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.
The Medical Implications of Nuclear War, Institute of Medicine. @) 1986 by the National Academy of Sciences. National Academy Press, Washington, D.C. Recent Assessments of the Environmental Consequences of Nuclear War RICHARD P. TURCO, PH.D. R&D Associates, Marina de! Rey, California NEW FINDINGS Since late 1984 a number of important scientific studies have considered the global-scale consequences of a major nuclear war. These studies look beyond the immediate and direct effects of nuclear explosions (blast, thermal radiation, and local radioactive fallout) to investigate the more widespread effects of dispersed radioactivity and severe climatic disturb- ances the nuclear winter (Crutzen and Birks, 1982; Turco et al., 1983~. Assessments of the relevant phenomena have been carried out by the U.S. National Research Council (NRC, 1985), the Royal Society of Canada (1985), and the Scientific Committee on Problems of the Environment (SCOPE) of the International Council of Scientific Unions (Pittock et al., 1986, Harwell and Hutchinson, 19851. Numerous physicists, atmospheric scientists, biologists, and physicians from around the world have contrib- uted to these projects. In each case the findings are similar. While cau- tioning that significant uncertainties remain to be resolved, each report concludes that a nuclear winter is a clear possibility following a nuclear exchange. The SCOPE report goes even further, describing the biological, ecological, and human implications of a nuclear war and its aftermath. This unique treatise is summarized in the paper by M. Harwell in this volume (also see Harwell and Hutchinson, 19854. The SCOPE executive summary of findings on the physical and at- mospheric effects of nuclear war is reproduced in the appendix to this paper. The report confirms that the coexistence of immense nuclear ar 96
THE ENVIRONMENTAL CONSEQUENCES OF NUCLEAR WAR 97 senals (about 24,000 warheads carrying some 12,000 megatons of explo- sives) and abundant combustible fuels concentrated in industrial and urban areas (up to several thousand million tons of petroleum and petroleum products and more than ten thousand million tons of wood and wood products) could bring about a global nuclear winter in the event of nuclear exchange. The physical mechanism would involve the ignition of large fires by nuclear bursts, followed by the insertion of unprecedented quan- tities of smoke into the atmosphere, where it could divert sunlight and trigger climatic perturbations. In addition to the major national and international studies mentioned above, a number of important individual scientific contributions to the nuclear winter problem have been made recently. Small and Bush (1985) reassessed the potential quantity of fuel that might burn in nuclear deto- nations over rural (nonurban) targets. They found quantities 10 to 100 times smaller than previous estimates (C~utzen and Birks, 1982; Turco et al., 1983; NRC, 1985~. Although the methodology of the study by Small and Bush is more thorough than earlier approaches, some of the key assumptions have been questioned; a subsequent analysis increases the fuel estimates of Small and Bush by a factor of about 10 (Pittock et al., 1986~. Regardless of the resolution to this problem, wildland fire smoke remains a secondary contributor to a nuclear winter, with industrial and urban smoke being the primary contributor. The heights of deposition of smoke in large fire plumes have been studied by Cotton (1985) and Manins (19851. It now seems clear that smoke from large nuclear-initiated fires would be injected into the upper troposphere, with some directly reaching the stratosphere. Most of the smoke injection would be above the low-altitude zone of normally rapid washout. However, the fraction of the smoke that might be removed immediately in the induced "black rain" is currently unknown; fractions of 30 to 50 percent assumed in recent assessments seem reasonable (NRC, 1985; Pittock et al., 19861. Even if the smoke is not initially injected into the stratosphere, several new studies show that heating produced by the absorption of sunlight will cause the smoke to rise into this region (Haberle et al., 19851. Such studies also indicate that solar heating could lead to the stabilization of elevated smoke layers and accelerate their spread into the Southern Hemisphere. The result would be more widespread and prolonged climatic disturbances. The expected climatic impacts of dense smoke layers include land sur- face temperature decreases of up to 35°C within one week during the summer half of the year, major shifts in global wind patterns, and sub- stantial decreases in precipitation in many continental regions (Thompson, 1985; Cess et al., 1985; Malone et al., 19861. When proper comparisons
