassland Temperate dry grassland 6. Tundra arctic/alpine Polar desert High arctic/alpine Low arctic/alpine 7. Desert and semidesert shrub Scrub dominated Irreversible degraded 8. Extreme deserts Sandy hot and dry Sandy cold and dry 9. Perpetual ice 10. Lakes and streams 11. Swamps and marshes . temperate Tropical 12. Bogs, unexploited peatlands 13. Cultivated land Temperate annuals Temperate perennials Tropical annuals , ~ . . . roplca. . perennla. .s 14. Human area Total aAnnual average values. hOfwhich 40 percent (or 0.8 x 1 on m2) productive. SOURCE: Adapted from Ajtay et al. (1979). Surface Area (1012 m2) 31.3 10 4.5 0.3 3 3 6.5 2.5 1.5 2 2.5 22.5 6 6 3.5 7 12.5 5 7.5 9.5 1.5 3.6 4.4 21 9 12 9 8 1 15.5 2 2 0.5 1.5 1.5 16 6 0.5 9 0.5 2b 149.3 NPP DM (g2/yr) 2300 1600 1000 1500 1300 850 650 1750 1500 800 2100 2300 1300 1200 1200 500 25 150 350 200 100 10 50 o 400 2500 4000 1000 1200 1500 700 1600 500 895 . Total Production DM (lOls g) 48.68 23 7.2 0.3 4.5 3.9 5.53 1.63 2.62 3 2 . 39.35 12.6 13.8 4.55 8.4 9.75 6 3.75 2.12 0.04 0.54 1.54 3 1.8 1.2 0.13 0.08 0.05 o 0.8 7.25 1.25 6 1.5 15.05 7.2 0.75 6.3 0.8 0.4 133.0 L. . 1vlng Phytomass DM (10 g/m ) 42 25 30 30 28 25 20 18 7 7.5 2.2 15 5 2.1 1.3 0.15 0.75 2.3 1.1 0.55 0.06 0.3 o 0.02 7.5 15 5 0.1 5 0.06 6 4 3.75

CRITICAL PROCESSES TABLE 5.2 Twenty-two Biomes, Sites, and Processes and Twelve Gaseous Species or Groups 61 NOx RON NH3 N2O CH4 CO RS H2S COS RX CH3M NMHC Sterile ocean Productive ocean Tropic wet Tropic dry Desert 1 Desert 2 Desert 3 productive Wet subarctic Dry subarctic Tundra Tropic agriculture Temperate agriculture Rice agriculture Temperate evergreen Temperate mixed Temperate grassland Wetlands Inland waters Sewage sources Feedlots Coastal shelf Biomass burning X X X X X X ' X X X X X X ~X X X X X X X X X X X X X X X ~X X X X X X X X X 1 X X X X X X X X X X X ~ ~X X X X X X X X X X X X X X X X X X X X X X X X X X ~X X X X X X X X X X X X X X X X X X X X X ~ X X NOTES: An "X" indicates that there is some reason to expect a significant source; a circled X indicates especially strong interest or evidence. The symbol "R" represents an organic group, RX means a methyl halide, and CH3M means a methylated metal. " 1 " includes termites and ruminants. 1. Alaskan arctic seaboard, vicinity of Point Barrow or Prudhoe Bay; 2. Interior Alaska, vicinity of Fairbanks; 3. Hudson Bay area, vicinity of Churchill. Species to be measured in the initial studies should include CH4, N2O, H2S, dimethyl sulfide [(CHINS], CO, and volatile halides. In later studies, such species as NH3 and volatile metals (e.g., mercury, arsenic, and selenium) should be measured. Ideally, gradient measurements should be made for flux determinations. During this initial phase, portable equipment with sufficient sensitivity will probably not be available, so bulk samples should be collected for analysis at cooperating laboratories. Samples of opportunity should be collected, prefera- bly during the spring ice breakup and during midsum- mer. Where possible, samples should be collected on a regular schedule throughout the year. Temperate Forests Observations on temperate forests will be of greatest value if accomplished at sites where ongoing research and monitoring programs will provide supportive infor- mation. A number of these are available within the con- tiguous United States. They provide representative sites for the forest types of interest. Forest Type coniferous Sierran mixed conifer Southern ., coniferous Mixed hardwood Possible Sites Pacific Northwest Oregon State collaboration with OSU School of Forestry Sequoia National Park (California) collaboration with UCSB and UCB scientists Tennessee collaboration with ORNL scientists Hubbard Brook (New Hampshire) Analytical Protocol This portion of the study could be done at different levels of intensity, but for the information needed, a rather elaborate and detailed (including micrometeoro- logical information) approach is most desirable. This approach would provide boundary layer gradient infor- mation on volatiles of interest, which, in turn, would make possible the estimation of emission and absorption rates. Instruments of the resolution and sensitivity needed for such a study exist, but they have not been applied specifically to any study ofthis sort. Initial appli- cation will emphasize the development of appropriate procedures and the calibration of instruments and pro- cedures, possibly at only one or two of the possible sites. As procedures are refined and important data are ob- tained, the direction and intensity of program develop- ment will become known.

62 Both laboratory and field studies of processes are de- sirable, and they involve both the canopy trapping and the gradient measurements cited above. Procedures for canopy trapping and bench analysis are at a reasonable state of development. Molecular Species of Interest NH3, CO, and NMHCs (including terpenes and other hydrocarbons) are of most interest. N2O, NOx, CH4, and reduced sulfur compounds probably are not present in significant quantities, but should be ex- amined. Topical Areas (Forests and Savannas) Tropical continental areas, both wet forested regions and drier sites such as savannas, are probably major sources of a variety of trace gases important in tropo- spheric chemistry. The potential importance of these regions stems from their geographical size, biological productivity, anoxic environments (in wet areas), and high turnover rates (largely by insects in dry areas). According to the criteria we have adopted in identifying biomes of potential importance, measurements of bio- logical emissions from both tropical wet and dry areas deserve a high priority. In tropical forests, investigations are needed on the fluxes of various nonmethane hydrocarbons, CH4, CO, N2O, and volatile sulfur-containing gases. Volatile spe- cies containing metals (probably methylated) should also be sought, and the potential for regional CO pro- duction from hydrocarbon oxidation must be explored through measurements and photochemical modeling. A variety of specialized approaches must be utilized: air- borne instruments must be deployed to determine the concentrations and fluxes of these gases above the forest canopy, sampling of emitted species from individual trees and plants is required, and airborne studies of the photochemical and cloud-mediated transformations in the forest plume must be undertaken. The latter studies would address questions related to the production and destruction of photochemical oxidants, e.g., 03, perox- ides, NOx, and CO. Similarly, investigations focused on cloud-water chemistry in these regions will result in a more complete understanding of the origins of a variety of organics and acids in precipitation collected in heavily forested areas remote from populated or industrialized centers. Because biological systems such as forests can act as sinks as well as sources for trace species in the troposphere, a final recommendation is for measure- ments of deposition of these species to tropical forests particularly of potential nutrients such as NH3 or am- monium ion (NH4 ), NOx or wet nitrate ion (NO3 ), SO2 PART II ASSESSMENTS OF CURRENT UNDERSTANDING or wet sulfate ion (SOT), 03, CO2 (to deduce exchange rates), and trace metals. The measurement of chemical fluxes from and to the tropical forests will be difficult. Base facilities must be established and local scientists with similar interests must be involved. Before a large, coordinated expedi- tion is undertaken in the tropics, methods for measuring concentrations and fluxes should be tested in more ac- cessible forests, for example, in North America. Be- cause qualitatively different emissions are likely to dis- tinguish tropical from temperate forests, investigations in both regions will be required. Dry tropical areas also display high cycling rates for nutrients. Although they store less material in their shrubs and grasses than is found in the wood of tropical forests, the biological material of savannas has a shorter lifetime and has higher nitrogen/carbon and sulfur/car- bon ratios than hardwood. From the rapid turnover rates, the chemical composition of the material, and the present data that suggest a large role for herbivorous insects, we conclude that significant volatile emissions are likely from certain dry tropical areas. In particular, there is potential for large emissions of CH4, CO, CO2, nonmethane hydrocarbons, N2O, and possibly methyl- ated metals. Initial measurements should focus on sites and processes that concentrate nutrients (termite colo- nies, for example), but the large land areas involved leave room for significant emissions from lower intensity sources distributed over large areas. The most difficult task will be to estimate total emis- sion rates from tropical forests. The general methods mentioned above and the use of meteorological towers are envisioned. About six full-time scientists and techni- cians would be needed. Measurement of chemical depo- sition to the tropical forests would require sustained ob- servations over a period of at least several months, both near the top of the forest canopy and at ground level for useful interpretation. In the dry tropical areas, wet ver- sus dry season differences may be marked. Two separate seasonal investigations or one longer investigation stretching through the wet and dry seasons would be required. Four full-time scientists would be needed for this study. Topical Areas Two distinct kinds of tropical blames are identified for concentrated research on biological sources important in tropospheric chemistry. These will be termed "wet tropics" and "dry tropics. " The potential importance of each of these environments and relevant investigations in each are outlined below. Separate discussions are presented for biomass burning and rice agriculture be- low.

CRITICAL PROCESSES Wet TropicalAreas Wet tropical areas are usually covered by tropical for- ests. These are evergreen or partly evergreen, they are frostfree and have average temperatures of 24°C or higher and rainfall rates of 100 mm or more per month for 2 of every 3 years. About 1.6 x 107 km2 of the earth's surface enjoys this climatic range, and (0.9 to 1 .1) x 107 km2 is actually covered by tropical forests now. These tropical forests contain about 60 percent of the global biomass, but 1 to 3 percent of it is permanently defor- ested annually, usually for agriculture, timber harvests, cattle grazing, and firewood. There are perhaps several million species of life in these forests, but only about 50O, 000 have been described. Barring irreversible cli- matic changes, it is possible that similar floristic species could be regenerated in perhaps 50 years after cutting. Approximately 80 percent of this biome occurs in only nine nations, five in South America and four in South- east Asia. These forests cycle nutrients rapidly through microrhizial (root) systems, and there are few inorganic reserves in their highly leached soils. Other indicators of the potential importance of tropical forests include their high net primary productivity, their potential~trigger effects of large changes, scientists' relative ignorance about them, and the existence of many anoxic sites. Several climatic effects could be triggered by changes in tropical forests because they mediate regional hydro- logic cycles. Cleared areas are more susceptible to pro- longed droughts and to much more erosion and flooding in wet periods than forested areas. A1SO7 these forests store CO2 and present a darker surface to sunlight than cleared areas. Potentially large emissions of gaseous hydrocarbons, N2O, CH4, reduced sulfur compounds, and possibly methylated metals and CO (see below) can be released from tropical forests. This potential arises from the prev- alence of anoxic sites in saturated soils and, during noc- turnal hours, in shallow surface waters. The largest po- tential source of CO could be the oxidation of biogenic hydrocarbons like isoprene (C5He), although several steps in the gas-phase oxidation of this and other hydro- carbons are not dear at this time. Effects of these emis- sions on the acidity of regional watersheds and rainfall, principally through oxidation of organics to formic and acetic acids, are likely and need investigation. D?:y TropicalAreas In dry tropical areas there are likely to be globally important biological sources of key gases. For example, in tropical savannas there is generally high net primary productivity although little of the fixed carbon enters into long-lived tissue (wood). The indigenous grasses 63 and shrubs are short-lived and are recycled by animals, largely insects. The chief species among these are ter- mites and ants; in some areas, termite mass densities per unit area exceed the highest mass densities of grazing animals anywhere. The potential for large emissions of CH4, CO2, CO, and perhaps other species from ter- mites, whose digestion is fermentative, is quite large. A1SO7 more speculatively, emissions of N2O and NOX gases are possible because nitrogen-fixing bacteria are known to live in termite guts. The roles of these crea- tures and of the relatively unexciting but extensive dry tropical areas in nature's atmospheric chemistry are ripe ~ . . . tor ~nvest~gat~on. Coastal Marsh, Estuary, ancl Continental Shelf Environments Coastal ecosystems are characterized by high biologi- cal productivity, active chemical and physical exchange, and transport driven by freshwater runoff, tidal forces, and wind-driven circulation. Existing data on emission of reduced sulfur gases, CH4, and N2O indicate that specific habitats in the coastal environment may be in- tense sources. Thus, though the areal extent of these sources may be relatively small, their contributions to the atmospheric budgets of certain reduced gases may be considerable. An additional consideration for placing priority on these environments is that higher fluxes and concentra- tions of many chemical species of interest place less de- mand on instruments (detection limits, response time, etc.~. For gases such as N2O, CH4, CO, COS, CS2, (CH3~2S, SO2, and a few others, it is reasonable to pro- pose initiating immediately a f~eld research program emphasizing basic processes of gas production from these source areas and their exchange with the atmo- sphere. For more reactive chemical species such as NH3, NO, and volatile metals, existing technology is probably inadequate for quantitative biosphere-atmosphere ex- change studies even in intense source areas. For almost all reduced gas species, measurement technology is cur- rently inadequate for studying very low-level sources and sinks. The primary objective ofthese studies is to develop an improved quantitative understanding of the processes that control the production and consumption of bio- genic gases in coastal wetland and aquatic environments and their exchange with the lower troposphere. When this information is available, it should be possible to extrapolate to regional and global flux estimates with supporting data from remote sensing and other geo- graphical and meteorological data bases. The proposed studies will require measurement of a wide range of biological, physical, and meteorological

64 variables at a selected set of coastal habitats. Five specific habitats of importance are (1) salt marshes, (2) man- grove swamps, (3) sea grass, (4) kelp, and (5) exposed mud and soil surfaces. Variables that can influence trace gas production and consumption in soils and sediments and their exchange with the atmosphere include biologi- cal community composition and dynamics, nutrient quantities, inputs of energy from different sources, re- moval of wastes by tidal forces and river runoff, temper- ature, salinity, pH-Eh, sediment physical properties, and wind stress on water and soil surfaces. Previous studies have demonstrated high variability in time and space. A quasi-continuous research pro- gram at each coastal habitat of interest will be required to accomplish the objectives of this experiment. One efficient way of conducting studies on biogenic gas sources in coastal environments would be to use NSF- LTER (Long-Term Ecological Research) sites for the proposed measurement program. Ongoing activities at these sites would provide supporting biological, mete- orological, and geochemical data required for the com- prehensive gas production and exchange studies pro- posed here. Point Sources There is a class of potentially important biological and surface sources that are either intense and spatially small, or qualitatively distinctive in their emissions. Ex- amples include lightning, industrial emissions such as combustion or waste plumes, volcanoes, and animal feedlots. A further source is biomass burning, an activ- ity that is widespread in certain regions at sites whose exact locations vary from year to year. Biomass Burning Although not a biome but a process, biomass burning is included in the present discussion because of its great potential importance and distinctive characteristics. Qualitatively, biomass burning may be regarded as a type of nonindustrial pollution. Many types of biomass burning combine to yield a large total of biomass burned annually. For example, biomass burning is used to clear tropical forests for agriculture, to prepare forested areas for settlements, and to dispose of agricultural wastes (e.g., sugar cane). Large quantities of biomass are burned as fuel in industries, for individual human needs, and in wildfires. Recent estimates of the annual global area involved in biomass burning range from 3 to 7 x 1 o6 km2, with estimates of the total biomass burned ranging from 4400 to 7000 Tg/yr. Although total bio- mass burning quantities are probably uncertain to within a factor of 2 or 3 and vary from year to year, they PART II ASSESSMENTS OF CURRENT UNDERSTANDING are almost certainly significant in the global atmo- spheric carbon cycle and probably in other cycles as well, e.g., oxygen and nitrogen. Ecologically, pro- nounced changes accompany deforestation through bio- mass burning, e.g., in flora, soil structure, and surface hydrology. Much biomass burning for deforesting oc- curs in areas that are not well characterized ecologically. Surface albedo values, surface winds, and turbulence are also affected. Certain types of biomass burning also produce charcoal, thus effectively constituting a CO2 sink. From a physical and chemical point of view, biomass burning is a high-temperature process that is dramatic both in quality and in quantity. Large amounts of mate- rial are transformed, mobilized, and volatilized quickly. Partially combusted particles become airborne, and a wide spectrum of gases are produced in the flames and through the process of smoldering. Gases containing carbon, hydrogen, oxygen, nitrogen, sulfur, halogens, phosphorus, and trace metals are involved. Many of the gaseous species so produced are highly reactive photo- chemically, but stable gases like CO2, CH4, N2O, and CH3C1 are also generated. The photochemically active species are known to give rise to rapid O3 production and probably yield other photochemical smoglike spe- cies, e.g., peroxyacetylnitrate(CH3COO2NO2 (or PAN)) Not surprisingly, a number of carcinogenic sub- stances are also produced in biomass burning; gases include benzene (C6H6) and toluene (C7He), and re- lated airborne solids are certain to be found. Further, it now appears that several oxygenated hydrocarbons from biomass burning yield organic acids in sufficient quantity to acidify precipitation and groundwaters re- gionally. There is a great deal of fundamental research to be done on the atmospheric chemistry effects of biomass burning. The full spectrum of compounds injected into the atmosphere in this way needs description. Quantita- tive production rates of C H4, C2H6, and other alkanes, N2O, C H3C1, aldehydes, ketones, nitrogen oxides, N H3, HCN,C H3CN, oxides of sulfur, CO, CO2, H2, and several other species must be determined. Methods need to be developed to quantify the various production and atmospheric injection rates; simply ratioing each species to CO2 and deducing CO2 emission rates might not suffice. Fortunately, some existing data suggest that the relative yields of some key gases (alkanes, alkenes, aldehydes, ketones) in temperate and tropical forest fires are not terribly different, so we can reasonably propose to concentrate initial research in temperate latitude ar- eas where logistics are less of a concern. For example, there are controlled (or prescribed) forest fires managed by experts in Georgia and Oregon that might be suitable for several studies. Obviously, certain distinctive fea

