Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 59
Studies in Geophysics: Energy and Climate II EFFLUENTS OF ENERGY PRODUCTION
OCR for page 60
Studies in Geophysics: Energy and Climate This page intentionally left blank.
OCR for page 61
Studies in Geophysics: Energy and Climate 3 Effluents of Energy Production: Particulates GEORGE D. ROBINSON Center for Environment and Man, Inc. 3.1 PARTICLES IN THE ATMOSPHERE Particles are a normal constituent of the atmosphere. If we exclude the larger water droplets in clouds or fog that have formed in slightly supersaturated air, particles can properly be called a trace constituent, with a global average mass mixing ratio near the surface of about 10−8, or 10 μg−3, in the units conventionally used in monitoring. We are concerned here with the proportion of this load that can be attributed to man’s activities and with any effects that a change of this proportion might have on climate. 3.2 SOURCES OF ATMOSPHERIC PARTICLES Particles may be characterized as wind-raised dust; wind-raised sea salt; direct products of combustion, soot, ash, condensed organic materials, etc.; indirect products of combustion, i.e., particles formed by chemical reactions in the atmosphere from the gaseous products of combustion— sulfates, organic nitrates, sulfuric and nitric acid; volcanic particles; and particles formed in the atmosphere from such products of plant and animal life and decay as terpenes, H2S, and NH3. Table 3.1 summarizes two attempts to estimate the annual production of particles by these various means. The material is several years old, but more recent work has not reduced the uncertainties indicated, which are certainly not overstated: the “nitrate” entry is particularly suspect. Table 3.1 also attempts to separate “natural” and “anthropogenic” sources; there is very little basis on which to make this separation in some categories, particularly “forest fires” and “soil dust.” Figure 3.1 shows the range of particle radius with which we are concerned and indicates the general nature of the radius-number distributions that are observed. 3.3 SINKS OF ATMOSPHERIC PARTICLES In the steady state, production and loss of atmospheric particles balance and a “mean residence time” may be defined as the ratio of loading to production rate. The estimate of mean loading given above was obtained from an assumed production of 1.8 × 1015 g yr−1 and a residence time of
OCR for page 62
Studies in Geophysics: Energy and Climate 10 days, which is the average residence time of a water molecule in the atmosphere. The major removal mechanisms are fallout under gravity without condensation; processes in which water condenses on a particle that later falls out in rain—conventionally termed rainout and the most effective sink; and processes in which the particle is captured by falling rain or snow and carried to the ground—conventionally termed washout. Some particles such as nitrates and certain organic materials may in some circumstances volatilize or decompose to gaseous products. The shape of the large radius end of the size-distribution curve is strongly influenced by dry fallout. The shape of the small-radius end depends to a great extent on coagulation processes, which do not change the mass mixing ratio but greatly affect optical properties. It will be clear from the nature of the removal processes, particularly those involving precipitation, that the residence time of a particle can be greatly influenced by the location and time of its production. A particle may exist in the atmosphere for a time measured in minutes or in years; this is true of both natural and man-made particles. Volcanic particles, for example, are often injected high in the atmosphere, away from the immediate influence of precipitation processes TABLE 3.1 Global Summary of Source Strengths for Atmospheric Particulate Matter Strength (Tg/yr) Source Natural Anthropogenic Primary Particle production Fly ash from coal — 36 Iron and steel industry emissions — 9 Nonfossil fuels (wood, mill wastes) — 8 Petroleum combustion — 2 (10–90) Incineration — 4 Agricultural emission — 10 Cement manufacture — 7 Miscellaneous — 16 Sea salt 1000 Soil dust (428–1100) 200 (?) Volcanic particles 4 Forest fires (3–150) 3 (?) SUBTOTAL 1207 92 Gas-to-particle conversion Sulfate from H2S (130–200) 204 — Sulfate from SO2 — 147 (130–200) Nitrate from NOx (60–430) 432 30 (30–35) Ammonium from NH3 (80–270) 269 — Organic aerosol from terpenes, hydrocarbons, etc. (75–200) 200 27 (15–90) SUBTOTAL 1105 204 TOTAL (773–2200) 2312 296 (185–415) FIGURE 3.1 Typical comprehensive size distribution for the principal tropospheric regimes and the size ranges important for turbidity, cloud formation, and mass concentration of particles. Curves a and b refer to possible variation of the size distribution with and without continuous production of very small particles. The arrow indicates the effect of pollution on the location of the maximum of the size distribution. After Inadvertent Climate Modification, Report of the Study of Man’s Impact on Climate. MIT Press, Cambridge, Mass., 1971. so that although they are a minor entry in Table 3.1, they have received considerable attention as a potential agent of climatic change. The same considerations apply to particles injected into the stratosphere by aircraft. 3.4 ANTHROPOGENIC PARTICLES The major sources of particles associated with man are industrial production and processing of materials such as metals and cement; combustion associated with industrial, commercial, and domestic needs including transportation; and agriculture. We look first at the industrial/domestic sector. Increasing populations, and increase in specific energy consumption, potentially increase particle production by processing and combustion, and increasingly within the last