98 PHYSICAL EFFECTS AND ENVIRONMENTAL CONSEQUENCES are made between these new results and the original nuclear winter cal- culations of Turco et al. (1983), the predicted temperature declines are found to be very similar.* The latest physical modeling of the radiative and climatic effects of dense clouds of smoke (Cess et al., 1985; Covey et al., 1985; Ramaswamy and Kiehl, 1985) reinforce the cooling mech- anisms postulated for a nuclear winter. In summary, a growing body of detailed technical analyses supports the general principles of the nuclear winter theory. Thus, although sub- stantial work remains to be done, the theory now stands on a much firmer scientific basis than ever before. In light of these newest confirmations of the possible physical aftermath of a major nuclear war, greater emphasis should be placed on understanding the biological and ecological conse- quences, particularly the agricultural and human impacts. These are, after all, of the most urgent concern to the global community (Harwell and Hutchinson, 19851. POTENTIAL CONTRIBUTION OF PLASTICS TO NUCLEAR WINTER Modern civilization is finding more uses for plastics in construction, durable goods, and packaging. When burned, plastics typically generate a sooty toxic smoke (and could therefore contribute to a nuclear winter). The industrialized world is producing plastics and noncellulosic synthetic fibers at the rate of about 60 million metric tons, or teragrams ( 1 teragram [Tg] = 10~2 g), per year (U.N. Statistical Yearbook, 1981; Handbook of Economic Statistics, 19841. In the United States, plastics output is pro- jected to increase by up to 5 to 6 percent per year through this decade (U.S. Industrial Outlook, 19851. Worldwide output has grown roughly 10-fold over the last two decades, and at current rates of growth output could redouble by the turn of the century. In 1983, production in the North Atlantic Treaty Organization (NATO) and Warsaw Pact alliances and Japan amounted to about 46 Tg of plastics and 8 Tg of noncellulosic fibers. Plastics are derived primarily from petroleum and account for about 2 percent of total petroleum consumption. The breakdown of production and usage for various plastics in the United States is given in Table 1. Considering the 70 percent of all plastics represented by the data in Table 1, about one-third goes into each of the packaging and construction industries, about one-quarter into consumer durables, and the rest into the transportation industry. These statistics are * That is, by taking into account the moderating effect of oceans on land temperatures and the normal seasonal differences in initial land temperatures.
THE ENVIRONMENTAL CONSEQUENCES OF NUCLEAR WAR TABLE 1 Plastics Production, Use, and Properties 99 Production/Usea (Tg/yr) Specific Fraction of Extinction Total Coefficient Pack- Construc- Consumer Transpor- Production b of Smoke C Plastics aging tion Durables tation Represented (m2/g-plastic) Polyethylene 2.87 0.77 0.70 0.12 0.70 0.5 Polyvinylchloride 0.30 1.89 0.49 0.10 0.81 1.0 Polystyrene 0.74 0.15 0.48 0 0.63 1.5 Polypropylene 0.42 0.04 0.87 0.16 0.68 0.5 Polyurethane 0 0.18 0.39 0.19 0.86 0.5 Polyester 0.40 0.15 0.33 0.09 0.95 1.0 Phenolic 0 0.99 0.08 0 0.91 0.5 Acrylonitrile- 0 0.09 0.16 0.11 0.69 1.2 butadiene styrene Epoxy/Melamine 0 0.49 0.25 0 0.90 0.3 Total 4.73 4.75 3.75 0.77 0.703 Weighted averagee .' SpeClIlC extinction coefficient of smoke (m2/g-plastic) 0.73 0.78 0.75 0.73 NOTE: Data on production and consumption apply to the United States in 1984 and are taken from Modern Plastics (1985). Only the predominant families of plastics are considered. aPackaging includes containers, wrappers, fillers, bottles, caps, cups, and so on. Construction materials include pipes, panels, insulation, wiring, tiles, moldings, fixtures, and so forth. Con- sumer durables include appliances, electronic gear, housewares, toys, records, tapes, and the like. Transportation refers to plastics used in cars, trucks, buses, aircraft, and so on. bThe fraction of production of each type of plastic that is included in the four use categories (i.e., excluding miscellaneous uses for which statistics are unavailable and exports). CThe specific extinction (scattering plus absorption) coefficients apply at visible wavelengths. The values are based on laboratory data obtained from the references cited in the text, but have a substantial uncertainty. (The coefficient for phenolics is only estimated.) The extinction coef- ficient, when multiplied by the quantity of a particular type of plastic that may be burned, and divided by the area over which the smoke may be dispersed, provides an estimate of the extinction optical depth of the resulting smoke layer. This is the fraction of the total production of plastics (all types and all uses) represented by the data in the table. According to Modern Plastics (1985), total U.S. plastics consumption in 1984 was nearly 20 Tg. eThe average extinction coefficient corresponding to the specific mix of plastics in each use category.