CRITICAL PROCESSES lures of tropical forests (e.g., smoldering in damp ar- eas), tropical agricultural products (e.g., sugar), and tropical grasses (e.g., the high cyanide content of sor- ghum) will eventually require targeted research expedi- tions. Global considerations such as for long-lived gases (CH4, CO2, N2O, CH3Cl, and COS) will also demand that the size of areas (biomass) being burned annually be quantified region by region. Lightning The most dramatic types of atmospheric electrical discharge undoubtedly produce certain gases and de- stroy others in their intense pulses of thermal and optical energy. Spectroscopic and chemical analyses have de- tected many interesting products of lightning. The sin- gle most pressing question today concerns the amount of fixed nitrogen (as NO) that is produced annually by atmospheric electrical discharges. Estimates of this quantity vary widely; they cover the range from major, i.e., comparable to combustion production of NOX, to smaller but significant fractions of this quantity. Key uncertainties are whether laboratory discharges simu- late the full range of atmospheric electrical phenomena adequately and the actual frequency of atmospheric dis- charges. Microdischarges near pointed biological sur- faces (e. g., pine needles) might also have important con- sequences. Volcanoes By mass fraction, the principal emissions from volca- noes are H2O and CO2. A variety of other gaseous and particulate substances is also emitted in quantities that are potentially significant to the regional and global at- mosphere. These include SO2, several volatile heavy metals, and ash particles. Through large explosive events, volcanic inputs and impacts can be enormous if short-lived. By their nature, these effects defy predic- tion, and once they occur, they defy averaging that is, it is not easy or particularly meaningful to compare volcanic emissions to annual averages from other sources. Animal Feedlots Volatile losses of feed nitrogen (principally as NH3, R-NH2, and possibly N2O and NO) and organic sulfur compounds from animal feedlots are appreciable, at least when expressed as a fraction of the feedlot's animal waste and possibly in absolute terms. In the United States, perhaps 40 percent of all fertilizer nitrogen is consumed by cattle, and the average length of time that cattle spend in feedlots or other concentrated popula 65 lions is high. It is likely that regional impacts of these emissions are significant, especially for levels of gaseous NH3, particulate NH4, and NH4 in precipitation. If so, the atmospheric transport of NH3 represents an air- borne fertilizer distribution system. Industrial Emissions This group includes combustion products, wastes from chemical production processes, mineral, gas and oil exploration and refining, burning or processing of wastes, and losses of solvents. As sources of atmospheric gases and particles, these processes can be distinctive both in kind and size, for example, by emitting unnatu- ral substances or natural ones in quantities that are somehow comparable to natural cycling rates. For sev- eral clearly important gases like NOX species and SO2, emission inventories are available for most industrial- ized countries, and some of these have been prepared with good spatial resolution (100 km x 100 km). Changing industrial intensities, processes, and practices dictate that these emission inventories will require up- dating. Rice Agriculture The potential importance of rice paddies as sources of atmospheric chemicals might not seem obvious. Several ~ - ~ . . . lines ot 0 Elective reasoning and preliminary field data combine to argue strongly in this direction, however. First, a number of our criteria (expressed above) indi- cate that rice agriculture represents a globally significant biome. These include physical area, reasonably high primary productivity rates, and the anoxic character of rice paddy soils. As a principal staple in world food diets, rice is extremely important in world agriculture. Ac- cordingly, large areas are cultivated- 1.3 x lo6 km2 in the late 1960s and 1.45 x 106 km2 in 1979. Further, in many countries new emphasis has been placed on multi- ple cropping through improved irrigation. Two crops per year is becoming the norm in tropical areas. One practical result of this is that rice paddy soils are under- water for perhaps 8 months instead of 4 months annu- ally, and the more negative redox potential of water- covered soils allows more reduced gases like CH4, NH3, and N2O to form. Indeed, the rice paddy environment is ideal for the evolution of a number of volatile species containing nutrient elements. The soils are oxygen-poor (strongly reducing), and they are nutrient-rich, often through fertilization. Direct indications of the impact of rice paddies on the global atmosphere have centered on CH4. In the early 1960s, a Japanese scientist showed that paddy soils, when cultured in the laboratory, released copious quan

66 titles of CH4, especially at the elevated temperatures that characterize midsummer temperate-latitude and tropical soils. Extrapolations ofthese results implied that rice paddies were responsible for perhaps one-third of all atmospheric CH4. Later field studies of CH4 emission rates continue to suggest the potential global importance of rice agriculture, but they find several complicating factors that require attention. For example, the emitted flux of CH4 depends on the amount of nitrogen fertilizer (and possibly on the type of fertilizer) used. The princi- pal mode of CH4 emission is direct transmission from rice roots upward through the (hollow) plant, and less by diffusive or bubble transport across the water-air inter- face (although all three modes are active). Relatively more CH4 is released late in the growing season. Other factors are doubtless involved, e.g., soil organic content Reliable global source estimates must recognize these . . . complexities. Volatile nitrogen emissions are also thought to be products of rice agriculture. This is mostly due to evi- dence that rice utilizes fertilizer nitrogen relatively inef- ficiently, 50 percent or less. On the basis of observations of low leaching rates, one concludes that nitrogen is lost in large amounts, probably as NH3 and/or N2O, because of the reducing nature of waterlogged, water- covered paddy soils. Nitric oxide emissions would not be inconceivable. Direct field studies are needed to deter- mine the principal nitrogen-bearing volatile species. If it is NH3, the effects would be regional; if emitted as N2O, then a global impact is possible. Once again, several complexities must be recognized: variations in fertilizer types, application protocols, soil types, and so on. While preliminary data suggest that ethane (C2H6) emissions from rice plants are not significant, those of isoprene (COHN) might be. Methylated metals could be released easily by rice, and volatile phosphorous compounds deserve some attention as well. Productive Oceans The oceans are a large-area, low-intensity source of reduced sulfur compounds to the lower troposphere. Preliminary studies have identified a number of com- pounds in seawater that are presumed to be of biogenic origin including H2S, COS, CS2, dimethyl sulfide (DMS), dimethyl disulfide (DMDS), and dimethylsulf- oxide (DMSO). Dimethyl sulfide appears to be one of the most abundant species, contributing an estimated flux of 34 to 56 TgS/yr to the marine boundary layer. Once in the marine boundary layer, DMS is probably oxidized by photochemical processes to produce SO2 with intermediates such as DMSO and methane sul- fonic acid. Qualitatively, the concentration of reduced sulfur compounds in surface seawater correlates with PART II ASSESSMENTS OF CURRENT UNDERSTANDING indicators of algal biomass. However, in one study no relationship was found between DMS in surface water and in the overlying atmosphere. A research effort to investigate the sulfur cycle in productive areas of the ocean should be initiated to elucidate sources of reduced sulfur in the ocean and their role in the global sulfur cycle and budget. A program to investigate the sulfur cycle in produc- tive areas of the open oceans should include a multidis- ciplinary team to study in situ biogenic production of sulfur species in the water column, fluxes across the sea- air interface, and chemical processes in the marine boundary layer determining transport and fate. Most of the research would be conducted from a major research vessel dedicated to the project, with simultaneous air- craft overflights to obtain information on the vertical distribution of sulfur species in the troposphere and rates of exchange between the boundary layer and free tropo- sphere. Initial sites for these studies might include the ocean area off the west coast of the United States, conti- nental shelf waters off the southeastern United States, and a major upwelling area. ISSUES OF CLOSURE Preceding discussions on biological and surface sources and following sections of this report on selected tropospheric chemical cycles illustrate the complexity of quantifying the role of biological processes and other surface sources in global atmospheric chemistry. Suff~- cient data are available on temporal and spatial variabil- ity in emissions of biogenic gases such as N2O, CH4, and (CHRIS to raise the issue of limits of understanding and predictability. For example, although it is established that emissions from soils are a significant source of global tropospheric N2O, measured emission rates vary by at least a factor of 400. Nitrous oxide is produced during both denitrification and vitrification processes with emissions to the atmosphere influenced by a wide range of biochemical, soil physiochemical, climatic, and land-use variables. Emissions of N2O may be episodic, with significant variations on time scales as short as hours and space scales of meters (e.g., see Figure 5.1~. Similar ranges of variability may be typical of CH4 emis- sions from diverse biological sources such as rice pad- dies, natural wetlands, and termite mounds (e.g., see Figure 5.2~. The complexity and variability of emissions of biogenic gases from soil derive from nonlinear inter- actions of the wide range of environmental factors that control the metabolism of the source organisms and the physical process of gas exchange between the soil and atmosphere. Biogenic gas emissions from aquatic envi- ronments (e.g., (CHRIS emissions from the ocean) are

CRITICAL PROCESSES 3 2 1 o 3 2 1 n (a) _'' _~'~14 ) ~ (b) . AA I ~ AN A ~ _ a, v Am, ~ ~ I ~ ~~ ~ ~ ~ Am_ ~ r A ~ ~ , Q o o <' ~ Q ~' Q a: ~ O Z ~ ~ 11 MONTH 1979-1980 FIGURE 5.1 Emissions of nitrous oxide from (a) corn field and (b) fallow field. Note the episodic character of the emissions (data from Duxbury et al., 1982~. less well studied, but can be expected to be equally varia- ble in time and space. The problem of "patchiness" in the distribution of marine bacteria and phytoplankton, which are two groups thought to be most involved in the production of reduced sulfur compounds in seawater, is a long-standing one in aquatic ecology. LL us I 20 oh ~ O 16 _ Cal ~ is ~ 12 _ d. o I ~ 8 _ ~ O 4 - CI) ° O _ X - ~ ,,0, 4 _ lo o us ~ 6 Z UJ I LL 8 ~ I LL ~ ~ O _~_ \J vet O z c] ~ ~4 ~ ~ ~ ~ ~ <( ~ o z ~ ~ Q MONTH 1980-1 981 FIGURE 5.2(a) Methane emissions from organic soils in the Great Dismal Swamp, Virginia (data from Harriss et al., 1982~. 67 Most research to date on surface source strengths has focused on "hot spots" (i.e., CH4 emissions fron anoxic sediments, mineral aerosols derived from intense dust storms). Inadequacies in flux measurement technology and techniques have limited studies of low-intensity, large-area sources (e.g., CH4 flux from slightly super- saturated open ocean surface layers and NH3 flux from nonagricultural soils). Recent developments in high- sensitivity, fast response detectors, aircraft micromete- orological measurement capabilities, and ground-based flux chamber design offer new opportunities for studies of sources of atmospheric O3, N2O, CO, and CH4. For most C, N. and S species, accurate flux measurements are not currently possible. To achieve the goal of a better understanding of global tropospheric chemistry requires that both the magni- tude of surface sources at their origin and the subsequent transport from the planetary boundary layer to the free troposphere be determined. For certain reactive re- duced gas species, removal processes in the boundary layer may restrict the significant influence of a particular species to regional or hemispheric scales (e.g., HIS, CO, and C5He). Processes driving exchange from the bound- ary layer to the free troposphere are largely decoupled from processes controlling emissions from the surface; thus an experimental design to quantify fluxes of gas or aerosol species from surface sources to the free tropo- sphere must take into account a very complex set of potentially nonlinear, episodic variables. Several problems raise the issues of the limits of pre- dictability and of scientists' ability to extrapolate from field measurements. These problems are inadequate methodologies for flux measurements and nonlinear in- teraction of most biological, geochemical, and meteoro- logical variables that determine flux rates at the surface and through the boundary layer. Q Q 14n 1 1 1 1 1 1 20 F Fertil izer / 100 _ 80 == 60 _ C: 40 20 o Fertil izer 140 kg~ 1_ 4~~~~ Fertilizer | - 0 1 2 3 4 ELASPED TIME (hrs) FIGURE 5. 2(b) Emissions of methane from fertilized and unfer- tilized rice plants (data from Cicerone and Shetter, 1981~.