OCR for page 63
Studies in Geophysics: Energy and Climate few decades steps have been taken to control emissions. Figure 3.2 illustrates the degree of success of control measures within the United States. Urban measurements of SO2 concentration may be as relevant to considerations of atmospheric particle load as are the urban particulate load measurements themselves. Figure 3.3 illustrates the effect of fuel controls on urban SO2 concentrations. Draconian control measures have been effective, urban SO2 concentrations declining in proportion to the controlled proportion of S in the fuel burned. Figure 3.4 shows the trend in SO2 concentration at a selection of stations. We emphasize the measurements of SO2 because of its fate in the atmosphere. Its mean global life as a gas is estimated to be two to three days. In the winter of 1968–1969, an estimate for the State of Connecticut was 2–3 hours (Hilst, 1970). Recent work in the St. Louis, Missouri, area suggests a few hours, with some indication that it disappears more quickly from the general urban plume produced by many small sources than from the highly concentrated stack plumes of major power plants. Possible sinks are absorption on ground surfaces and conversion to sulfuric acid with production of acid particles, some of which are in turn converted to (NH4)2SO4, NH4HSO4, or CaSO4. Such particles are universally present in the atmosphere. The production rate of Table 3.1 seems to imply that about 75 percent of SO emission is converted to sulfate particles before leaving the atmosphere, but some recent measurements FIGURE 3.2 Composite levels of total suspended particulate at urban and nonurban NASN stations. From “Monitoring and Air Quality Trends Report 1972,” U.S. Environmental Protection Agency, EPA-450/1–73–004. in highly polluted atmospheres find only about 10 percent conversion. Altshuler (1973) shows that at nonurban sites in eastern United States, concentrations of gaseous SO2 and particulate SO4−2 are about equal and of order 10 μg−3. Concentrations of SO4−2 of up to 30 μg−3 have been measured at rural sites in Sweden.* Investigation of details of the conversion is just beginning, but there seems little doubt that it occurs with a time constant, to some extent controlled by the concentrations of other pollutants, of hours to days and at efficiencies of not less than 10 percent in regions of high emission and perhaps more than 50 percent global average. We briefly mention particle production by agricultural practices. The first major contribution to the global load of particles is by wind-raised dust. The American dust-bowl phenomenon of the 1930’s has not recurred, in part because the accompanying climatic conditions have not recurred in their full severity, in part because of the success of positive countermeasures. There may now be a similar problem in China, and there is certainly a major windborne dust problem in overgrazed arid regions in Africa and India. The second contribution from agriculture is to the particle load due to combustion, particularly by the “slash and burn” type agriculture in less arid tropical and subtropical regions of Africa and Southeast Asia. The particle load from these agricultural practices is not negligible in the context of local climatic effects, but the associated immediate primary resource depletion is such as to call for remedial action on a short time scale without regard to possible longer-term climatic effects. In this discussion of climate and energy, we will therefore not further consider the sources associated with agricultural malpractice. 3.5 ATMOSPHERIC PARTICLES AND WEATHER AND CLIMATE The properties of atmospheric particles that affect the processes of weather and climate are those concerned with condensation of water on the particle, which, for convenience, we will term “nucleation properties,” and with interaction of the particle with solar and terrestrial radiation, which we will term “optical properties.” Since condensation or evaporation of water on or from the particle change its size and composition, the nucleation properties are not independent of the optical properties. NUCLEATION PROPERTIES Many of the constituents of atmospheric particles are hygroscopic or deliquescent. Figure 3.5(a) shows how the size of such particles might be expected to vary with the surround- * The measurements were made by L. Granat and S. Larssen and examined and communicated to me by R. J. Charlson, A. H. Vanderpol, and A. P. Waggoner of the University of Washington, Seattle, Washington.