100 PHYSICAL EFFECTS AND ENVIRONMENTAL CONSEQUENCES consistent with other surveys of plastics consumption (U.S. Statistical Abstracts, 19851. There are two lifetimes of plastics that are of interest: their functional lifetime from manufacture to disposal as trash and their lifetime in the environment after disposal. It is assumed that plastics used in packaging (about 33 percent of the total produced, or 16 Tg/year at present) have an average functional lifetime of 3 months. Thus, relatively little packaging material is stockpiled at any time. Of the remaining plastics, it is assumed that equal thirds have functional lifetimes of 10, 20, and 40 years (i.e., about 22 percent of total production, or 10 Tg/year of each type at present). The longer-lived materials would be used mainly in construction as well as in certain items of furniture and appliances. In addition, poly- meric fibers, used principally in carpeting, are assigned a lifetime of 10 years (8 Tg/year is produced at present). While the useful lifetimes of plastics in society are difficult to determine accurately, our estimates based on broad use categories should be adequate. In disposal, plastics may be burned, buried, or simply discarded as loose rubbish. Only about 10 percent of all collected municipal waste is burned, and thus incineration represents a minor sink for plastics (Guillet. 19731. No meaningful recycling of plastics occurs. Most polymers are also immune to decomposition in soils. However, solar ultraviolet radia- tion causes bond incisions in many polymers, leading to chemical activity and degradation, although continuous direct exposure to sunlight is re- quired for effective decomposition (Guillet, 1973~. Today, discarded plas- tics continue to accumulate in dumps, salvage yards, and landfills. Some of this material would be accessible to burning in a nuclear war. Production records indicate that about 750 Tg of plastics and noncel- lulosic fibers were manufactured from 1963 to 1983 in the United States, Europe, the USSR, and Japan. With the assumed lifetimes discussed above, it is estimated that about 400 Tg of polymers are in use today. Given the production rates and lifetimes of various plastics products and the amounts available in 1983, estimates of future accumulations can be made (Figure 11. If production were to remain constant at the 1983 level, the quantity of plastics in use would continue to grow, approaching 700 Tg in several decades. If production were to increase by 5 percent per year until the turn of the century (as might be anticipated based on past perfo~ance; U.S. Industrial Outlook, 1985), about 800 Tg could be in use in the NATO and Warsaw Pact alliances by the turn of the century, with the quantities in use in these countries continuing to grow well into the century. Greater amounts would have been produced and discarded, and much of this would be accessible at trash disposal sites. More than 1,600 Tg of plastics and fibers will have been manufactured in the countries of interest by the turn of the century.
THE ENVIRONMENTAL CONSEQUENCES OF NUCLEAR WAR 4000 ' ' ' 1 ' ' ' ' 1 ' ' ' ' 1 ' 1' ' ' 1 ' ' ' ' // ' ~ b , ~ In m 3000 ILL cam o J z o 2000 - 101 / PRODUCED - - I N USE - O. 1983 1993 2003 2013 2023 2033 YEAR FIGURE 1 Projections of the total production and accumulation of plastics and noncellulosic fibers in the United States, Europe, the U.S.S.R., and Japan. The calculations correspond to the following assumptions about future plastics pro- duction: (a) production is constant at the 1983 rate for an indefinite period, and (b) production grows annually by 5 percent of the 1983 rate from 1983 to 2003 and is constant thereafter. The assumed plastics lifetimes are discussed in the text. The smoke produced by burning polymers has been studied in the laboratory (Tewarson, 1982; Seader and Einhorn, 1976; Morikawa, 1978; Hilado and Machado, 1978; Quintiere, 1982; Bankston et al., 1981~. For the materials in Table 1, measured smoke emission factors (grams of smoke per gram of material burned) range from several percent to 20 percent or more. Under smoldering conditions, as much as 50 percent of some plastics may be converted into smoke particles. Under flaming con- ditions, the smoke is typically black and sooty, with a large optical ab- sorption coefficient. Table 1 summarizes data on the specific extinction coefficient of the smoke generated by a variety of plastics during flaming combustion (the specific extinction coefficient is expressed in terms of the square meters of cross section produced by each gram of material burned). The values for individual plastics differ by a factor of 5 or more. However, the average weighted coefficients corresponding to the four principal use categories in Table 1 are very similar. On this basis, a specific extinction coefficient