68 Prediction of global emissions of most biogenic gases from surface sources may be somewhat analogous to weather prediction short-term, local-to-reg~onal proc esses are amenable to fundamental understanding through long-term studies with a combination of in situ and remote sensing techniques; long-term, global pre dictions will evolve through the development of a statis tical chemical climatology of the troposphere. BIBLIOGRAPHY Ahrens, L. H. (1979~. Origin and Distribution of the Elements. Perga- mon, New York, 537 pp. Ajtay, E. L., P. Ketner, and T. Duvigenaud (1979~. Terrestrial primary reduction and phytomass, in The Global Carbon Cycle. Wiley, New York. Andreae, M. O., and H. Raemdonck (1983~. Dimethyl sulfide in the surface ocean and the marine atmosphere: a global view. Sczence 221: 744-747. Bonsang, B., B.-C. Nguyen, A. Gaudry, and G. Lambert (1980~. Sulfate enrichment in marine aerosols owing to biogenic gaseous sulfur compounds. I. Geophys. Res. 85: 7410-7416. Bremmer, J. M., and A. M. Blackmer (1978~. Nitrous oxide: emis- sion from soils during vitrification of fertilizer nitrogen. Science 199:295-296. Broda, E. (1975~. The Evolution of the Bioenergetic Process. Pergamon, Oxford, 211 pp. Broecker, W. S., T. Takahashi, H. M. Simpson, and T.-H. Peng (1979~. Fate of fossil fuel carbon dioxide and the global carbon budget. Science 206:409-418. Burns, R. C., and R. W. F. Hardy (1975~. Nitrogen Fixation in Bacteria arid Higher Plants. Springer-Verlag, New York, 189 pp. Chapman, V. I. (1977~. Wet Coastal Ecosystems. Elsevier, Amster- dam. Cicerone, R. I., and I. D. Shetter (1981~. Sources of atmospheric methane: measurement in rice paddies and a discussion. I. Geophys. Res. 86:7203-7209. Crutzen, P. I., L. E. Heidt, I. P. Krasnec, W. H. Pollock, and W. Seller (1979~. Biomass burning as a source of atmospheric gases CO, H2, N2O, NO, CH3C1 and COS. Nature 282:253-256. Delwiche, C. C. (1981), (egg. Denitrification, Nitrification and Atmo- sphenc Nitrous Oxide. Wiley, New York, 286 pp. Delwiche, C. C., and G. E. Likens (1977~. Biological response to fossil fuel combustion products, in Global Chemical Cycles and Their Alterations by Man, Werner Stumm, ed. Dahlem Konferenzen, Berlin, pp. 73-88. Duxbury, J. M., D. R. Vauldin, R. E. Terry, and R. L. Tate (1982~. Emissions of nitrous oxide from soils. Nature 298:462- 464. Ehhalt, D. H., and U. Schmidt (1978~. Sources and sinks of atmo- spheric methane. Pure Appl. Geophys. 116:452-464. PART II ASSESSMENTS OF CURRENT UNDERSTANDING Friend, J. P. (19733. The global sulfur cycle, in Chemistry of the Lower Atmosphere, S. I. Rasool, ed. Plenum, New York, 335 pp. Garrels, R. M., A. Lerman, and F. T. Mackenzie (1976~. Controls of atmospheric O2 and CO2: past, present and future. Amer. Sci. 64:306-315. Harriss, R., D. I. Sebacher, and F. T. Day, Jr. (1982~. Methane flux in the great dismal swamp. Nature297:673-674. Holland, H. D. (1978~. The Chemistry of the Atmosphere and Oceans. Wiley, New York, 351 pp. Hutchinson, G. L., A. R. Mosier, and C. E. Andre (19829. Ammo- nia and amine emissions from a large cattle feedlot. I. Environ. Qual. 11:288-293. Jorgensen, B. B. (1977~. The sulfur cycle of a coastal marine sedi- ment (Limfjorden, Denmark). Limnol. Oceanogr. 22:814-832. Junge, C. E. (1972~. The cycle of atmospheric gases natural and man-made. Quart. I. Roy. Meteorol. Soc. 98:711-729. Kellogg, W. W., R. D. Cadle, E. R. Allen, A. L. Lazrus, and E. A. Martell (1972~. The sulfur cycle: man's contributions are com- pared to natural sources of sulfur compounds in the atmosphere and oceans. Science 175:587-596. Kvenvolden, K. A., ed. (1974~. Geochemistry and the Origin of Life. Dowden, Hutchinson & Ross, Stroudsburg, Pa., 422 pp. Li, Y.-H. (1972~. Geochemical mass balance among lithosphere, hydrosphere, and atmosphere. Amer. J. Sci. 272:119-137. Lovelock, J. E., R. J. Maggs, andR. A. Rasmussen(1972~. Atmo- spheric dimethyl sulfide and the natural sulfur cycle. Nature 237:452-453. Mann, K. H. (1982~. Ecology of Coastal Waters: A Sys~erns Approach. University ofCalifornia Press, Berkeley, Calif., 322 pp. Margulis, L., and J. E. Lovelock (1978~. The biota as ancient and modern modulator of the earth's atmosphere. PureAppl. Geophys. 116:239-243. Martell, E. A. (1963~. On the inventory of artificial tritium and its occurrence in atmospheric methane. J. Geophys. Res. 68:3759- 3769. Ponnamperuma, C. (1977~. Chemical Evolution of the Early Precam- bnan. Academic, New York, 221 pp. Rasmussen, R. A. (1974~. Emission of biogenic hydrogen sulfide. Tellus 26: 254-260. Rasmussen, R. A., L. E. Rasmussen, M. A. K. Khalil, and R. W. Dalluge (1980~. Concentration distribution of methyl chloride in the atmosphere. J. Geophys. Res. 85: 7350-7356. Redfield, A. C. (1958~. The biological control of chemical factors in the environment. Amer. Sci. 46: 205-221. Singh, H. B., L. J. Salas, and R. E. Stiles (1983~. Methyl halides in and over the eastern Pacific (40°N-32°S). J. Geophys. Res. 88:3684-3690. Soderlund, R., and B. H. Svensson (1976~. The global nitrogen cycle, in Nitrogen, Phosphorus and Sulphur-Global Cycles. SCOPE Report 7. 13. H. Svensson and R. Soderlund, eds. Ecol. Bull. Stockholm 22:23-72. Walker, J. C. G. (1977~. Evolution of the atmosphere. Macmillan, New York,318 pp. Whittle, K. J. (1977~. Marine organisms and their contribution to organic matter in the oceans. Mar. Chem. 5:381-411.

69 GLOBAL DISTRIBUTIONS AND LONG-RANGE TRANSPORT BY I. M. PROSPERO AND H. LEVY II The ultimate objective of the Global Tropospheric Chemistry Program is to understand the global chemi- cal cycles of the troposphere to such a degree that the response of the atmosphere to natural or man-made perturbations can be predicted. In order to do this, a thorough knowledge of the chemistry of the species in question and of the physical behavior of the atmosphere is needed. At the simplest level, the linkage to atmo- spheric conditions is implied in any chemical process that involves solar radiation or water in any of its states. The linkage between meteorology and atmospheric chemistry becomes more explicit when the variable of time becomes a factor in the problem. Because the atmo- sphere is never still, the time dependence of chemical processes will also invoke a space dependence. Meteorological processes are intimately linked with chemical processes over a wide range of time and space scales. In the early days of atmospheric chemistry, when the subject was synonymous with pollution chemistry, field experiments combining chemistry and meteorol- ogy were carried out on a scale commensurate with that of the perceived problem; consequently efforts focused on studies of pollutant dispersion from relatively local- ized sources such as smoke stacks or industrial com- plexes. Later, pollution began to be viewed as a regional problem, and experiments were designed to study pollu- tant dispersion on a regional scale. A number of experi- ments of this type were carried out with some degree of success. It has come to be realized that the scale ofthe pollution problem is much greater than that of a particular urban region. For example, the area affected by acid rain en- compasses the entire northeastern United States and much of eastern Canada. Some of the major industrial sources for the sulfur and nitrogen oxides that are re- sponsible for the increased acidity are located in the central United States. A similar situation was recog Sized much earlier In Europe, where the acid rain pron- lem in some countries is clearly attributable to a major extent to materials injected into the atmosphere in an- other country. Thus the subject of atmospheric chemis- try in general, and atmospheric pollution in particular, must be viewed as one that transcends national bounda- ries. Transport studies must be planned accordingly. Before the long-range transport of chemical species in the atmosphere can be understood and predicted, a good understanding of global meteorology is needed. Fortunately, tremendous progress has been made in this area over the last decade or so. A number of models have been developed that have been used (or could be used) for predicting transport. At this time a number of gen- eral circulation (climatological) transport models exist that have been applied to a limited range of tropospheric chemistry studies with mixed results. Unfortunately, there is no chemical data set that can be used to validate these models. Thus the future development of these and other models is severely hampered. It is because of this need that we established our first objective: 1. To obtain data on the distribution of selected chemical species so as to be able to characterize impor- tant meteorological transport processes and to validate and improve the ability of models to simulate the long- range transport, global distribution, and temporal vari- ability of selected chemical species. This first objective is primarily meteorological in na- ture. The second objective focuses on the needs of the atmospheric chemists studying the various cycles: 2. To obtain data on the distribution of those chemi- cal species that play an important role in the major chemical cycles of the troposphere. The third objective derives from a concern about the possible impact that some chemical species might have on weather and climate. To this end we recommend measurements: 3. To establish, quantify, and explain long-term trends in the concentration and distribution of environ- mentally important trace gases, especially those that are radiatively active. These three objectives could be met by establishing a global network comprising three different types of sub- networks: the Global Distributions Network; the Sur- face Source/Receptor Network; and the Long-Term Trends Network. The configuration of these networks and the species to be measured in each are discussed in the Global Distributions and Long-Range Transport experiment description in Part I, Chapter 3 of this re- port. At this time we will discuss only the broader philo- sophical aspects of setting up networks of this sort. A1- though the general subject of atmospheric models is fundamental to our topic, it will not be discussed here, as it is adequately covered later in Part II, Chapter 6. Likewise, we will not deal with any of the chemistry of the species that are designated in the network protocol because this subject is covered in Part II, Chapter 7 of this report.

o L ~ j.: . ~ hi_ or o o o o CD ~ (: '1 1 w~ 1 1 . it, ~·; . if' o CO o .,~L ., · ~ . ., ~ ~- . ·f . for ~ . _~ o To . or o Cal C9 z so Cal o o o CD °c - w . ~ . ~ ~ l OR it 1 . o A. o In or oo ; , ~- , - ~ l ~7 ~ . o ~ - ~l o ~ ~ ~ - = = v : ~ = . -- ~ ~ · ·~: '1 ~ ~ · -~ o o o o o o CD ~ 70 i. ~ ~n z - .s U) z o ~_ J~ O~ o CL(,, _ CE ~ ~ o ~r ° ° - ,_,, ~ - ~ Z Q ~ o O O ~ o - UJ rr · ~ O z O 6 ~n z O J J o CL - o 111 ~ 7 {1) _ Q J O LL ~ cn - m ~ o o 00 ~o 1 0 1 ~ ~100 loo <1 .° 1 ~ 1 -o 1 ~ _I 1 s°- 1 1 ~ ~Q \1 lo J o 1 0 ~ 1 ~o I I ~ 1 , ° °o ~n 0 1 · ~ 1 ~ 1 0 1 ~ ~o ~ 1 ~ 1 ~ . , ~ I ~ 1 ~ ~oo ~ 1 ~ 1 1 1 10 ~mo ~ - ) ~ o - ~ o o ' - ~5 - ~ o t ~ ~ - ~ o 10 1 0 o 1 o l

CRITICAL PROCESSES MONITORING CONCEPTS The word "monitor" is defined as "to watch, ob- serve, or check especially for a special purpose. " Thus a network of stations engaged in a routine program of sampling could be rightfully described as a monitoring program. Monitoring is an activity that in the past has been viewed with a certain amount of disdain by many scientists. The poor reputation that monitoring had earned for itself was due perhaps to the fact that some . . monltormg programs were poorly conceived and badly executed; nonetheless, they were permitted to continue. This poor image persists in some circles despite the fact that some of the most exciting developments in atmo- spheric chemistry in recent years have been derived from programs that involved a monitoring effort. The most outstanding example is that of CO2. The excellent and vitally important record that exists for CO2 concen- trations in the atmosphere is primarily due to the efforts of one person, David Keeling, who persisted in his work despite the initial indifference of many in the scientific community and despite the sometimes flagging fund- ing. It is relevant to our thesis that the CO2 record is important not only for what it says about the carbon cycle but also for what it says about meteorological and oceanographic processes. An excellent example of a monitoring program that has had a major impact on knowledge of the atmo- spheric circulation is the Department of Energy's (for- merly Atomic Energy Commission's) high-altitude sampling program for radioactive materials. This pro- gram was initiated in response to the intensive nuclear weapons testing of the 1950s and 1960s. In effect, the individual nuclear detonations injected a point-source tracer pulse into the stratosphere. Circulation patterns could be determined by monitoring the subsequent dis- persion of the tracer. By monitoring the total budget of radioactive materials in the stratosphere as a function of time, especially after the prolonged interruption and subsequent cessation of atmospheric testing, it was pos- sible to calculate the residence time for stratospheric materials. As a consequence of this program of system- atic aircraft and balloon flights, important stratospheric transport processes were identified and quantified. Without these data it would have been extremely diff~- cult to make any realistic evaluation of the impact of anthropogenic emissions on stratospheric chemistry such as those assessments made for supersonic-transport emissions, halocarbons, and other materials. The cur- rent stratospheric chemistry models are, in effect, based on meteorology derived from these radionuclide tracer measurements. To provide further support for the concept of monitor- ing for the sake of investigating processes, we cite an 71 example from the field of oceanography. For over a de- cade, a number of programs have focused on the study of the distribution of i4C and 3H (tritium) in the oceans. These radionuclides were produced in very high yields during the period of intense weapons testing. Much of this material was subsequently deposited in the oceans in the high latitudes. After injection into the sea surface, the nuclides follow the movement of the water masses. The distribution of these radioactive materials has been measured periodically and systematically in a series of programs that have been carried out over the years. As a result of this work, there is now a much improved pic- ture of transport and mixing processes in the oceans, and detailed ocean-mixing models have been devel- oped. This new knowledge of oceanographic processes has been vitally important in developing the models of the global CO2 cycle that are critical to assessing the impact of CO2 on climate. Thus it is clear that carefully planned and executed monitoring efforts can produce data that can be used to develop new insights into chemical and physical proc- esses. The Global Tropospheric Chemistry Program network will, in effect, use chemical tracers in much the same way as radionuclides were used in the cited exam- ples. There are a number of monitoring networks cur- rently in existence. The most extensive and ambitious is carried out under the United Nations Environment Program (UNEP), one of whose tasks is global environ- mental assessment. As a part of the task of information gathering, a Global Environmental Monitoring System (GEMS) was established. Some of the activities carried out under GEMS are as follows: 1. Health-related monitoring of pollutants in air, food, water, and human tissues (e.g., SO2, heavy metals, chlorinated hydrocarbons); 2. Climate-related monitoring ofvariables that could effect climate change (e. g., CO2, atmospheric turbidity, "lacier messes, albedo); 3. Monitoring of long-range transport of pollutants (e.g., sulfur and nitrogen oxides and their transforma- tion products). In pursuit of these objectives, UNEP supports the Background Air Pollution Monitoring Network (BAP- MoN), which is operated under the World Meteorologi- cal Organization (WMO). The BAPMoN network consists of three types of stations: regional, continental, and baseline (or global). Regional stations are located in rural areas far from population centers so as to be mini- mally affected by anthropogenic emissions. Continental stations are situated in places where no significant changes in land use practices are anticipated for a con- siderable time. Baseline stations are located in remote

72 pristine areas such as islands and mountains. The distri- bution of BAPMoN stations is shown in Figure 5.3. At regional stations the program consists of sampling of wet precipitation by automatic rain gauges and the measurement of atmospheric turbidity by sun photome- ters. At a limited number of locations, measurements of suspended particles are made by means of high-volume air samplers. Generally, the precipitation (and sus- pended particle) samples are grouped and analyzed on a monthly basis. However, some national programs ana- lyze on a shorter term basis. In the United States, there are 12 BAPMoN stations, which are operated by NOAA. These stations (and those in Canada) operate on a weekly schedule. The United States and Canadian stations do not sample suspended particles. Baseline and continental stations, in addition to following the same program as the regional stations, also measure COo and, in some cases, other constituents such as SO2 and NOX. Baseline stations are usually operated as research stations by the managing country. Some stations such · ~ · . ~ T ~ ~ a_ ~ ~ ~ ~ ~ · · ~ _ ~ as those in the up.;. Floral monitoring tor Climate Change (GMCC) program, also monitor other trace gases (such as the halocarbons, CO, and 03~. As of 1981, the WMO network consisted of 100 re- gional, 12 continental, and 12 baseline stations operat- ing in 49 countries (Figure 5.3~. It is hoped that the network will ultimately include about 160 stations in 90 countries. The WMO network is an impressive achievement considering the limited funding available to support its operation. In 1981, the entire GEMS program was budgeted at $2 million a year. However, from the standpoint of the objectives of the Global Tropospheric Chemistry Program in general and the Global Distribution/Long-Range Transport Program in particular, the WMO network leaves much to be desired. First, the major emphasis in the network protocol is placed on pollution. Consequently, many species that are important to atmospheric chemical processes are ignored. Second, according to the network protocol, the chemical analysis activity focuses primar- ily on precipitation; suspended particle sampling is car- ried out at only a select subset of stations as a result of national initiatives. Gas sampling (other than for CO2) is even more limited and is usually carried out only at baseline stations. Third, the BAPMoN protocol is based on a monthly sampling frequency; consequently, these data could not be used for developing and validating event-transport models. Last, the BAPMoN network is concentrated on continental areas in the northern hemi- sphere. There are only seven operational stations in the southern hemisphere, and of these only four are located south of 15°S: Cape Grim, Tasmania; Lopez, Argen- tina; Amsterdam Island; South Pole. This unbalanced PART II ASSESSMENTS OF CURRENT UNDERSTANDING distribution of stations is a major deficiency from the standpoint of the Global Tropospheric Chemistry Pro . ~ gram 0 electives. There are a number of other networks operating in the United States. The most widespread is the National Air Surveillance Network (NASN), operated by the En- vironmental Protection Agency. The primary responsi- bility for the operation of this network resides with the states; the primary purpose of the network is to monitor air quality and to verify compliance with the national air quality standards. A number of species measured in the NASN program are relevant to the Global Tropospheric Chemistry Program: SO2, CO, NO2, 03, hydrocar- bons, and total suspended particles. The basic device for measuring total suspended particles is the high-volume (hi vol) filter sampler; as of the late 1970s, there were several thousand of these in operation at NASN sites. The number of stations measuring the various gaseous species exceeds several hundred, with over a thousand measuring SO2 routinely. Most NASN stations are situ- ated in urban areas; however, a significant number are located in suburban and remote areas. Some of the data obtained from the NASN program will be very useful for modeling purposes especially from the standpoint of providing information on atmospheric source strengths. Unfortunately, the data obtained from the high-volume particle samplers are almost useless for any purpose other than determining total suspended particle concentrations. First, sampling is too infre- quent (one day in six), and second, the filter sampling medium generates artifacts that preclude meaningful chemical analysis for important species such as SO4-, NO3, and other trace materials. A rather extensive network of precipitation-sampling stations has evolved in the United States over the past few years as a consequence of increased concerns about acid precipitation. These studies are carried out as a part of the National Acid Precipitation Assessment Pro- gram. The national network is actually a composite of many smaller networks, each of which is operated by various agencies (EPA, NOAA, the Department of En- ergy, and the Department of Agriculture) and university groups. The operational protocol does not call for any concurrent air-sampling studies, although some work of this nature is carried out as a part of individual research efforts. The WMO BAPMoN program and other related pollution-monitoring programs have been criticized for their strategy and their tactics. The deficiencies in these programs are to a certain extent inherent in the terms of reference mandated by the governing bodies. Such pro- grams evolved in the crisis atmosphere that was gener- ated by the environmental movement in the 1960s. As a consequence of the demands to take action on these