OCR for page 64
Studies in Geophysics: Energy and Climate FIGURE 3.3 Comparison on SO2 trends at Bayonne, New Jersey, with regulations governing percent sulfur content in fuel. From “Monitoring and Air Quality Trends Report 1972,” U.S. Environmental Protection Agency, EPA–450/1–73–004. ing relative humidity. Charlson et al. (1974) have confirmed this type of behavior for particles in the atmosphere. Figure 3.5(b) illustrates some of their observations, taken near St. Louis, Missouri. The size actually attained by hygroscopic particles, and by all particles in a supersaturated atmosphere, is controlled by the rate of supply of water to the particle—the equilibrium radius may not be reached at any time as relative humidity changes at humidities below 100 percent, and at higher humidities only the larger particles may be effective in forming cloud droplets. Effectively, all particles in the atmosphere are condensation nuclei, but in practice only a proportion of them—“cloud condensation nuclei”—may be active in cloud formation on any one occasion. The number and size distribution of cloud or fog droplets forming in cooling air depend on the nature and size distribution of the particles present—clouds and fogs forming in polluted air will have properties different from those forming in unpolluted air. Subsequent precipitation from and dissolution of the clouds may be affected, and bulk optical properties will be different. Certain rare atmospheric particles have the property of initiating the freezing process in supercooled droplets at higher temperatures than that at which they freeze spontaneously (about −40°C). Certain pollutant particles might FIGURE 3.4 Composite levels of sulfur dioxide at 32 NASN stations. From “Monitoring and Air Quality Trends Report 1972,” U.S. Environmental Protection Agency, EPA–450/1–73–004.
OCR for page 65
Studies in Geophysics: Energy and Climate have this property. Silicates in wind-raised dust are perhaps the most common “natural” freezing nuclei. Some material processing, e.g., steel manufacture, might produce such nuclei artificially. Artificial introduction of particularly effective freezing nuclei is the basis of “rainmaking” operations, but there is no evidence at present of consistent substantial production of similarly effective nuclei by industrial and domestic activity, although localized effects have been reported. Direct sampling from aircraft and radar observations have demonstrated changes in the rain-producing processes and particle numbers and size distributions in summer clouds in the St. Louis area. There is also some indication that the areal distribution of precipitation is affected by the presence of the city, although in the latter case it is not yet possible to distinguish clearly between added particles and added heat as causative agents. FIGURE 3.5 (a) Relation of size and relative humidity for a typical sample of urban pollutant particles containing hygroscopic and deliquescent material, e.g., H2SO4 and (NH4)2SO4. Note the hysteresis and the indeterminacy of the relation for decreasing relative humidity, (b) Measured relation between relative humidity and the scattering coefficient of air samples. Tyson, Missouri. (Data from D. S. Covert, University of Washington.) Curve a, 2330h, September 24, 1973 (deliquescent particle behavior). Curve b, 1223h, September 23, 1973 (hygroscopic particle behavior). OPTICAL PROPERTIES OF SINGLE PARTICLES The interaction of a spherical homogeneous particle with electromagnetic radiation can be computed to any required degree of accuracy in terms of the ratio of its radius to the wavelength of the radiation and the “complex refractive index,” which characterizes refractive and absorptive properties of the material. The properties of meteorological interest are the optical cross section of the particle, the proportion of radiation incident on this cross section that is absorbed, and the polar diagram of the intensity of scattered radiation. The absorbed fraction is usually expressed as (l – ω0), where ω0 is the “albedo for single scatter.” The complex refractive index must be known for all wavelengths of solar and terrestrial radiation carrying significant energy. Figures 3.6 and 3.7 are examples, respectively computed and observed, of optical cross sections and polar scattering diagrams of particles found in the atmosphere. OPTICAL PROPERTIES OF ASSEMBLIES OF PARTICLES Many of the problems of the optical effects of suspensions of particles in the atmosphere can be discussed with sufficient accuracy by single scattering approximations—the supposition that a photon traversing the depth of the atmosphere is very unlikely to encounter more than one particle. This is particularly true if we are interested only in the total energy carried by the various streams of radiation. In this approximation, Figure 3.8 represents, schematically, the radiative effect of a layer of particles. The layer is illuminated from above by the direct solar beam and by a diffuse flux of scattered radiation. Some of the incident solar radiation has also been absorbed. The layer itself absorbs some radiation and scatters some upward and downward. Of the radiation transmitted downward, direct and diffuse, some is absorbed and some scattered upward, both from the atmosphere and the ground. Some of this is trans- FIGURE 3.6 Ratio of optical and geometric cross section for particles with a refractive index of 1.5.
OCR for page 66
Studies in Geophysics: Energy and Climate FIGURE 3.7 Phase function for atmospheric scattering. mitted, with or without scattering by the particle layer, and some is rescattered downward. The process continues as illustrated. The proportion scattered downward or upward varies with the direction of the incident radiation, as shown in Figure 3.9, so that the proportion of the radiation in the direct solar beam that is scattered upward and lost to space is greater for low than for high solar elevation. In particular, the proportion of backscatter from a diffuse flux is greater FIGURE 3.8 Effect of a particle layer on albedo (schematic). FIGURE 3.9 Upward and downward scattering of solar radiation by a particle (schematic). than that from a normally incident beam and less than that from a beam at grazing incidence. For high solar incidence and a sufficiently high underlying albedo, the possibility thus arises that a layer of particles might redirect to earth more energy from the upward diffuse flux than it redirects to space from the predominantly direct beam incident from above. A particle that itself absorbs no radiation and scatters some to space may in this way actually increase the radiation absorbed by the planet. If the particle itself absorbs, the likelihood of an increase in planetary absorption is even greater. For a detailed numerical solution of this problem we require, for each wavelength of radiation, a specification of the number density, the size distribution, the distribution with height, and the single scatter albedo of the particles (or the complex refractive index of the material of the particles). If there are present particles of different materials or if the refractive index varies with size or if there are non-spherical particles, we are faced with difficulties that have not yet been resolved. We must also know the absorptive and scattering properties of the overlying atmosphere and the albedo and angular variation of scattering of the underlying atmosphere-surface system. Although complex, the problem can be solved to a useful degree of approximation in a horizontally homogeneous atmosphere. Several authors have attacked the problem, and we will examine some examples of their work in Section 3.7. In general, they show that the interaction of atmospheric particles and solar radiation affects climate and weather processes in two ways: it may change (either increase or decrease) the energy absorbed by the planet, and it may change the pattern of heating and cooling in the surface-atmosphere system, initially perturbing the static stability of the atmosphere. Change in the planetary albedo means change in the equivalent radiative temperature of the planet. The difficulty of inferring conditions on a planet’s surface from its equivalent radiative temperature is well illustrated by considering the planet Venus, which is irradiated at about twice the intensity of earth and has about half its absorptivity. Its equivalent radiative temperature is thus not greatly different from that of the earth, but its surface temperature is believed to be about 700 K.
OCR for page 67
Studies in Geophysics: Energy and Climate Computation of the transmission and reflection of radiation from clouds is not different in principle from that for any other collection of particles. The single scatter approximation is, of course, inappropriate, and the assumption of spherical particles must be made so that there are uncertainties where ice clouds are concerned. The optical properties of pure water are reasonably well known, and the albedo of a cloud containing a given amount of pure water can be shown to increase as the number of cloud particles increases. Clouds formed in polluted air containing a large number of cloud condensation nuclei would, therefore, be expected to have a higher albedo than those formed from cleaner air, other factors being unchanged. The simple picture might, however, be changed if absorbing material were present in the polluting nuclei. The same equations of radiative transfer apply to terrestrial as to solar radiation, but in practice very different computations are required because of the nature of the source and absence of a direct beam. The effects of particles (excluding, of course, cloud droplets) are mainly determined by their small size relative to the wavelengths of concern. For the radiation carrying maximum energy flux, the parameter 2πa/λ is lower by a factor of 20 in the terrestrial radiation than in the solar radiation; and for a given refractive index, the optical cross section and single scatter albedo are much smaller at terrestrial than at solar wavelengths. The major difficulty is to know the complex refractive index at terrestrial wavelengths. In summary, we have in principle the means for computing the effects of homogeneous spherical particles on energy transfer in a stratified horizontally homogeneous atmosphere. We need to know, and to predict, the number, size, and material of all the particles and the refractive index of this material over a wide range of wavelengths. We need to know where the particles are and will be in the atmosphere. We need to know the reflectivity of the earth’s surface over a wide range of wavelengths. In principle, we can compute the required radiative properties of clouds; but in practice, because some limit must be set to computation (and because the atmosphere is usually not homogeneously stratified), it is desirable to know the bulk reflectivity and absorptivity of clouds and regard them as bounding surfaces. 