CRITICAL PROCESSES issues, programs were instituted that have produced tremendous quantities of data. Unfortunately, many of the data are difficult to interpret in terms of chemical and physical processes and of environmental impact. The shortcomings of these national and international monitoring programs are now generally recognized. In the United States, a National Research Council study group (Committee on National Statistics, 1977) on en- vironmental monitoring has recommended changes in the EPA's programs, which they found to be deficient on a number of grounds: they were not firmly based on scientific principles; they gave insufficient attention to discovering or anticipating pollutants; their efforts were often fragmented and uncoordinated. Similar study groups in other countries have reached similar conclu- sions. As a consequence of these perceived shortcom- ings, a number ofchanges are imminent in some ofthese programs. Most significant from the standpoint of the objectives of the Global Tropospheric Chemistry Pro- gram is a move to implement high-volume air sampling for suspended particles at BAPMoN stations and to place all sampling on a weekly rather than monthly schedule. There are a number of monitoring programs of more limited scope that have been relatively successful. The NOAA GMCC program and the Atmospheric Lifetime Experiment (ALE, sponsored by the Chemical Manu- facturers Association) have monitored concentrations of a number of halocarbons including fluorocarbon- 1 1 and fluorocarbon-12 since 1977 and 1978, respectively. Methyl chloroform has been monitored at a number of locations since 1979. These data clearly show that the concentrations of the halocarbons have been increasing as a result of man-made emissions. Likewise, extended time records of measurements for a number of other trace species are being developed; most notable are the data for N2O, CH4, 03, and CO. Although the quality and extent of these data sets are variable and the inter- pretation is in some cases debatable, it is clear that tem- poral and areal variations are indeed occurring and that some changes can be attributed to natural processes and others to human activities. A number of surface time-series data sets are also being developed for airborne particulate matter. Many ofthese data sets show large variations that can be attrib- uted to large-scale transport phenomena. For example, the concentration of sulfate and other species in aerosols in the Arctic has been found to increase sharply with the arrival of air masses from Europe and the industrialized areas of Asia; these increases are also associated with the occurrence of widespread haze in the Arctic. More than 15 years of mineral aerosol measurements made over the western tropical North Atlantic show dear seasonal trends; the seasonal maximum concentrations are about 73 100 times greater then the seasonal minima. These min- eral aerosols are derived from sources in Africa some 5000 km distant. Long-term trends in concentration have been related to drought in Africa. Likewise, min- eral aerosol measurements in a network of seven SEAREX stations in the North Pacific show similar seasonal variations that are attributable to mineral dust transported from Asian sources that are over 8000 km away. In both the Asian and African dust studies, sharp day-to-day variations in dust can be related to specific synoptic events; in some cases, the dust can be traced to specific dust sources. Aerosol data such as these could be used to develop event-type meteorological transport models. NETWORK DESIGN In order to properly design a global network, it is necessary to have a fairly good idea of the temporal and areal variability of the concentration of the species in question. With a few exceptions, the data base for spe- cies of interest to the Global Tropospheric Chemistry Program is utterly inadequate. The situation is espe- cially desperate with regard to the question of vertical · . . ~ . c lstr1 button ot trace species. To further complicate matters, there is also a dearth of meteorological data over large areas of the globe. Espe- cially troubling is the sparsity of upper atmosphere data over most of the oceans and much of the continents. Thus some types of meteorological studies (especially event studies) that require the input of real meteorologi- cal data will be handicapped. The other problem is logistical. The earth is large, and field operations are difficult, time consuming, and expensive. It is unrealistic to expect to be able to dispatch simultaneously many large teams of highly trained sci- entists and technicians to many different parts of the globe and to keep them there for extended periods of time. Even if such resources were available, it is not at all clear what species these teams should measure, where they should measure them, and at what frequency. In- deed, many of these questions will be addressed by the individual research programs that are focused on spe- cific processes; consequently, this essential information will not be available for some years to come. Faced with these gaps in the knowledge of atmo- spheric chemistry and meteorology and because of the problems with logistics, we took a pragmatic approach. We surveyed the list of important species in the sections on cycles in Chapter 7 and also in Appendix C and asked the following question: Ofthe species that are relevant to our objectives, which are relatively easily measurable at ambient concentrations by using existing technology? It is difficult to accurately define the criterion of being

74 "relatively easily measurable. " To give some indication of what we mean, we shall cite a few examples. Current network operations in SEAREX in the North Pacific have shown that it is possible for untrained personnel to routinely collect filter samples that can be successfully analyzed for a number of species such as excess SO4-, NO3, mineral particles, sea-salt aerosol, CH4, sulfonic acid, Pub, and some trace elements such as vanadium. On the other hand, field experience in this same region has led the SEAREX investigators to conclude that it is impossible to use these same sampling procedures to make measurements of trace species such as lead, mer- cury, and chlorinated hydrocarbons the possibilities of . · , . . - . contamination during t. he sampling operation are so overwhelmingly great that the data would be unreliable and hence useless. Another example of a relatively easy sampling and analysis procedure is that for some of the halocarbons, which can be "grab sampled" in the field by using flasks, and subsequently analyzed in the laboratory by means of gas chromatographic procedures. Although this sampling procedure does require the use of specially constructed and treated flasks and some operator train- ing, the procedure has been used more or less routinely by a number of investigators. Another example is 03, which can be easily measured at ground level with off- the-shelf equipment. However, it is a much more di~- cult task to obtain vertical profiles of O3. Although ozonesondes are readily available, their use is not so straightforward, and trained technicians are necessary, although they need not be scientists. The types of measurements that we categorically ex- clude from our network protocol are those that require a highly trained scientist who has detailed knowledge of the chemistry of the species being measured and the idiosyncrasies of-the measurement technique-an ex- ample would be the current measurement of NO at concentrations typically found in remote areas. Finally, there is the type of measurement where the technique is still under development such as that for the hydroxyl radical. At this time, such measurements are completely out of the question for any sort of routine network field operation. In order to circumvent the problems of a limited data base and difficult logistics, we propose a program that calls for a dynamic plan of growth for the networks. We would start with a few stations sited in grossly different environmental regimes. These stations would serve sev- eral purposes. They would yield data that would serve as initial values for testing models; in turn, the output from the initial model tests would be used to further refine the sampling protocol and to provide guidance in the place- ment of new stations in the network. Also, the start-up stations would serve as operational environmental test PART II ASSESSMENTS OF CURRENT UNDERSTANDING sites for the instrumentation and measurement proto- col. Finally, the operation ofthese stations would serve to develop the logistical support base for the network. An assumption that is central to the proposed pro- gram is that selected stations in existing networks such as BAPMoN would evolve along with or as a part of the Global Tropospheric Chemistry Program global net- work. The most likely mechanism would be for cooper- ating nations to augment the operations at suitable exist- ing stations. An essential requirement would be that the protocol is identical to the Global Tropospheric Chemis- try Program protocol and that analytical techniques are completely validated for accuracy and precision. TEMPORAL CONSIDERATIONS The question of the frequency of sampling was not addressed in a very explicit manner in the description of the proposed program in Part I. The sampling fre- quency will be a function of a number of factors includ- ing the lifetime of the species in the atmosphere, the rate of change of the mean concentration of the species in the atmosphere (assuming that a concentration trend is be- ingmonitored), and the magnitude of the transient con- centration changes that are generated by source injec- tions and by transport or removal processes. As an example, consider N2O, which has a rather uniform global distribution (except for urban areas). The mean concentration of N2O is about 300 ppbv; concentrations at the surface in the northern hemisphere are about 0. 5 ppbv greater than those in the southern hemisphere. Moreover, the concentration of N2O appears to be in- creasing at a rate of about 0.2 percent per year. Clearly, it should not be necessary to measure N2O very often or at many different stations in order to characterize these trends. At the other extreme, aerosol particles have a residence time of 1 to 2 weeks. Thus the concentration of aerosol particles can change by orders of magnitude over the period of ~ day as the synoptic situation changes; for example, when a cloud of particles is ad- vected to a sampling site. Consequently, in order to characterize an aerosol event in context with the meteo- rology, a relatively high sampling frequency is required, ranging from 1 day to 1 week. It is because of their short residence time that aerosols can serve as good tracers for use in event models. To illustrate the difficulty in defining a sampling pro- tocol at the outset of a program, we cite the example of CO2, which is a relatively long-lived species. The con- centration of CO2 in the atmosphere is increasing stead- ily at the rate of about 0.3 percent per year because of human activities. However, the concentration of CO2 varies seasonally, the magnitude of the variation being geographically dependent. For example, at Manna Loa

CRITICAL PROCESSES the annual variation is about 2 percent whereas at the South Pole it is about 0.5 percent. The annual variations are primarily attributable to the impact of the biosphere. Because ofthese variations, it is necessary to sample at a relatively high frequency in order to adequately charac- terize the inherent variance of the data. In the current CO2 monitoring program, weekly CO2 samples are col- lected. Nonetheless, it is the variability of the concentration of the species that contains the information about the controlling processes. In the case of CO2, this variability can be traced to the seasonal activity and distribution of biological sources and sinks and to the controlling mete- orological and oceanographic processes. Indeed, the year-to-year fluctuations in the CO2 growth rate have shown that the oceanographic processes in equatorial regions of the Pacific Ocean play a major role in deter- mining the variability of atmospheric CO2 concentra- tions. The CO2 sampling program is an excellent exam- ple of how a carefully planned and executed monitoring . . . . program can provide great mslg. ats mto important proc- esses. The ultimate test of the knowledge of the CO2 system (or any other chemical system) is the capability to use the collected data to construct a model that will predict future trends. The way in which the CO2 program was imple- mented is also a good example of the general approach that we espouse. Stations were established at a relatively slow rate, and sampling and analytical techniques were thoroughly tested. Finally, modeling has played a large role in the CO2 program, and it has provided some important insights into the CO2 system. VERTICAL DISTRIBUTION MEASUREMENTS The most difficult problem in the network operations will be that of obtaining vertical concentration profiles. In many cases, the boundary layer concentrations will not be representative of those in the free troposphere- the major sources or sinks may reside in the boundary layer, and the chemistry there could be different from that in the free troposphere. Long-lived gases are thought to be well mixed in the vertical. However, the medium-lived gases may require vertical profile mea- surements. Shorter-lived gases such a 03, C2C14, and C2HC13 will almost certainly require measurements of vertical profiles. Also, vertical distribution data will be necessary for many of the aerosol species. Unfortunately, it is difficult and expensive to make vertical profile measurements on a routine basis. Other than the data obtained from a small number of ozone- sonde stations, there is remarkably little data in the liter- ature on vertical profiles. The most impressive and extensive data set on the distribution of a broad range of 75 important atmospheric chemistry species was that obtained in a series of flights made in 1977 as part ofthe Global Atmospheric Measurement Experiment on Aerosols and Gases (GAMETAG). A team of 39 scien- tists from nine different institutions measured a large number of species on flights over the western United States and Canada and over large areas of the North and South Pacific. Profiles were made up to 6-km altitude. GAMETAG was clearly a pioneering effort in the field of atmospheric chemistry. Aircraft operations of that type will be a mandatory part of the Global Tropo- spheric Chemistry Program as spelled out elsewhere in this report. However, it is completely unrealistic to ex- pect to carry out flights of the GAMETAG type in con- junction with routine network operation-indeed, it would be impossible to do so because there are not enough aircraft and scientific personnel to do the work. Instead, we propose a stepped approach to develop- ing the knowledge of vertical distributions. The vertical distributions in a limited number of source and sink environments must be measured. The objective is to characterize the concentration profiles and their tempo- ral and spatial variance and to relate these measure- ments to the sources (or sinks) and the local meteorology. Measurements of this type are already planned as pri- mary components of the other major programs in this work. The data obtained from these studies will enable an appropriate protocol to be specified for obtaining useful vertical profile data on a global scale. · . . · ~ . Another approach is to make a concurrent series of measurements above the boundary layer and at the sur- face using stations located on mountaintops and at ground level. Considerable care must be exercised in operating a sampling station on a mountain because the mountain generates its own meteorology, which, in turn, can affect the chemistry. Nonetheless, the program at the NOAA GMCC station on Mauna Loa, Hawaii, has shown that this problem can be circumvented. The station is located at an altitude of 3400 m, which is nominally above the marine boundary layer. However, it is clear that on many occasions, especially during the day, the station is definitely impacted by boundary layer air transported through the inversion by upslope winds. On the other hand, very extensive tests have shown that during downslope wind conditions the air at the station is derived from above the boundary layer. There exists an extensive set of measurements made concurrently at a mountaintop and the surface. Mineral and sea-salt aerosol concentrations have been measured for the past several years at Mauna Loa by a group from the University of Maryland. Independently, a SEAREX group has been making measurements of some of the same species at a coastal site in Oahu. These two records reveal some very interesting similarities and