3.6 SOME OBSERVATIONS OF THE RADIATIVE EFFECTS OF PARTICLES At present, we cannot clearly identify any effect of changes in atmospheric particle content on climate. There are some suggestive correlations of volcanic activity and surface temperature, but they are far from establishing a causal relation. There are numerous phenomena that can be described as effects of particles on weather. The most obvious concerns visibility—the distance at which a black object large enough to be comfortably resolved can be seen against the horizon. Contrast determines visibility, and contrast is reduced by scattering of light into the line of sight to the black object. For an average limiting contrast, the visual range is where σ is the scattering coefficient of the atmosphere. For a particle-free atmosphere at sea level, σ is about 1.5 × 10−2 km−1, so V is about 270 km. Such ranges have very occasionally been reported in the Antarctic. If there are 10 μg−3 of sulfate particles of density 1.67, all of radius 0.3 μm at sea level, V is about 65 km. (This is the maximum optical effect of the loading that we have suggested as a global average.) For the current U.S. urban average of about 80 μg–3, the minimum visibility would be 8 km if the particles were of this material and size. These numbers, although not unrealistic, are computed for the size distribution that produces the maximum optical effect. Recent observations in Sweden (see footnote in Section 3.4) have shown a tenfold variation of scattering coefficient for a given mass loading of SO4=, the maximum effect corresponding to that computed here. The size distribution of the particles is all important. Figure 3.10 (from Munn, 1973) shows the number of hours with visibility 10 km or less in an industrial region (Windsor, Ontario) and two locations remote from industry (Mont Joli, Quebec, and Gander, Newfoundland). Windsor reflects the urban improvement shown in Figure 3.2, but the frequency of haze at the remote stations is increasing. Numerous observations of the direct and diffuse components of solar radiation in cities show the effect of particles in the atmosphere. Robinson (1962) examined records from the suburbs of London and Vienna in a period around 1950 and found on average both scattering and absorption in FIGURE 3.10 Secular trend in reports of smoke, haze, or dust at Windsor, Mont Joli, and Gander Airports. (May–October) Source: Munn (1973).
OCR for page 68
Studies in Geophysics: Energy and Climate cloudless atmospheres to be about 10 percent in excess of the expectation for clean air. A few observations from aircraft suggested that not more than one fifth of the excess scattering was directed upward. These observations, made before the institution of clean air acts, suggest a modal single scatter albedo of about 0.5 for the particles concerned. Direct measurements of the optical properties of “industrial haze” in southern England were reported by Waldram (1945), his method of observation being visual photometry of searchlight beams. He made the first direct measurements of absorption of visible radiation by these particles. Figure 3.7 illustrates one of the polar scatter diagrams that he measured, on an occasion when the effective ω0 was about 0.55. [Figure 3.7 also contains an example of a polar scatter diagram from the extensive observations of Barteneva (1960).] Charlson and his collaborators (e.g., Lin et al., 1973; Charlson et al., 1974) have recently perfected portable equipment that allows investigation of both optical and nucleating properties of particles and have used it in urban and rural areas. Their investigation is in its early stages, but for urban particles their preliminary measurements indicate single scatter albedos for visible light, mainly in the range 0.8 to 0.4 with the mode near 0.7. They have confirmed that many particles are composed mainly of sulfuric acid and the ammonium sulfates—the nature of the absorbing constituents is not clear. Dzubay and Stevens (1973) collecting particles with radii less than about 1 μm in urban St. Louis write, “… the small [particles] … make a black deposit … consistent with the notion that … the small particles consist of combustion and secondary aerosols. At least 75 percent of the sulphur—is contained in the small particles—significant amounts of the sulphur may be bound to light cations…H+, NH4+.” This suggests absorbing material originating in combustion either carried on sulfate particles or associated with the same size range. The combustion concerned is in a city subject to EPA regulation. Although Robinson (1947) claimed to have observed it in