76 differences, which, for the most part, appear to be inter- pretable on climatolog~cal and meteorological grounds. In another effort, a limited number of gases are being sampled concurrently at Mauna Loa and at Cape Ku- mukahi, Hawaii. These data suggest that a similar ap- proach of concurrent mountain/surface measurements might be fruitful for other species as well. It would seem logical that an extended set of such measurements be initiated at the Mauna Loa GMCC and at a suitable surface site in the Hawaiian Islands. Some other candi- date sites are available. A BAPMoN site is planned at the observatory on Tenerife in the Canary Islands. This station is at an altitude of 2370 m, which places it above the height of the nominal trade-wind inversion. There are other BAPMoN stations planned for mountain sites, some of which might be suitable for such studies. While paired mountain/surface measurements will provide useful free troposphere/boundary layer data, extensive vertical profile data will require the use of aircraft (except for O3 and water vapor, where sondes can be used). What sort of aircraft program could we reasonably expect to carry out a routine network operation? First, it will be necessary to use locally based lightweight com- mercial aircraft. Second, we assume that only slight modifications could be made to the aircraft. Third, flight times would have to be relatively short, on the order of an hour or two; this means that one could spend only a relatively short time at each sampling level. Be- cause of these limitations, it appears that only grab- sampling procedures would be possible for routine pro- f~ling work. Thus the protocol would be limited initially to the grab-sampling of gases and to those aerosol mea- surements that are feasible with low-volume samples. Thus the vertical distribution studies within the Global Tropospheric Chemistry Program would consist of four types of activities: 1. Intensive short-term field experiments associated with the major Global Tropospheric Chemistry pro- grams these would involve surface experiments and highly instrumented aircraft. 2. Ozone and water vapor sondes launched from a limited number of bases on a fairly frequent schedule several a week. 3. Mountain/surface stations carrying out the com- plete network protocol on a fairly frequent schedule (one or more samples a week). 4. Aircraft studies of grab-sample gases and parti- cles. A major task in developing the network protocol is to ascertain the optimum mix of these activities. A special effort must be made to relate the light aircraft work using PART II ASSESSMENTS OF CURRENT UNDERSTANDING grab samples to the ground-level program. We antici- pate that this is an appropriate area for exploratory research in the early phases of the program. It might be appropriate for such a study to be carried out in the Hawaiian Islands in conjunction with the routine opera tions at Mauna Loa and a corresponding operation set up at sea level. For completeness, an ozonesonde pro gram should be carried out concurrently with the a~r craft and ground station study. BIBLIOGRAPHY Bodhaine, B. A., B. G. Mendonca, I. M. Harris, and I. M. Miller (1981~. Seasonal variations in aerosols and atmospheric trans- mission at Mauna Loa. J. Geophys. Res. 86: 7395-7398. Broecker, W. S., and T.-H. Peng (1982~. Tracers in the Sea. Lamont- Doherty Geological Observatory, Columbia University, Pali- sades, New York, 630 pp. Cawse, P. A. (1982~. Inorganic particulate matter in the atmo- sphere, in Environmental Chemistry, Vol. 2, H. J. M. Bowen, ed. Royal Society of Chemistry, London, pp. 1-69. Committee on National Statistics (1977~. Environmental Monitoring. National Academy of Sciences, Washington, D.C., 181 pp. Davis, D. D. (1980~. Project GAMETAG: an overview. I. Geophys. Res. 85:7285-7292. Duce, R. A., et al. (1980~. Long-range atmospheric transport of soil dust from Asia to the tropical North Pacific: temporal varia- bility. Science 209: 1522- 1524. Geophysics Study Committee (1977). Energy and Climate, Studies in Geophysics. National Research Council. National Academy of Sciences, Washington, D.C., 158 pp. Mahlman, I. D., and W. J. Moxim (1978~. Tracer simulations using a global general circulation model: results from a midlati- tude instantaneous source experiment. I. Atmos. Sci. 35:1340- 1374. Marland, G., and R. M. Rotty (1979~. Carbon dioxide and cli- mate. Rev. Geophys. Space Phys. 17:1813-1824. Newell, R. E. (1963~. The general circulation of the atmosphere and its effects on the movement of trace substances. /. Geophys. Res. 68:3949-3962. Oltmans, S. I. (1981~. Surface ozone measurements in clean air. l. Geophys. Res. 86:1174-1180. Parrington, I. R., W. H. Zoller, and N. K. Aras (1983~. Asian dust: seasonal transport to the Hawaiian Islands. Science220:195-197. Pinto,~J. P., Y. L. Yung, D. Rind, G. L. Russell, J. A. Lerner, J. E. Hansen, and S. Hameed (1983~. A general circulation model study of atmospheric carbon monoxide. J. Geophys. Res. 88:3691-3702. Prinn, R. G., P. G. Simmonds, R. A. Rasmussen, R. D. Rosen, F. N. Alyea, C. A. Cardelino, A. J. Crawford, D. M. Cunnold, P. I. Fraser, and I. E. Lovelock (1983~. The atmospheric lifetime experiment, I: introduction, instrumentation, and overview. J. Geophys. Res. 88:8353-8368. Prospero, i. M., and R. T. Nees (1977~. Dust concentration in the atmosphere of the equatorial North Atlantic: possible relation- ship to the Sahelian drought. Science l 96: 1196-1198. Rahn, K. A. (1981~. Relative importance of North America and Eurasia as sources of Arctic aerosol. Atmos. Environ. 15:1447- 1456.

CRITICAL PROCESSES Turco, R. P., R. C. Whitten, andO. B. Toon(1982~. Stratospheric aerosols: observation and theory. Rev. Geophys. Space Phys. 20:233-279. Uematsu, M., R. A. Duce, J. M. Prospero, L. Chen, J. T. Merrill, and R. L. McDonald (1983~. Transport of mineral aerosol from Asia over the North Pacific Ocean. I. Geophys. Res. 88:5343- 5352. 77 Wallen, C.-C. (1980~. Monitoring potential agents of climatic change. Ambio 9:222-228. Whitten, R. C. (ed.) (1982~. TheStratosphericAerosolLayer. Springer- Verlag, New York, 152 pp. World Meteorological Organization, Environmental Pollution Monitoring Program (1981~. Summary Report on the Status of the WMO Background Air Pollution Monitoring Network as of April, 1981.

78 HOMOGENEOUS AND HETEROGENEOUS TRANSFORMATIONS BY D. DAVIS, H. NIKI, V. MOHNEN, AND S. LIU As discussed in the section above by Cicerone et al., a variety of biological and geological processes results in the emission of trace gases into the troposphere. For the mostpart, the key elements(e.g., carbon, nitrogen, and sulfur) making up these trace gases are in reduced oxida- tion states. By contrast, when these elements are re- turned to the earth's surface, via precipitation and/or dry deposition, they most frequently are found in their thermodynamically stable oxidized forms. The atmo- spheric transformations that lead to the chemical oxida- tion of trace gases are complex and encompass homoge- neous gas-phase, homogeneous aqueous-phase, as well as heterogeneous processes. Outlined in the text that follows are several of the more prominent features of each of these transformation types. The authors have drawn special attention to those outstanding questions that will require futher study in the near future. HOMOGENEOUS GAS-PHASE CHEMISTRY Cyclic Photochemical Transformations: HxO' Most gas-phase oxidation processes are either directly or indirectly initiated as a result of the atmospheric ab- sorption of ultraviolet solar radiation. One of the more important oxidizing agents formed from this absorption process is now believed to be the hydroxyl (OH) radical. Its pivotal role is illustrated in Figure 5.4. It is seen that OH-initiated reactions provide the major pathway for transforming a large number of tropospheric com- pounds into their oxidized forms. The OH species there- fore plays a major role in controlling the chemical life- times ofthese "reduced" compounds. The primary production of OH is initiated by the photolysis of O3. Solar photons having wavelengths be- tween 315 and 1200 nm dissociate O3 to produce an oxygen atom in its ground electronic state: O3 + hv(1200 > ~ >315nm) ~ 0(3P) + O2. (5.1) The 0(3P) atom combines rapidly with O2 in a three- body reaction to reform O3: 0(3P) + O2 + M ~ O3 + M (M = N2, 02)- (5 2) Thus the sequence of reactions (5. I) and (5.2) is a null cycle with no net chemical effect. On the other hand, when O3 absorbs a photon in the near-ultraviolet, with a wavelength shorter than 315 nm, an electronically ex- cited oxygen atom is produced. O3 + hv (X < 315 nm) ~ O(~D) + O2 (5 3) The O(iD) ~ 0(3P) transition is forbidden and there- fore O(~D) has a relatively long radiative lifetime (t = 110 s). In the troposphere, rather than relaxing ra- diatively, O(~D) most often collides with N2 or 02, i.e., reaction (5.4), and ultimately leads to the regeneration of O3 via reaction (5. 2) and another null cycle. O(iD)+M~O(3P)+M(M=N2,O2). (5.4) Occasionally, however, O(iD) collides with H2O and causes the generation of two OH radicals: O(~D) + H2O ~ 2HO. (5.5) In anthropogenically unperturbed regions of the troposphere, OH reacts overwhelmingly with CO and CH4; i.e., OH + CO ~ CO2 + H (5.6) OH + CH4 ~ CH3 + H2O. (5 7) The hydrogen atom and CH3 radical rapidly com- bine with O2 to form HO2 and CH3O2 radicals, respec- tively. Even so, reactions (5.6) and (5.7) do not necessar- ily lead to removal of OH from the atmosphere, since both HO2 and CH3O2 radicals can be partially con- verted back to OH via a complex series of chain reac- tions. This chemistry is illustrated in Figure 5.5. In this reaction scheme, the HO2 radical formed by the reac- tion of HO and CO, regenerates OH via and HO2 + NO ~ NO2 + OH (5.8) HO2 + O3 ~ 2O2 + OH. (5.9) Alternatively, the HO2 species can result in radical chain . . . term~nat~on v~a HO2 + OH ~ H2O + O2 (5.10) HO2 + HO2 ~ H202 + O2 (5.11) However, a small fraction of the H2O2 formed in reac- tion (5.11) may be photolyzed to again regenerate two OH radicals, thereby serving as a temporary reservoir for OH. In addition to reactions (5. 10) and (5. 1 1), both OH and HO2 may be removed by combination reac- tions with NO2, i.e., HO + NO2 + M ~ HNO3 + M (5.12) HO + NO2 + M ~ HONO3 + M. (5.13)

CRITICAL PROCESSES H2SO4 Mult steps SOW so2 Multisteps HSO3 HS, H2O ~ H2S/ CO \ ~CC13 NH3\ co2 NH2, H2O Multisteps NO, H2O Depending on the temperature, the HONO3 species can be relatively unstable, dissociating back to HO2 and NO2. Its lifetime at 300°K is 10 s, whereas at 250°K it is 104 s. At high altitude, reaction (5.13) may therefore provide a significant chain termination step. The HNO3 formed in reaction (5.12) is very efficiently converted back to OH radicals via photolysis and typically is re- moved by wet or dry deposition processes. The chemistry of the CH3O2 radical resulting from the OH-CH4 reaction is quite complex, with the rate coefficients for several of the elementary reactions in- volved in this chemistry remaining unmeasured. One possible atmospheric degradation scheme for this spe- cies is that shown in Figure 5.6. As in the chemistry of HO2 species, CH3O2 can react with either NO or HO2, depending upon the tNO]/tHO2] ratio: CH3O2 + NO ~ CH3O + NO2 (5.14) CH3O2 + HO2 ~ CH3OOH + O2. (5.15) Simple kinetic considerations indicate that reaction (5.14) becomes equal to reaction (5.15) for an [NO]/ tHO2] ratio of ~1.0 at 300°K (the uncertainty here is at least a factor of 3). As shown in Figure 5. 6, the eventual Mule~steps CxHy-~'H2O \ 79 CO, H2O Multisteps CO HNO~ ~5 3 XO, O2 HCi, H2O FIGURE 5.4 The central role of OH in the oxidation of tropospheric trace gases. fate of the CH3O appears to be the formation of CO via the intermediate product, formaldehyde (CH2O): CH3O + O2 ~ CH2O + HO2 (5.16) (5.17) (5. 18) CH2O + hv (or OH) ~ CHO + H CHO + O2 ~ CO + HO2. From this sequence of reactions, it is evident that the complete oxidation of the CH3OO radical potentially can provide an additional source of HO2 radicals to the HxO' system. However, in the presence of low NO lev- els, reaction (5.15) dominates, and CH3OOH is formed. This species may be removed from the gas phase by heterogeneous processes or may undergo vari- ous homogeneous gas-phase reactions. At present, the relative importance of the various reaction pathways for CH3OOH is unknown. Among the minor products that may be formed from further CH3OOH chemistry are methanol (CH3OH) and formic acid (HCOOH). Complicating the HxOy fast-photochemical cycle still further is the possibility that other OH-initiated reac- tions also generate free radical species, some of which could feed back into the main HxOy cycle. One of the

80 o3 |hp o(1 D) H2O l he \tor HONO2 HO2NO2 Dry Removal Wet or Wet or Dry Removal Dry Removal FIGURE 5.5 OH/HO2 radical chain reactions. most likely classes of compounds that might fit this role are the nonmethane hydrocarbons, NMHC (CxH', where x andy > 1~. In general, the atmospheric chemis- try of the NMHCs is far more complex than that of CH4. And, despite some recent progress, mechanisms describing this chemistry remain highly uncertain. co NO co2 O2 ~ ~ 1 | HCOOH | 1 O2CH2OH 1 ~ 102 | CH2 OH | CH 3( )H HO2 OH _ :_ N r r OH FIGURE 5.6 A possible tropospheric degradation scheme for CH3O2 radicals, formed from CH4. The current lack of under- standing of this chemistry defines one of the major uncertainties in the understanding of fast HxOy photochemistry. PART II ASSESSMENTS OF CURRENT UNDERSTANDING Two potentially important types of NMHCs in the "remote" troposphere are isoprene (COHN) and ter- penes. Emitted by deciduous and coniferous trees, re- spectively, these compounds react rapidly with OH radi- cals. Their reactivity toward OH as well as O3 suggests that they should be short-lived in the atmosphere; there- fore, their concentrations will likely decrease rapidly as one moves away from their source regions. As noted previously, whether the oxidation of these compounds serves as a net sink or source for HxO' species can only be speculated on at the present time. Illustrating the poten- tial complexity ofthis problem is the atmospheric degra- dation scheme for the compound CsHs, Figure 5.7. Two of the by-products from this chemistry that might affect the HxOy cycle are CO and O3. For this impact to be realized, however, would require that significant fluxes of C5Hs be present such as those that might originate from a tropical rain forest. Although it is perhaps understandable that the com- plex mechanisms involving NMHCs are not yet known, CH3 isoprene CH2 = C-CH = CH2 1 OH, O2, NO OH CH 1 13 CH2-C-CH = CH2 1 O. ':\ unimolecular '/ ~,o, 1 l ,CH4 ~ C: ~ _ O2 CH2OH I CH3C-CH = CH2 1 /2 ~ OH, O2, NO CH2O 1 + H2O 1i \ NO, O O. OH I! 1 1 CH3C-CH - CH 1 2 \02 | unimolecular O2 \ /< CH OH ~ ~ O O 2 CH3C-CH hv OH, O2, NO OH HO2 hu ~ 1 ~O CH3 + ~'' [~ 1°2 1 ~ _ ~ CO + HO2 FIGURE 5.7 A possible reaction scheme for isoprene oxidation in the presence of NOX.

CRITICAL PROCESSES the point should be stressed that even the simplest HxOy cyclic scheme has not yet been fully tested. Thus the central role of OH in atmospheric chemistry, even though strongly supported by laboratory kinetic mea- surements, has not been convincingly demonstrated. This state of affairs reflects the current absence of signifi- cant field data on many of the critical species involved in HxO' fast photochemistry. The absence of field data, in turn, has reflected an absence of available measurement technology for many of the critical species involved in this cycle. Cyclic Photochemical Transformations: NxO' Like the HxOy system, there are several NxOy species (e.g., NO, NO2, NO3, MONO, N2O5, and HOONO2) that are believed to be present in the atmosphere at concentration levels corresponding to those predicted by photo-stationary-state chemistry. As previously noted, this NxO' cyclic system also has chemical coupling with the HxO' system. Figure 5.8 summarizes several of the key aspects of this cycle. It is seen that during daylight hours NO and O3 are constantly produced from the photolysis of NO2: NO2 +hv(285 Chad 375nm)~ NO + O(3P) (5.19) followed byreaction(5.2~. NO and O3, inturn, continu- ously react to regenerate NO2 via reaction (5.20~: NO + O3 ' NO2 + 02, (5.20) OH HNO2 ~ 3 wo/ro l ~ o3 ho o ~ ~ o, _ 0 `3 pi HO ha, NO 3 ___ ~ I__- HO2NO2 o3 ~ ~ O3 L 3 _ J t 1( )2 ,_ _ NO2 i__ 2_5 _i ~ wo/ro FIGURE 5.8 Major atmospheric reactions of NxOy species: Solid line boxes indicate major daytime nitrogen species; broken line boxes indicate significant nighttime nitrogen species. It is still uncertain whether PAN is a major species in the free troposphere. The notation wo/ro denotes washout/rainout process. 81 thus forming a null cycle in which an ultraviolet photon is converted into heat. In addition to reaction (5.20), both reactions (5.8) and (5.14) may also make signifi- cant contributions to the NO to NO2 conversion proc- ess, provided high-solar-flux conditions are prevalent. The NO2 formed from reactions (5.20), (5.8), and (5.14) may be photolyzed to regenerate NO via reaction (5.19) or, alternatively, react with OH or HO2 (Reac- tions (5.12) and (5.13~) to form nitric acid species. Both of the latter reactions can lead to the removal of NxO from the atmosphere via wet or dry deposition. Yet an- other reaction possibility involving NO2 is reaction with o3 NO2 + O3 ~ NO3 + O2 (5.21) The NO3 formed in reaction (5.21) may undergo pho- tolysis (regenerating NO2) or react with additional NO2 (reaction (5.22~) and generate still another new NxO species, N2O5. NO2 + NO3 + M ~ N2O5. (5.22) Although N2O5 may thermally decompose, it is now believed that some fraction of it reacts heterogeneously with H2O, forming HNO3. During daylight hours, the dominant loss process for NO3 is photolysis. Under nighttime conditions, reac- tions (5.21) and (5.22) become dominant. This night- time chemistry thus predicts that NO3 should be one of the major NxOy species in the atmosphere. Limited field observations of this species, however, indicate the levels to be far lower than those predicted from the chemistry shown in Figure 5.8. Possible NO3 scavengers remain unidentified at this time. A final uncertainty in the understanding of NxO' chemistry involves the coupling of this chemistry with complex hydrocarbon species known to react with NO2 to form peroxyacetyl nitrate (PAN), a well-known air pollutant. CH3C(=O)OO- + NO2 ~ CH3C(=O)OONO2. (5.23) PAN can thermally decompose with a lifetime of ap- proximately 45 min at 300°K; but it has a lifetime of 200 days at 250°K. There is now some evidence, in fact, that suggests that this species could define a major reservoir for NOX in the "remote" free troposphere. Like the HxOy fast-photochemical cycle, there is now a great abundance of laboratory kinetic data that suggest that the fast NxOy cycle is one of the key photochemical cycles operating in the troposphere. Also like the HxOy cycle, no quantitative field test has yet been performed that has demonstrated this fact. Such tests need to vali- date basic relationships such as that shown in equation (5.24~:

82 [NO] J23 = [NO2] k24 [03] + ks [HO2] + kit [CH3O2] (5.24) The absence of these tests to date again reflects the ab- sence of appropriate field measurement technology in years past. Ozone liansformations/Photochemical Sources and Sinks As shown in Figure 5.9, the fast-photochemical cou- pling between NO, NO2, and O3 produces a null cycle in which an ultraviolet photon is converted into low- grade heat. When, however, NO is converted to NO2 without the use of an O3 molecule, there can be net production of O3. The latter chemistry occurs when reaction (5.8) and/or reaction (5. 14) becomes dominant over processes (5.11) and (5.15), respectively. These chemical conditions are found to be quite common in large urban population centers throughout the United States. As such, there are frequently times when the levels of photochemically generated O3 far exceed those found in the natural environment. Evaluating whether significant O3 production occurs in the "remote" troposphere is a far more difficult task. As shown in Figure 5.9, the primary formation pathway for OH requires the consumption of one O3 molecule. Whether this ultimately represents a net destruction pathway for O3 depends on the subsequent chemistry of the HO2 and CH3O2 radicals. Recall that these species NO + O2 NO3 NO (Net Or Loss) '~ (Net O3 Gain) \3 /~ OH, O2, O2 \// HO2,/~ H O2, O2 ( Pr°bable Net O3 Loss) 4~: /// ~ O2 ( ~ ~ ) I, O2 (Net O3 Loss) ~O3 1 1 o(1 D) 1 / O23E - 2 OH (Probable Net O3 Loss) FIGURE 5.9 Photochemical production and destruction of O3. Key oxygen species are shown in solid boxes. PART II ASSESSMENTS OF CURRENT UNDERSTANDING are formed as a by-product from the reaction of OH with CH4 and CO. Thus, if HO2 and CH3O2 predominantly react to form H2O2 and CH3OOH (reactions (5 . 1 1 ) and (5.15~), the formation of OH from O3 defines an O3 sink. Alternatively, the HO2 species can react with O3 itself, and this again would define a significant O3 photo- chemical sink. Only when appreciable NO is present can reactions (5.8) and (5.14) (involving HO2 and CH302) dominate (5. 1 ~ ~ and (5. 15), leading to a net photochemical source of O3. Thus, whether the coupled HXo'/NxOy cycles define a net sink or source for O3 in the "remote" tropo- sphere stands as one of the major unanswered scientific questions in global tropospheric chemistry. Noncyclic Transformations Earlier, it was noted that the OH radical appears to be the principal species responsible for initiating the oxida- tion of numerous "reduced" trace gas species. Follow- ing the initiating step, most of the reaction products move systematically through one or multiple steps to a final oxidized product. This final product is then re- moved from the troposphere by wet or dry deposition. Illustrative of this reaction sequence is the degradation of complex NMHCs. The initial products resulting from the attack of OH or O3 on the parent hydrocarbon compound are aldehydes, ketones, and/or organic ac- ids. Still further chemistry involving these species may result in the final production of CO and CO2. Alterna- tively, some ofthe larger organic radicals generated may undergo combination reactions or chain-polymeriza- tion-type reactions. Both of the latter type of reactions may result in the formation of organic aerosols. At present, the branching ratio for a given hydrocarbon undergoing complete degradation to form CO2 versus its forming organic aerosols is still poorly understood. Ammonia in the troposphere can react with OH to form NH2 radicals. These radicals, in turn, may react with O2 to form NH2OO. The chemistry of this species is unknown. Possible final reaction products may in- volve N2 or NO. A very significant class of compounds for which non- cyclic transformations are important are those that con- tain sulfur. For virtually all of these compounds (the one possible exception being COS), the initiating step in- volves reaction with OH. In the case of hydrogenated sulfur, the subsequent reactions of the initially formed sulfur radical species involve several possible reactants: O2 ~ O3 ~ NO, HO2, or RO2. In each case, this appears to lead to the formation of SO2. The SO2 species, in addi- tion to being removed from the gas phase by heteroge- neous processes, may react with the OH to form the intermediate HOSO2 radical. The fate of the HOSO2

CRITICAL PROCESSES radical is believed to be reaction with O2 to form a peroxy radical or, alternatively, SO3 and HO2, i.e., o H-O-S + O o o S = 0 + HO2 o The peroxy species may react with NO or HO2 or per- haps hydrate as a result of collisions with H2O. The only certainty in this chemistry now appears to be that, like SO3, the final product is some form of sulfuric acid. The latter species is rapidly removed from the gas phase by various heterogeneous processes. In each of the above systems (e.g., NMHC, NH3, and the sulfur compounds), the mechanism following the initiating step leading to the formation of a final oxidized product is unknown. The absence of this ki- netic information reflects, to a large degree, the absence of adequate methodologies to study the kinetics of poly- atomic free radical species. New kinetic information will be essential to achieving an acceptable level of under- standing of these important oxidative atmospheric path ways. HOMOGENEOUS AQUEOUS-PHASE TRANSFORMATIONS The aqueous phase is most frequently considered in the context of physical removal processes. Like the gas phase, however, this medium also encompasses exten- sive chemical transformations And, like the gas phase, these chemical transformations are oxidative in their chemical nature and involve some of the same reactive agents i.e., OH, HO2, and 03, although the aqueous phase is far more complex in its chemistry than the gas phase. Not only are there a large number of single-step elementary-type reactions to contend with, but there are also numerous fast equilibria. Furthermore, this chem- istry involves the reactions of neutral free radicals, free radical ions, On-free-radical ions, and nonradical, non- ionic, reactive species such as H202 and O3. Making this chemistry still more complex is the fact that the aqueous phase is distributed in the atmosphere in the form of a broad spectrum of aqueous aerosols. The two most general classes of liquid aerosols may be identified as (1) those found in clouds or fogs, and (2) those present under clear air conditions. In the first case, the most important size range is 2 to 80 ~m, although rain droplets up to a few millimeters can be found. The second category encompasses particles ranging from the size of critical clusters (10 angstroms) up to a few mi- crometers. The number density of aqueous aerosols as a function of size is also highly variable, being critically 83 influenced by exact environmental conditions. How the chemistry of aqueous aerosols differs as a function of size is currently one of many poorly understood characteris- tics of these species. Traditionally, the approach taken in unraveling this chemistry has involved studies of closed chemical reac- tor systems in which only two or perhaps three major chemical species are added to solution reactors. (In many respects, these studies have their analogue in gas- phase smog chamber investigations.) This has been par- ticularly true of studies designed to elucidate the oxida- tion pathways of SO2 and nitrogen oxides. In one of the most extensively investigated systems, involving aque- ous SO2 mixtures, added oxidizing agents have in- cluded O2-saturated solutions with and without added metal ion catalysts, O3-saturated solutions, and H2O2 solutions. The qualitative as well as semiquantitative data generated from these investigations have shown that each reaction system could potentially be important in the aqueous-phase oxidation of S(IV) to S(VI). All show some pH dependence, but of these the O3 system appears to have a particularly high sensitivity to changes in pH level (see Figure 5.10~. For the most part, mechanistic details on the S(IV) to ~o-6 1010 ~ 10~1 ~ 4 4.5 1 1 1 5 5.5 6 pH 7~ O2 / Without / Catalysts / - 6.5 7 7.5 FIGURE 5.10 Conversion of S(IV) to S(VI). The pH depend- ence of the reaction rate is for the systems H2O2, 03, and O2.

84 S(VI) aqueous chemistry have been lacking. Thus it is unclear how the bulk chemical conversion rates mea- sured for the 03, O2/metal ion catalyst, and H2O2 oxi- dizing systems may be combined for the case of a real aqueous aerosol environment. There is growing evi- dence, in fact, that many ofthe same reactive intermedi- ates (especially free radicals) exist in all three systems. More recently, an integrated modeling approach has been attempted on these aqueous-phase systems. In this approach, as in modeling studies of homogeneous gas- phase chemistry, the entire reaction system is built up from a large number of elementary reactions. Sulfur, nitrogen, and/or carbon chemistries are taken to occur simultaneously and continuously in time. Illustrative of the latter approach is the chemical scheme shown in Figure 5.11. This chemical scheme portrays aqueous- phase chemistry as being made up of both numerous very fast equilibria and individual rate controlling ele- mentary reactions. Chemical intermediates include both ionic and free-radical-type species. HxO' oxidizing agents in the system may result from aerosol scavenging of H202, 03, OH, and HO2 or by the in situ liquid- phase photolysis of the species H2O2 and O3. The aque- ous aerosol model may also be expanded to include the chemistries of NO2 and NO3 as well as soluble carbon species such as formaldehyde (CH2O). The use of building block elementary reactions to construct the complex chemistry of aqueous aerosols now appears to offer considerable potential. Facilitating this approach is the availability of a large volume of rate _~ ~ ~ ~ i ~ Cl FIGURE 5.11 Primary chemical pathways for a cloud droplet containing reduced sulfur, carbonate, and C1- ion, and a source of reactive HxOy The symbol ~ indicates fast equilibria, an ~ either an elementary reaction or multistep fast aqueous-phase process, and the dotted enclosures indicate various types of micro- chemistries taking place within the larger overall aqueous system. PART II ASSESSMENTS OF CURRENT UNDERSTANDING coefficients for elementary solution reactions. Most of these have been generated over the past 15 years by radiation chemists using, in particular, pulse radiolysis techniques. Even so, there remain numerous reactions of possible importance to this chemistry that still are without rate constants. Others, which have been mea- sured, need to be reexamined with more advanced ki- netic tools, especially with regard to establishing their temperature dependence. In all cases, the question may be raised: Are rate coefficients measured in bulk liquid phase applicable to the broad spectrum of aqueous aerosols in the environ- ment? Certainly, there would appear to be a need to investigate this chemistry under conditions where indi- vidual aerosol species could be studied as a function of time. Such studies will challenge the best technology, but must be viewed as a critical step in advancing the under- standing of this science. - HETEROGENEOUS PROCESSES Normally, a heterogeneous reaction implies one oc- curring at an interface between two phases, e.g., gas- liquid, gas-solid, or liquid-solid. Several interfaces may be involved in an overall process. Of course, solid-solid and liquid-liquid interfaces are also possible, but are not likely to be of great importance in atmospheric chemis- try. The study of interfaces, with their interesting and significant physical and chemical processes, is an old discipline that has recently reawakened. To appreciate the role of interfaces in many phenomena, one need only recognize that, in any multiphase system, communica- tion between the bulk phases occurs through the surfaces that connect them. Even when these surfaces make up a small fraction of the total volume as is usually the case for particles suspended in ambient air- they may have a dominant effect. A classical example can be seen in the phase transition of supercooled water droplets to frozen droplets (contact freezing) or ice crystals (sublimation freezing). This fundamental precipitation-producing process (Bergeron process) is believed to be induced and controlled to a significant extent by particles with very specific surface characteristics (ice nuclei). As in all chemistry and solid-state physics, measure- ments at the atomic and molecular level lie at the heart of a satisfying description of surface structure and compo- sition. Processes that occur at surfaces are described in terms of the time evolution of reactant, product, and intermediate structures. Without definition of surface structures in terms of equilibrium bond lengths and bond angles, as well as the potential energy functions that describe their variations, an adequate description of reactions at the molecular level is impossible. Because

CRITICAL PROCESSES even the simplest heterogeneous reactions are very com- plex at the molecular level, understanding at this level requires the application of many complementary exper- imental and theoretical tools. As a result, attempts to resolve heterogeneous processes of importance to atmo- spheric chemistry are still quite rudimentary. An at- tempt to present a comprehensive picture of heteroge- neous atmospheric chemistry can be found in various workshop documents listed in the bibliography. Most reactions that occur on such surfaces are thought to be noncatalytic. These include chemical re- actions in which both phases participate as consumable reactants, or physical processes involving either trans- port or growth, or both. The process of absorption or adsorption is a heterogeneous process. Heterogeneous catalytic processes normally imply the "conserved" par- ticipation of the interface material or a species adsorbed on it. Heterogeneous catalysis requires, among other things, the demonstration of "turnover" numbers 'far in excess of unity. The turnover number is essentially the number of repeated reactions conducted per unit time at a catalytic site. Reactions can be heterogeneous overall but locally homogeneous, as represented by reaction within the bulb of an aerosol particle where reactants are trans- ported in from the gas phase. These reactions might better be termed multiphase rather than heterogeneous because the reactants react in one phase, although some originate from another phase. One special class of heterogeneous reactions is that Direct Emission or Produced by Gas BOX 1 Water Vapor, Atmospheric Trace Gases (NH3, NOx, H2S, SO2'03' Unsaturated Hydrocarbons) Depletion by Gas-Phase Reactions and Other Sinks 85 referred to as gas-to-particle conversion. These reac- tions cause the transfer of a chemical species from the gas phase to an aerosol or liquid droplet suspended in the atmosphere, or they may cause formation of new parti- cles. Figure 5.12 shows a box diagram for gas-to-particle conversion processes including the following: 1. Homogeneous, homomolecular nucleation (the formation of a new stable liquid or solid ultrafine particle from a gas involving one gaseous species only); 2. Homogeneous, heteromolecular nucleation (for- mation of a new particle involving two or more gaseous species); 3. Heterogeneous heteromolecular condensation (growth of preexisting particles due to deposition of mol- ecules from the gas phase). The coexistence of homogeneous and heterogeneous reaction paths governing the distribution of key chemi- cal species is shown in Figure 5 .13 . Of the many gas-to-particle conversion processes be- lieved to occur in the atmosphere, such as those depicted in Figure 5. 13, one of the most interesting is the conver- sion of gas-phase SO2 to sulfate. Because this process generates two hydrogen ions, it often is responsible for producing acid rain in regions containing high levels of so2. Heterogeneous reactions may also be of importance in aqueous-phase atmospheric chemistry. The potential for transition metals, commonly found in atmospheric Radiation and Other Energy Input Constant or Time Dependent Rate of Production Heterogeneous Processes Coagulation Sorption Relative Humidity Constant or Vary- ing with Time Relative Humidity Constant or Vary- ing with Time 1 ~. Heteromolecular BOX 2 Heteromolecular BOX 3 . . Condensation Products of Low Volatility Nucleation Rate . BOX 4 (Formed by Gas- of Production Embryon~c Embryonic Phase Chemical ~Critical Size Brownian Droplets of Reactions, Hydrated Gas-to-Particle Coagulation LargerSize Species Included) Formation) Mixing and Coagu ration BOX 5 Preexisting and Newly Produced Aerosol, Liquid or Solid (Size Variation Possible Due to Fluctuations of Relative Humidity, Coagulation, Deposition of Trace Gas Molecules, etc.) FIGURE 5.12 Box diagram for gas-to-particle conversion. The boxes contain the substance, and the arrows describe the process.