the London atmosphere, there are yet no incontrovertible observations of modification of the terrestrial radiation field by particles. The effect sought is small, measurement is difficult, and alternative explanations of anomalies are available. There is no reason to question the results of the computations that indicate modification. Observations in the St. Louis, Missouri, area have shown modification of the nature of convective clouds by the particles in the urban atmosphere, but there has been no direct confirmation of the expected increase in the albedo of clouds forming in polluted air; and because of the wide range of observed cloud albedo, a very extensive statistical investigation would be required. Satellite cloud observations should contain appropriate material. On the other hand, several observers (e.g., Robinson, 1958) have reported absorption of solar radiation in cloud considerably in excess of that which would be expected from pure water clouds—an effect that can be explained by the inclusion of absorbing particles of the type observed in urban polluted air. 3.7 COMPUTATION OF THE RADIATIVE EFFECTS OF POLLUTANT PARTICLES SURFACE TEMPERATURE We consider, as examples of the radiative effect of realistic pollutant particle loadings, two theoretical studies, one of particles in the lower atmosphere and one of stratospheric pollution. Atwater (1970b) has studied the effect on both solar and terrestrial radiation of a low-lying layer of particles simulating urban pollution. He used a size distribution corresponding to that measured in a city atmosphere by Peterson et al. (1969) and made computations for a range of complex refractive index. We select from Atwater’s tables the case with refractive index for solar radiation (λ = 500 nm), m = 1.5 − 0.03i and for terrestrial radiation (λ ~ 10 μm), u = 1.5 − 0.2i. The real part is appropriate to hydrated sulfuric acid. The imaginary part for solar radiation is within the range measured later by Lin et al. (1973) in urban aerosol. The imaginary part for terrestrial radiation is appropriate to aqueous solutions. The scattering and absorption coefficients computed by Atwater are i.e., an “effective single scatter albedo” for the layer of 0.83, and The results for solar radiation correspond to a visibility of 12 km. For a 1000-m layer and the global average solar power at the surface, which is around 150 Wm−2, the mean heating rate due to absorption of solar radiation is about 0.4 K per day. The perturbation of terrestrial radiation depends not only on local conditions in the polluted layer but on the temperature throughout the atmosphere and particularly on cloud conditions. Atwater made computations for a standard case without cloud, with σabsorption of 0.01 km−1, and he found the mean cooling within the layer of particles to be about 0.5 K per day, approximately equal in magnitude to the solar heating. Atwater’s work suggests that in midlatitude cities temperature changes due to perturbation of solar and terrestrial radiation averaged over the year almost exactly balance. The balance is, however, the result of a redistribution of heating and cooling. Daytime heating is reduced at the surface and increased within the particle layer. Nighttime cooling is reduced at the surface and increased within the particle layer. At the surface, there is a moderation of the diurnal temperature cycle and a net decrease in solar radiation. The potential effect on crop growth, particularly early and late in the growing season, deserves more careful examination.
OCR for page 69
Studies in Geophysics: Energy and Climate PLANETARY ALBEDO Atwater (1970a) also computed the effect on planetary albedo of a layer of particles in the lower atmosphere. He found the relation, which we discussed qualitatively above, between planetary albedo change, surface albedo, absorption by the particle layer, and backscattering by the particle layer. Figure 3.11 illustrates some of his conclusions. In this diagram a particle layer is characterized by absorption and backscattering coefficients, and for each surface albedo a line divides those layers that increase planetary albedo from those that decrease it. Some observations of the properties of particle layers by Waldram and Robinson are plotted on this diagram; actual atmospheric particle layers appear to be close to the neutral line for the average land surface albedo of 0.15. All the observed layers would decrease planetary albedo over snow, all would increase it over quiet ocean waters with albedo around 0.05. B. M. Herman of the University of Arizona (private communication) has studied the influence of a layer of particles in the lower stratosphere. His computations are for a size distribution similar to that of naturally occurring stratospheric particles as observed by Friend and for visible solar radiation (λ = 500 nm). He allows for a tropospheric particle loading with optical depth in the zenith of 0.1, which is consistent with our estimate of a mean mass mixing ratio of about 10−8 for particles with refractive index (real part) 1.5 and radius 0.5 μm. He