86 FIGURE 5.13 Gas-phase constituents and major reaction pathways (solid lines). Interactions between chemical families are indicated by dashed lines. Heavy (double) arrows show key heterogeneous pathways involving aerosols (A) and precipitation (P) (Turco et al., 19824. aerosols and cloud droplets, to function as heteroge- neous catalysts is relatively well known. Based on their abundance, chemical form, stable oxidation states, and bonding properties, one can speculate that iron, manga- nese, and perhaps copper are most likely to function in this way (especially in the case of rural or urban atmo- spheric environments). For an atmospheric reaction heterogeneously catalyzed by a transition metal, the rate-determining step is likely to be either reaction at the catalyst surface or permeation of the reactant through the organic film that is sometimes observed on aqueous atmospheric aerosols. Heterogeneous catalysis involv- ing transition metals is likely to be of little consequence for species with rapid gas-phase or homogeneous liquid- phase reaction pathways, but may be significant for slower processes such as the aqueous-phase oxidation of so2. Soot-catalyzed SO2 oxidation may be another impor- tant mechanism for sulfate formation in the atmo- sphere. Soot is synonymous with primary carbonaceous particulate material. It appears that this material is present not only in urban atmospheres, but also in re- mote regions such as the Arctic. It is a chemically com- plex material consisting of an organic component and a component variously referred to as elemental, graphitic, or black carbon. Soot has properties similar to those of activated carbon, which is well known to be a catalytic surface active material. The above discussion makes it quite apparent that the inclusion of heterogeneous processes is essential to PART II ASSESSMENTS OF CURRENT UNDERSTANDING A HO2 ~ HN024 p NO ~ ~a' l i ~ H NO3 ~ p, A H2O2-OHMS 1~ __ NO2 ~ HO2 NO2 A, P I ~ ~-` ----_ A' ,>~` \ NO3 ~ ~ N2O5 ! \ O3 ~-O ~ / I '. ~ L _'' A, P \~ I'---------______ C H4 CH3 O2 NO2 _ _- --- CH3:CH3 0 ~ CH2O ~ P ~ CIO P. A (CH3 )2S S_SO-- _~SO2 ' HSO3 C1CS CS HIS ~SO3 CS2 H2S H2SO4 at, A, P achieving a complete understanding of the tropospheric cycles of sulfur, nitrogen, chlorine, carbon, and so on. However, because ofthe complexity ofthis chemistry, its quantification in existing models has not yet been satis factorily accomplished. Thus both extensive laboratory and extensive field studies are needed. BIBLIOGRAPHY Homogeneous Gas-Phase Transformations Atkinson, R., K. R. Darnall, A. C. Lloyd, A. M. Winer, and i. N. Pitts, Jr. (1979~. Kinetics and mechanism of the reaction of the hydroxyl radical with organic compounds in the gas phase. Adv. Photochem. 11 :375-488. Chameides, W., and J. Walker (1973~. A photochemical theory of tropospheric ozone. J. Geophys. Res. 78:8751. Crutzen, P. J. (1983~. Atmospheric interactions homogeneous gas reactions of C, N. and S containing compounds. Chapter 3 in The Major Biogeochemical Cycles and Their Interactions. SCOPE 2 1, B. Bolin and R. Cook, eds. Wiley, New York, pp. 67-112. Demerjian, K. L., and J. G. Calvert (1974~. The mechanism of photochemical smog formation. Adv. Environ. Sci. Technol. 4:1- 262. Fishman, i., S. Solomon, and P. Crutzen (1980~. Observational and theoretical evidence in support of a significant in situ photo- chemical source of tropospheric ozone. Tellus31: 432. Levy, II, H. (1974~. Photochemistry ofthe troposphere. Adv. Pholo- chem. 9:5325-5332. Liu, S. C., D. Kley, M. McFarland, J. D. Mahlman, and H. Levy, II (1980~. On the origin of tropospheric ozone._. Geophys. Res. 85: 7546-7552. Logan, J. A., M. J. Prather, S. C. Wofsy, and M. B. McElroy

CRITICAL PROCESSES (1981~. Tropospheric chemistry: a global perspective. I. Geophys. Res. 86:7210-7254. Seller, W. (19741. The cycle of atmospheric CO. Tellus 26: 1 16. Seinfeld, I. H. (Chairman) (1981~. Report on the NASA Working Group on Tropospheric Program Planning. NASA Reference Publica- tion 1062. Wofsy, S. C. `1976~. Interactions of CH4 and CO in earth's atmo- sphere. Annul Rev. Earth Planet. Sci. 4:441-469. Homogeneous Aqueous-Phase Transformations Chameides, W., and D. D. Davis `1982~. The free radical chemis- try of cloud droplets and its impact upon the composition of rain. I. Geophys. Res. 87 4863-4877. Farhataziz, and A. B. Ross `1977~. Selected specific rates of reac- tions oftransients from water in aqueous solution. III, Hydroxyl radical and perhydroxyl radical and their radical ions. NSRDS NBS 59, special publication. Department of Com- merce, National Bureau of Standards. Graedel, T. E., and C. I. Weschler (1981~. Chemistry within aque- ous atmospheric aerosols and raindrops. Rev. Geophys. Space Phys. 19:505-539. Heiko, B. G., A. L. Lazrus, G. L. Kok, S. M. Kunen, B. W. Grandrud, S. N. Gitlin, and P. D. Sperry (1982~. Evidence for aqueous phase hydrogen peroxide synthesis in the troposphere. J. Geophys. Res. 87:3045-3051. Junge, C. E., and T. A. Ryan (1958~. Study ofthe SO2 oxidation in solution and its role in atmospheric chemistry. Quart. I. Roy. Meteorol. Soc. 84:46-55. Penkett, S. A., B. M. R. {ones, K. A. Brice, and A. E. i. Eggleton (1979~. The importance of atmospheric ozone and hydrogen peroxide in oxidizing sulfur dioxide in cloud and rainwater. At- mos. Environ. 13:123-13 7. Pruppacher, H. R., and I. D. Klett (1978~. Microphysics of Clouds and Precipitation. Reidel, Boston, Mass., pp. 1-714. 87 Ross, A. B., and P. Neta (1979~. Rate constants for reactions of inorganic radicals in aqueous solution. NSRDS NBS 65. De- partment of Commerce, National Bureau of Standards, pp. 1- 55. Scott, W. D., and P. V. Hobbs (1967~. The formation of sulphate in wafer droplets.~. Atmos. Sci. 24:54-57. Stedman, D. H., W. L. Chameides, and R. i. Cicerone (1975~. The vertical distribution of soluble gases in the troposphere. Geophys. Res. Lett. 2:333-336. Taube, H., and W. C. Bray (1940~. Chain reactions in aqueous solutions containing ozone, hydrogen peroxide and acid. I. Amer. Chern. Soc. 62:3357-3375. Heterogeneous Processes fames, D. E., ed. (1979~. U.S. NationalReport, 1975-1978, Seven- teenth General Assembly International Union of Geodesy and Geophysics, Canberra, Australia, December 2-15, American Geophysical Union, Washington, D.C. Kiang, C. S., D. Stauffer, V. A. Mohnen, I. Bricard, and D. Vigla (1973~. Heteromolecular nucleation theory applied to gas-to- particle conversion. A tmos. Environ. 7: 1279- 1283. Schryer, David R., ed. (1982~. Heterogeneous Atmospheric Chemistry. Geophysical Monograph 26. American Geophysical Union, Washington, D.C. Turco, R. P., O. B. Toon, R. C. Whitten, R. G. Keesee, and P. Hamill (1982~. Importance of heterogeneous processes to tro- pospheric chemistry: studies with a one-dimensional model, in Heterogeneous Atmospheric Chemistry. Geophysical Monograph 26. David R. Schryer, ed. American Geophysical Union, Washing- ton, D.C., pp. 231-240. Vali, Gabor, ed. (1976~. Proceedings of the Third International Workshop on Ice Nucleus Measurements, Subcommittee on Nucleation, Inter- national Commission on Cloud Physics, International Associa- tion of Meteorology and Atmospheric Physics, International Union of Geodesy and Geophysics, Laramie, Wyo., ~an. 1976.

88 WET AND DRY REMOVAL PROCESSES BY B. HICKS, D. LENSCHOW, AND V. MOHNEN In simple terms, the tropospheric concentrations of many species are determined by their rates of emission and removal. The species source is not usually a single term; it typically includes contributions from both natu- ral and anthropogenic sources, as well as in situ produc- tion. Likewise, the removal rate is made up of both transformation and transport terms. However, deposi- tion to the earth's surface constitutes the major sink for many tropospheric trace gases and aerosols. This sec- tion discusses removal at the earth's surface, which in many cases is the major factor limiting tropospheric trace gas concentrations. The residence time of aerosol particles ranges from the order of a day in the atmospheric boundary layer (the lowest ~ 1000 m of the troposphere, which is closely coupled to the surface by convection and mechanical mixing) to more than a week in the upper troposphere. These residence times suggest that the physical removal processes are equivalent to chemical transformation rates of about ~ percent per hour. It is convenient to differentiate between wet and dry deposition processes. The process by which falling hy- drometeors (e.g., rain, snow, and sleet) carry atmo- spheric trace constituents to the surface is known as wet deposition. The processes of gravitational settling of particles and of turbulent transport (and subsequent impaction, interception, and absorption to exposed sur- faces) of particles and gases to the surface are collectively known as dry deposition. There are several potentially important processes that do not fit neatly into either category. These include fog droplet interception, scav- enging by spray droplets at sea, and processes associated with dewLall. In practice it is sometimes not possible to apportion total deposition between wet and dry components. Moreover, in some circumstances it is clear that wet dominates dry, while in other cases the opposite appears to be true. Such generalities should be modified according to the chemical and physical nature of the species under con- sideration. For example, submicrometer particles are poorly captured by falling raindrops and are ineff~ci- ently deposited by dry mechanisms. However, they can enter into the in-cloud nucleation, coagulation, and coa- lescence processes that precede precipitation. Soluble trace gases (such as HNO3 vapor) are easily scavenged by falling raindrops and are rapidly adsorbed at exposed surfaces. WET DEPOSITION Wet deposition constitutes a very intermittent but highly efficient mechanism for transforming and even- tually removing trace gases and aerosol particles from the troposphere. Aerosol particles act as nuclei for the condensation of water in warm clouds and for the gener- ation of ice crystals in supercooled clouds. Subsequent coalescence and accretion lead to a wide range of droplet sizes, the largest of which initiate the precipitation proc- ess (e. g., snow and hail). The droplets collect other par- ticles and gas as they fall, especially when passing through urban plumes or through a polluted boundary layer. The terms rainout and washout are sometimes used to differentiate between in-cloud and subcloud scavenging, but their use is dropping from favor. Airborne particles are removed by falling raindrops below cloud base by much the same physical processes as cloud droplets scavenge particles within clouds. Scav- enging efficiencies are related to particle size and chemi- cal composition. In-cloud nucleation processes scavenge soluble, hydroscopic particles more easily than particles that do not have an affinity for water. Likewise, soluble and chemically reactive trace gases are more readily removed than less reactive species. There has been considerable effort to document and model cloud scavenging systems. For example, a deep- rooted convective cell feeds on air from the boundary layer, which is normally the most polluted portion of the atmosphere, whereas some stratiform cloud systems form above the boundary layer and thus exist in a rela- tively cleaner environment. Scavenging characteristics of the two kinds of cloud systems will certainly be differ- ent; futhermore, the trace gases and aerosol particles accessible to them will differ. Because of the differences between scavenging within clouds by nucleation (and related cloud-physical processes) and subcloud scaveng- ing by impaction and adsorption by falling hydromete- ors, special care must be taken to interpret correctly the results of experimental case studies. The results of a study of particle washout by raindrops falling through a smokestack plume may not necessarily be applicable to the case of long-range transport and subsequent precipi- tation scavenging in remote regions. The manner in which trace gases and aerosol particles are scavenged by clouds and by falling precipitation determines the preferred parameterization for inclusion in models. Steady precipitation falling through a pol

CRITICAL PROCESSES luted air mass will deplete pollutants at a rate propor- tional to the instantaneous concentration, so that an exponential decay of concentration with time will result. Measurements of the composition of precipitation through the duration of such a simple precipitation epi- sode will display the same exponential time decay. Thus, for some short-term studies an exponential "scavenging rate" is used to relate precipitation quality to air chemis- try, analogous in form to the decay constant of radioac- tive decay. This scavenging rate for small particles is typically of the order of 10-5 to 10-4 per second. If average concentrations of chemical species in pre- cipitation are of concern (with respect either to space or to time), then it is usual to assume a first-order linear relationship between concentrations of the species of interest in precipitation and their concentrations in air. The "scavenging ratio" defined in this way (i.e., the precipitation concentration divided by the air concen- tration) is expressed either on a volumetric or on a mass basis, and sometimes the precise definition is not made clear. Further confusion arises from the influence of me- teorological factors, especially precipitation type and in- tensity, and the frequent uncertainty concerning the rel- ative contributions of in-cloud and subcloud processes. The term "washout ratio" is frequently used synony- mously with " scavenging ratio. " Early studies of radioactive fallout showed that in- cloud mechanisms result in highly efficient scavenging of many types of airborne gaseous and particulate mate- rial. Contemporary studies of precipitation acidity have shown that in-cloud reactions can be rapid, and that experimental determinations of scavenging ratios can be strongly affected by these reactions. Ambient SO2 can interact with other chemical constituents in hydro- meteors (e.g., H2O2 and nitrogen oxides) and can be deposited as sulfate. The role of clouds as sites for accel- erated chemical reactions is a major emphasis of the research program described elsewhere in this report. Evaporation of falling hydrometeors is sometimes sufficiently rapid that none of the precipitation leaving the cloud base reaches the ground. This process (virga) is a familiar example of cloud-related mechanisms for transforming material chemically and physically and for relocating it in the troposphere. The overall effect of clouds that do not rain is not well understood. An illustration ofthe uncertainty regarding wet depo- sition processes is the case of SO2 scavenging. It is known that SO2 is absorbed in rain droplets at a rate that is strongly affected by the pH of the droplet. This ab- sorption causes sulfur scavenging to be dependent on all other factors that influence precipitation acidity, many of which are not yet known. Temperature is acknowl- edged to have a strong influence on the rate at which 89 dissolved SO2 is oxidized; the results of scavenging stud- ies carried out in winter must be expected to differ from those obtained in summer. Finally, it is certain that scav- enging characteristics depend on the physical nature of the precipitation. Most studies to date have been of rain. Freezing rain, hail, and snow have yet to receive much attention. Research conducted on the relationships between precipitation chemistry and air quality has often been hampered by the lack of chemical data at cloud height. There are obvious difficulties involved in using ground- level air chemistry observations as a basis for calculating scavenging ratios. Scavenging ratios for materials of surface origin are likely to be underestimates if ground- level air concentrations are used in their derivation, be- cause air concentrations near the surface will generally be greater than those characteristic of the air from which the material is being scavenged by precipitation. Simi- larly, experimental evaluations of scavenging ratios for substances with sources in the upper troposphere will tend to be too high if ground-level air concentration data are used. Unless this source of error is eliminated by appropriate use of aircraft sampling or remote probing to measure chemical concentrations in the air that is being scavenged, there is little hope of resolving ques- tions regarding the role of synoptic variables and cloud chemistry. The mechanism for generating precipitation clearly affects precipitation quality. If rain falls through a pol- luted layer of air beneath cloud level, then a first-order dilution effect results. Thus the concentration of some soluble trace gas in rain sampled at ground level would tend to vary inversely with the amount of rain that fell. On the other hand, if air from the same polluted layer were drawn into an active orographic cloud scavenging material from a constant air stream and depositing it in a steady rain, then the concentrations in the rain would be far less influenced by the amount of rain that fell. In general, the relationship between precipitation chemis- try and precipitation amount is indeed found to lie be- tween the extremes corresponding to these two concep- tual examples. lust as the quality of rain depends on the quality of the air from which it falls, the total deposition of chemical species in precipitation is closely linked to the quantity of precipitation. Precipitation is a highly variable phenom- enon that cannot be predicted with accuracy. The net deposition of chemicals associated with precipitation is more variable and even more difficult to predict. It seems unlikely that the capability to predict wet deposi- tion at a single location on an event basis will ever be developed, since no organized prediction scheme can hope to reproduce the details of the random factors asso