integrates for solar position over the day for each month of the year in each 10° square of the northern hemisphere, using realistic estimates of the underlying albedo, neglecting optical effects above the perturbing layer. He performs the computations for several values of single scatter albedo. Figures 3.12 and 3.13 show the nature FIGURE 3.11 Effect of particles on planetary albedo. FIGURE 3.12 Albedo increase by latitude and month following addition of a 10-km layer of 0.08 μg m–3 of particles, ω0 = 1 (no absorption). Units are 105 × fractional albedo change. FIGURE 3.13 Albedo decrease by latitude and month following addition of a 10-km layer of 0.08 μg m–3 of particles, ω0 = 0.9. Units are 105 × fractional albedo change. of his results for no absorption and for a single scatter albedo of 0.9. In the first case, the particle layer increases the planetary albedo at all locations; in the second case there is a decreased albedo everywhere. The albedo perturbation scales linearly with the mass loading of particles up to loadings more than 20 times that used in preparing Figures 3.12 and 3.13. To extend Herman’s calculations to a realistic situation, it is necessary to know the exact nature of the size distribution of the particles, their complex refractive index over the whole solar spectrum, and the albedo of the underlying surface-atmosphere-cloud system, again over the whole solar spectrum. This we do not know, except that there must be considerable differences between values in the solar infrared and those observed in the visible radiation. 3.8 SUMMARY AND PROSPECTS FOR THE FUTURE We have seen that the burning of fossil fuel containing sulfur leads to the formation of sulfuric acid particles in the
OCR for page 70
Studies in Geophysics: Energy and Climate atmosphere. They have a radius around 0.5 μm and are efficient scatterers of solar radiation. If they are pure sulfuric acid particles, they would be expected to absorb about 1 percent of the energy of the solar radiation falling on them, mainly in the infrared. Measurements, however, indicate that in the atmosphere they are often, perhaps always, associated with other material, which causes absorption of more than 10 percent of the intercepted solar radiation. These conditions exist in and near cities in which regulation of particulate emissions has been enforced. Visibility is a sensitive indicator of atmospheric particle loading, and we have noted that while emission control has improved visibility in and near industrial cities, there is a trend toward reduced visibility in locations about 1000 km downwind of regions of major industrial activity. There is no evidence of recent increases in more remote areas (e.g., the southern Pacific Ocean). Particles of the type in question would be expected to change the optical properties of clouds— their absorption and reflection of solar radiation. There are a few observations of unexpectedly high absorption, but increase of cloud albedo has not yet been firmly established. There is some evidence of a modification of the processes leading to precipitation in clouds formed in air containing combustion-produced particles. Known mechanisms of removal lead us to expect an average lifetime in the atmosphere of a few days, but for an individual particle this lifetime would be expected to have a very wide range. There should be a close analogy with the life in the atmosphere of water molecules. For these the mean life is 10 days. Many have only a few hours’ stay, but some of the small proportion that enter the stratosphere may stay there for several years. Computation of the energetic consequences of the radiative properties of particles has concentrated on two areas: perturbation of the planetary albedo and perturbation of local heating rates. If observed properties of particles are used in the computations (rather than the optical constants of pure sulfuric acid solutions), we find that the sign of the perturbation of planetary albedo is uncertain: the magnitude of the perturbation must be small. This result is independent of particle loading, at least up to loadings considerably greater than those currently observed, and it holds whether the particles are in the troposphere or the stratosphere. We can greatly increase emission of the kind of particle now produced by combustion in industrial communities without greatly changing the integrated radiative properties of our planet’s disk as seen from space, unless, as may be the case, the increased particle loading changes the albedo of cloud. (We should remember that a change of optical properties of the constituent droplets of a cloud does not necessarily mean a change of its albedo.) Current air-quality and emission standards for particles and SO2 are set by reference to effects on human health, not to effects on weather and climate. There is little doubt that an increase in particles and sulfurous emissions to a magnitude that might have global climatic consequences would be intolerable from the point of view of community health. Unless there is considerable improvement in the physical and economic efficiency of techniques for removal of particles and particle-producing gases, health rather than climatic considerations may be the ultimate constraint on the burning of low-grade fossil fuels. Computation also shows that although the amount of solar radiation reaching the ground is decreased, the net perturbation of local heating rate by industrial pollution in and near a midlatitude city is small when averaged over the day and year. This aspect of the energetic effects of particle content cannot be satisfactorily studied in isolation. A simple analysis demonstrates this. Consider the combustion of 1 ton (1000 kg) of coal containing 2.5 percent S. Current practice concerning particulate emission control is observed, but SO2 in not removed. Twenty-five kilograms of the S is burned. Assume that 50 percent of the resulting SO2 is converted to H2 SO4 · 4H2O particles of radius 0.5 μm. There are 50 kg of these particles. We assume a density of 2 × 103 kg m−3 for the material of the particles. The geometrical cross section of this 50 kg of particles is The optical cross section is at least twice this, i.e., We assume that the particles have a life in the atmosphere of one day. The global mean (night and day through the year and over the planet) of solar power reaching the surface for 50 percent cloudiness is about 150 W m−2. To compute the solar energy absorbed by the particles, we make two alternative assumptions: (a) They are pure H2SO4 · 4H2O with a single scatter albedo for the whole solar radiation of 0.97. (b) They are contaminated by or associated with other material, and their effective single scatter albedo is 0.9. (This is rather cleaner than the current particle load near St. Louis.) The absorbed solar energy is, for Case (a), ~ 104 kWh; for Case (b) ~ 3 × 104 kWh. The calorific value of a ton of coal, when its products have mixed with the atmosphere, is ~ 8 × 103 kWh. Burning 2.5 percent sulfur coal with no control of SO2 emission, even with complete removal of all solid emission, leads on average to a net loss of energy at the earth’s surface. This is probably compensated by adjustments in the field of terrestrial radiation, but our exercise suggests that climatic effects of particle production should not be considered independently of the effects of heat production, which are the subject of other chapters in this volume, nor of the effects of CO2 production on the terrestrial radiation field. 3.9 ACKNOWLEDGMENTS I wish to thank M. A. Atwater of the Center for the Environment and Man, Inc., Hartford, Connecticut; R. J. Charlson
OCR for page 71
Studies in Geophysics: Energy and Climate and A. P. Waggoner of the Department of Civil Engineering and the Institute for Environmental Studies, University of Washington, Seattle, Washington; and B. M. Herman of the Institute of Atmospheric Sciences, University of Arizona, Tucson, Arizona, each of whom allowed me to use results of his unpublished work during preparation of this paper. REFERENCES Altshuler, A. P. (1973). Atmospheric sulphur dioxide and sulphate. Distribution of concentration at urban and non-urban sites in the United States, Environ. Sci. Technol. 7, 709. Atwater, M. A. (1970a). Planetary albedo changed due to aerosols, Science 170, 64. Atwater, M. A. (1970b). Investigation of the radiation balance for polluted layers of the urban environment, Ph.D. Thesis, New York U. Barteneva, O. D. (1960). Scattering functions of light in the atmospheric boundary layer, Izv. Geophys. Sci. 12, 1852 (English transl. p. 1237). Charlson, R. J., A. H. Vanderpol, D. S. Covert, A. P. Waggoner, and N. C. Ahlquist (1974). H2SO4/(NH4)2SO4 background aerosol: Optical detection in St. Louis region, Atmos. Environ. 8, 1257. Dzubay, T. G., and R. K. Stevens (1973). Applications of x-ray fluorescence to particulate measurements, preprint from National Environmental Research Center, U.S. EPA. Hilst, G. R. (1970). Sensitivities of air quality prediction to input errors and uncertainties, in Proceedings of Symposium on Multiple-Source Urban Diffusion Models, U.S. EPA, APCO Publ. No. AP–86. Lin, C. I., M. Baker, and R. J. Charlson (1973). Absorption co-efficient of atmospheric aerosol: A method for measurement, Appl. Opt. 12, 1356. Munn, R. E. (1973). Secular increases in summer haziness in the Atlantic provinces, Atmosphere 11, 156. Petersen, C. M., H. J. Paulus, and G. A. Foley (1969). The number-size distribution of atmospheric particles during temperature inversions, J. Air Pollut. Control Assoc. 19, 795. Robinson, G. D. (1947). Notes on the measurement and estimation of atmospheric radiation, Quart. J. R. Meteorol. Soc. 73, 127. Robinson, G. D. (1958). Some observations from aircraft of surface albedo and albedo and absorption of clouds, Arch. Meteorol. Geophys. Bioklimatol. B9, 28. Robinson, G. D. (1962). Absorption of solar radiation by atmospheric aerosol, as revealed by measurements at the ground, Arch. Meteorol. Geophys. Bioklimatol. B12, 19. Waldram, J. M. (1945). Measurement of the photometric properties of the upper atmosphere, Quart. J. R. Meteorol. Soc. 71, 319.