go elated with the location and intensity of single storm cells. These deposition "footprints" have been the sub- ject of some study; first as a consequence of concern about radioactive fallout, but most recently under the aegis of acid rain. Precipitation quality recorded during a single period of uninterrupted precipitation (an event) will display features corresponding to a cross section through the event. As a consequence, interpretation of the fine structure of time-sequence records observed at a single station is quite difficult, since it is often not possi- ble to determine which part of the observed behavior is due to meteorological or air chemical processes and which is a result of the vagaries of the sampling cross section. However, the prediction of average patterns and of statistical variability (both with time and space) are achievable goals, provided appropriate information becomes available on the physical and chemical proc ~. esses ot Importance. Because the deposition "footprints" of different chemical compounds in single precipitation events tend to look alike, comparisons between deposition records of different chemical species must be expected to yield high correlation coefficients. Time records of sulfate deposi- tion in rain at some specific site should be highly corre- lated with nitrate, for example, without the need to , . Imagine some cause-and-effect relationship between these two species. In this regard, the determination of a low correlation coefficient may be as informative as de- tecting an unusually high value. Recent emphasis on precipitation acidity has tended to divert attention from the basic questions of precipita- tion scavenging of particular trace gases and aerosol particles. High rainfall acidity does not necessarily mean very high concentrations of dissolved trace species in the rain, nor does a pH of 7 mean that the rain is completely free of dissolved material. Precipitation col- lected at remote sites is usually somewhat more acidic than expected solely on the basis of equilibrium with atmospheric CO2 (pH about 5.6) as a result of back- ground levels of nitrates and sulfates. The worrying feature of acid deposition over North America, for ex- ample, is not only its pH but also the concentrations of chemicals in the solution being deposited. In contrast to the case of dry deposition, wet deposi- tion rates can be monitored with existing techniques. Wet/dry collectors, which protect precipitation samples from contamination by dryfall processes during periods between rain events, became popular during the era of radioactive fallout studies and are now familiar instru- ments in most deposition measurement programs. The use of bulk collection devices is discouraged for studies of long-term wet deposition, because of the considerable uncertainty about the effect of dry deposition between . . . precipitation events. PART II ASSESSMENTS OF CURRENT UNDERSTANDING Collection of precipitation for chemical analysis of trace constituents, although conceptually simple, is sus- ceptible to many problems. Many trace species of inter- est have extremely low concentrations, particularly in the remote regions that are often areas of concern when studying global biogeochemical cycles. Precipitation . . . . .. . samp es contammg these species are easily contaml- nated during collection and subsequent sample han- dling. Problems with wall losses in the collection and storage vessel, biological activity in the samples, loss of volatile species, and so on, demand that the greatest care and preparation be taken before undertaking the seem- ingly simple task of collecting rain for chemical analysis. FOG AND DEWFALL Precipitation collection devices fail to provide repre- sentative data on deposition via fog interception and dewfall. Fog droplets can contain relatively high concen- trations of pollutants; the physical and chemical proc- esses involved are precisely those that contribute to the in-cloud component of normal precipitation scaveng- ing. Iffogforms in polluted air, significant deposition vie fog droplet interception and deposition is likely. It is not obvious whether this process best fits under the general category of wet or dry deposition, and this uncertainty sometimes causes the process to be overlooked. Studies ofthe acidity of cloud liquid water have shown that droplet interception by forest canopies can be a major route for acid deposition. Exceedingly low pH values have been reported, presumably in circum- stances (such as high-altitude, stratiform clouds) in which there is minimal buffering and negligible access to the trace metals and NH3 compounds that can serve as neutralizing agents. It is clear that even uncontaminated fog droplets will cause dry deposition rates to be modified by wetting surfaces. Dewfall (and other processes that cause liquid water to form on exposed surfaces) will modify dry depo- sition rates in much the same manner, and for some chemical species net deposition rates can be significantly affected. . DRY DEPOSITION Dry deposition rates are influenced strongly by the nature of the surface and by source characteristics. Sur- face emissions are held in closer contact with the ground than emissions released at greater altitudes, so that in the former case concentration loss by dry deposition would be expected to be greater. Consequently, dry deposition fluxes tend to be highest near sources, whereas the high- est rates of wet deposition ofthe same substances may be found much further downwind.

CRITICAL PROCESSES Dry deposition rates are intimately related to atmo- spheric concentrations in the air near the surface A . first-order linear relationship is usually assumed. The coefficient of proportionality between atmospheric con- centrations and dry deposition rates, which is known as the deposition velocity, clearly depends on the meteoro- logical conditions, the chemical nature of the substance in question, and the nature of the surface on which it is being deposited. The term "deposition velocity" suggests an analogy with gravitational settling that is usually incorrect. In most instances, deposition through the atmosphere is accomplished by turbulent mixing to within a very short distance of the final receptor surface, followed by diffu- sive transfer across a layer of near laminar flow immedi- ately next to the surface. Turbulent transfer very near the surface is possibly influenced by the presence of small roughness features of the surface, electrostatic forces, and other mass and energy exchanges that are occurring. Discussion of the relationship between these (and other) potentially important factors is simplified by use of a resistance analogy, in which the inverse of the deposition velocity is viewed as a resistance to transfer in direct analogy with electrical resistance as described by Ohm's law. Individual resistances are associated with each process contributing to the dry deposition phenom- enon, and these individual resistances are combined in a network whose structure reproduces the conceptual linkage between the various contributing mechanisms. A total resistance to transfer is then evaluated by using the electrical analog. The analogy is not perfect, how- ever, it permits the processes involved in trace gas and aerosol particle deposition to be compared and com- bined in a logical manner. Particles already deposited on a dry surface can be resuspended by wind gusts exceeding some critical value related to the size and density of the particle. Soil grains and particles of surface biological origin can be en- trained in the lower atmosphere under some conditions. Suitable circumstances are not necessarily unusual. In arid regions, a surface saltation layer is frequently visible in strong winds, and it has been demonstrated that such aeolian particles can be carried into the upper tropo- sphere by deep convection and transported horizontally for considerable distances. The generation of particles as a result of chemical reactions occurring within vegetated canopies has been postulated as a cause for the blue haze phenomenon associated with forests in many parts of the world. Ocean spray is another well-known example of surface generation of particles. Resuspended particles constitute another form of atmosphere-surface interac- tion, thus sharing many of the features normally associ- ated with dry deposition. There is considerable scientific disagreement about 91 the mechanisms involved in dry deposition. Models (such as the resistance models mentioned above) that combine knowledge of individual processes to simulate natural phenomena occasionally omit processes that are sometimes considered to be important. However, all such models enable a test to be made of scientists' ability to simulate nature on the basis of their understanding of its component parts. For some circumstances and for some chemical species, the most important factors af- fecting dry deposition have been formulated well enough to permit fairly accurate modeling. The sum- mary of the dry deposition of certain chemical species that follows is based on a contribution to the Critical Assessment Review Papers on acid deposition, soon to be released by the Environmental Protection Agency. SO2. Uptake by plants is largely via stomates during daytime, but about 25 percent is apparently via the epidermis of leaves. At night, stomata! resistance in- creases substantially. When moisture condenses on the surface, resistances to transfer should decrease substan- tially. Deposition to masonry and other mineral surfaces is strongly influenced by the chemical composition of the surface material. To water, snow, or ice surfaces, deposi- tion rates are influenced by the pH of the surface water and by the presence of liquid films. O3. Dry deposition to plants is much like SO2, but with a significant cuticular uptake at night and with the presence of surface moisture minimizing deposition rates. Deposition to water surfaces is generally very slow. NO2. Similar to O3 for deposition to plants, but with a somewhat greater resistance to transfer. Even though NO2 is insoluble in water at low concentrations, deposi- tion to water surfaces might be quite efficient. NH3. No direct measurements are yet available, but a similarity to SO2 appears likely. Submicrometer particles. Deposition to smooth sur- faces is a minimum for particles of about 0. 5-pm diame- ter. Deposition velocities increase as particle size in- creases, until the terminal settling velocity predicted by the Stokes-Cunningham formulation is reached. Very small particles are deposited at rates that are controlled by Brownian diffusivity across a limiting quasi-laminar layer in contact with the surface. For rougher surfaces, deposition velocities tend to increase. Supermicrometer particles. Turbulence can cause particles to be deposited by inertial impaction and inter- ception, with deposition velocities greater than the Stokes-Cunningham prediction. Particle shape is an im- portant factor.

92 Sulfate particles. A value of 0.1 cm/s is often used for the deposition velocity for sulfate particles. However, recent experiments have demonstrated that deposition velocities for sulfate aerosol vary with the roughness of the surface. Values less than 0.1 cm/s seem appropriate for snow and ice, and about 0.2 to 0.3 cm/s for growing pasture and grassland. There is considerable disagree- ment concerning forests. Some workers use large depo- sition velocities (approaching 0.7 cm/s), while others prefer to continue to use the value 0.1 cm/s used in early transport and dispersion models. Phenomenological differences appear likely. ~1 Dry deposition to the oceans remains a major un- known. Data obtained in laboratory experiments on trace gas exchange between the atmosphere and water surfaces indicate that exchange rates are limited by fac- tors associated with the liquid phase, especially with the Henry's law constant. The deposition of hydroscopic particles is known to be influenced by their growth upon entering the region of very high relative humidity near the water surface. However, the practical significance of the effect is still being debated. Of major importance is the fact that exceedingly little information is available for dry deposition under typical open ocean conditions. The average wind speed at sea is about 8 m/s, with a highly disturbed surface and much spray. In such condi- tions the relevance of experimental data obtained in laboratory experiments seems open to question. In some areas of the world ocean, such as the "Roaring Forties," the surface is sufficiently agitated that the con- cept of a distinct, identifiable surface between the air and the ocean becomes difficult to defend. Rather, there is an interracial layer with properties somewhat like a gas- liquid suspension. In such conditions, exchange oftrace gases and aerosol particles between the atmosphere and the ocean may be quite rapid but bidirectional. Limiting processes cannot yet be identified with confidence. Although detailed knowledge of many of the proc- esses involved is lacking, the ability exists to measure dry deposition fluxes in some circumstances, for some substances. Dry deposition to some surfaces can be mea- sured directly, e.g., in the cases of accumulation on snowpacks or ice, or on some mineral and vegetative surfaces. For very large particles, deposition can be measured by exposing artificial collection surfaces or vessels since the detailed nature ofthe surface plays a less important role. However, until recently there has been little information on the rate of deposition of small parti- cles and trace gases to natural surfaces exposed in natu- ral surroundings. In the last decade, methods developed for measuring the meteorological fluxes of heat, mois- ture, and momentum have been extended to 03, CO2, SO2, nitric acid vapor, nitrogen oxides, and various PART II ASSESSMENTS OF CURRENT UNDERSTANDING particulate pollutants, with varying degrees of success. Some of these experiments have been intensive case studies, using instrumented meteorological towers, and were intended to identify and quantify factors control- ling the deposition. Other studies have used instru- mented aircraft to measure spatial averages of deposi- tion fluxes over terrain of special interest. None have yet demonstrated a capability for routine monitoring. There are essentially two schools ofthought on moni- toring dry deposition. The first advocates the use of collecting surfaces and subsequent careful chemical analysis of material deposited on them. The second in- fers deposition rates from routine measurements of air concentration of the trace gases and aerosol particles of concern and of relevant atmospheric and surface quanti- ties. Collecting vessels have been used for generations in studies of dustfall and gained considerable popularity following their successful use in studies of radioactive fallout during the 1950s and 1960s. The inferential methods assume the eventual availability of accurate deposition velocities suitable for interpreting concentra- tion measurements. In the era of concern about radioactive fallout, dust- fall buckets were used to obtain estimates of radioactive deposition, especially of so-called local fallout immedi- ately downwind of nuclear explosions. It was recognized that the collection vessels failed to reproduce the micro- scale roughness features of natural surfaces, but this was not viewed as a major problem because the emphasis was on large "hot" particles and the need was to deter- mine upper limits on their deposition so that possible hazards could be assessed. Much further downwind, so-called global fallout was found to be associated with submicrometer particles similar to those likely to be of major interest in studies of global tropospheric chemistry. However, most of the distant radioactive fallout was transported in the upper troposphere and lower stratosphere, and its deposition was mainly by rainfall. The acknowledged inadequacies of collection buckets for dry deposition collection of global fallout were of relatively little concern because dry fallout was a small fraction of the total surface flux. The acknowledged limitations of surrogate-surface and collection vessel methods for evaluating dry deposi- tion have caused an active search for alternative moni- toring methods. In general, these alternative methods have been applied to studies of specific pollutants for which especially accurate and/or rapid response sensors are available. The philosophy of these experiments has not been to measure the long-term deposition flux, but instead to develop formulations suitable for deriving average deposition rates from other, more easily ob- tained information such as ambient concentrations, wind speed, and vegetation characteristics. Neverthe

CRITICAL PROCESSES less, several initiatives are under way to develop micro- meteorological methods for monitoring the surface fluxes of particular pollutants. Surrogate surface meth- ods are also being improved. Although these devices share many ofthe conceptual problems normally associ- ated with collection vessels, they appear to have consid- erable utility in some circumstances. It has been shown that deposition of small particles to surrogate surfaces is sometimes similar to that of foliage elements. However, none of the surrogate-surface or micrometeorolog~cal methods that have been identified to date has been suc- cessfully demonstrated to monitor the dry deposition of a pollutant being slowly deposited. BIBLIOGRAPHY Beille, S., and A. l. Alshout (1983~. Acid Deposition. D. Reidel, Dordrecht, Holland, 250 pp. Engelmann, R. I. (1968~. The calculation of precipitation scaveng- ing, in Meteorology and Atomic Energy, D. H. Slade, ed. U.S. Atomic Energy Commission. Galloway, I. N., and D. M. Whelpdale (1980~. An atmospheric sulfur budget for eastern North America. Atmos. Environ. 14:409-41 7. Galloway, I. N., I. D. Thornton, S. A. Norton, H. L. Volchok, and R. A. N. McLean (1982~. Trace metals in atmospheric deposi- tion:areviewandassessment.Atmos.Environ.16:1677-1700. 93 Greenfield, S. M. (1957~. Rain scavenging of radioactive particu- late matter from the atmosphere.~. Meteorol. 14:115-123. Hales, i. M. (1972~. Fundamentals of the theory of gas scavenging by rain. Atmos. Environ. 6:635-659. Hardy, Jr., E. P., and J. H. Harley, eds. (1958~. Environmental Contamination from Weapons Tests. Health and Safety Laboratory Report HASL-42A. U.S. Atomic Energy Commission. Hicks, B. B., M. L. Wesely, and J. L. Durham (1980~. Cntique of Methods to Measure Dry Deposition: Workshop Summary. EPA-600/9- 80-050. U.S. Environmental Protection Agency, 69 pp. (NTIS PB81-126443.) Lindberg, S. E., R. C. Harriss, and R. R. Turner (1982~. Atmo- spheric deposition of metals to forest vegetation. Science 215: 1609-1611. Liss, P. S., end W. G. N. Slinn, eds. (1983).Air-SeaExchangeof Gases and Particles, NATO ASI Series, Series C. Mathematical and Physical Sciences No. 108. D. Reidel, Dordrecht, Holland, 561 PP. Owens, I. S. (1918~. The measurement of atmospheric pollution. Quart. Hi. Roy. Meteorol. Soc. 44:149-170. Pruppacher, H. R., R. G. Semonin, and W. G. N. Slinn, eds. (1983~. Precipitation Scavenging, Dry Deposition, and Resuspension, Vol. 1, Precipitation Scavenging. Elsevier, New York, 729 pp. Pruppacher, H. R., R. G. Semonin, and W. G. N. Slinn, eds. (1983~. Precipitation Scavenging, Dry Deposition, and Resuspension, Vol. 2, Dry Deposition and Resuspension. Elsevier, New York, 731 PP Sehmel, G. A. (19803. Particle and gas dry deposition: a review. Atmos. Environ. 14:983-1012. Shannon, i. D. (1981~. A model of regional long-term average sulfur atmospheric pollution, surface removal, and net horizon- tal flux. Atmos. Environ. 13:1155-1163.

Next: 6 THE ROLE OF MODELING IN UNDERSTANDING TROPOSPHERIC CHEMICAL PROCESSES »
Global Tropospheric Chemistry: A Plan for Action Get This Book
×
Buy Paperback | $50.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

In a giant step toward managing today's pollution problems more effectively, this report lays out a framework to coordinate an interdisciplinary and international investigation of the chemical composition and cycles of the troposphere. The approach includes geographical surveys, field measurements, the development of appropriate models, and improved instrumentation.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!