102 PHYSICAL EFFECTS AND ENVIRONMENTAL CONSEQUENCES of 0.75 m2/g-plastics burned may be employed to make estimates of optical effects. Plastics can be treated with flame retardants and smoke retardants. In the United States in 1984, 200,000 tons of flame retardants were used (about 1 percent of the weight of the plastics produced) (Modern Plastics, 1984~. Flame retardants do not generally reduce the smoke emissions of plastics and can increase emissions dramatically in some cases (Bankston et al., 1981~. Practical and economical smoke retardants for plastics have not yet been developed. Such retardants may eventually reduce the smoke emissions of certain plastics by up to 50 percent (Modern Plastics, 1984) but will never eliminate the smoke altogether. Accumulated plastics are concentrated in urban zones where, because of their variety, utility, and durability, they find countless practical uses. In the United States, the top 277 metropolitan areas hold about 75 percent of the population (U.S. Statistical Abstracts, 1985) and, presumably, at least this fraction of all plastics. The bunting of hundreds of teragrarns of plastics would produce enough sooty smoke to contribute to a nuclear winter, perhaps even leading to one in the absence of other sources of smoke and dust. Of course, a substantial portion of the urban areas of the warring nations would have to be ignited in fires, as has been postulated in a number of recent studies (Crutzen and Barks, 1982; Turco et al. 1983; NRC, 1985; Pittock et al., 1986~. Table 2 defines a plausible range of optical perturbations that might result from plastics combustion in a future nuclear war. The range of potential global-average absorption optical depths resulting from plastics smoke accounting for a plausible range of parameter variations is about 0.1 to 1.2. It has been shown that smoke injections leading to equivalent global absorption optical depths of ~0.2 could greatly perturb the climate (Thompson, 1985; Cess et al., 1985; Malone et al., 1986~. The baseline scenario of Turco et al. (1983) projected an equivalent global absorption TABLE 2 Projected Smoke Potential of Plastics in 2000 A.D. Mass of Plastics (Tg) Total Extinction Cross-section (1012 m2) Fraction Activated by Firesa Equivalent Global Extinction Optical Depthb Global Absorption Optical DepthC TTAPS Baseline Urban Smoke Absorption Optical Depth (global) 1,000-2,000 750- 1 ,500 118-1/2 0.2-1.5 0.1-1.2 ~0.5 aThis factor assumes that one-fourth to two-thirds of all plastics in the warring nations could be burned in a major nuclear war and that one-fourth to one-half of the smoke would be removed immediately by precipitation (Pittock et al., 1986). bThis assumes that the smoke is uniformly distributed around the entire planet. CThis corresponds to a range of 0.2 to 0.5 for the single-scatter albedo of the smoke.
THE ENVIRONMENTAL CONSEQUENCES OF NUCLEAR WAR 103 optical depth of about 0.5, which produced a nuclear winter. Accordingly, smoke and toxic gases from plastics alone would seem to guarantee some nuclear winter-like environmental effects in the event of a future nuclear conflict. Additional sources of climate-altenng aerosols would be soot from oil and structure fires, smoke from wildfires, and dust from surface detonations. The estimates presented here are uncertain because, among other things, extrapolations of trends in plastics consumption have been extended sev- eral decades into the future. However, the drift toward greater worldwide dependence on plastics and synthetic fibers is very clear. Nuclear arsenals may be reduced substantially in the coming decades, but there is currently little reason for optimism. Thus, in a politically divided world, we must seriously consider how advanced technologies may someday interact to threaten the future of mankind. REFERENCES Bankston, C. P., B. T. Zinn, R. F. Browner, and E. A. Powell. 1981. Aspects of the mechanisms of smoke generation by burning materials. Comb. Flame 41:273-292. Cess, R. D., G. L. Potter, S. J. Ghan, and W. L. Gates. 1985. The climatic effects of large injections of atmospheric smoke and dust. J. Geophys. Res. 90:12937-12950. Cotton, W. R. 1985. Atmospheric convection and nuclear winter. Am. Sci. 73:275-280. Covey, C., S. L. Thompson, and S. H. Schneider. 1985. "Nuclear winter": A diagnosis of atmospheric general circulation model simulations. J. Geophys. Res. 90:5615-5628. Crutzen, P. J., and J. W. Birks. 1982. Twilight at noon: The atmosphere after a nuclear war. Ambio 11:114-125. Guillet, J. (ed.). 1973. Polymers and ecological problems. Polymer Science and Tech- nology, Volume 3. New York: Plenum. Haberle, R. M., T. P. Ackerman, O. B. Toon, and J. L. Hollingsworth. 1985. Global transport of atmospheric smoke following a major nuclear exchange. Geophys. Res. Lett. 12:405-408. Handbook of Economic Statistics, 1984. 1984. Washington, D.C.: U.S. Government Print- ing Office. Harwell, M. A., and T. C. Hutchinson. 1985. Environmental Consequences of Nuclear War, Volume II: Ecological and Agricultural Effects. SCOPE 28. Chichester, U.K.: John Wiley & Sons. Hilado, C. J., and A. M. Machado. 1978. Smoke studies with the Arapahoe chamber. J. Fire Flammability 9:240-244. Malone, R. C., L. H. Auer, G. A. Glatzmaier, M. C. Wood, and O. B. Toon. 1986. Nuclear winter: Three-dimensional simulations including interactive transport, scaveng- ing and solar heating of smoke. J. Geophys. Res. 91:1039-1053. Manins, P. C. 1985. Cloud heights and stratospheric injections resulting from a thermo- nuclear war. Atmos. Environ. 19: 1245- 1255. Modern Plastics. 1984. 61:62-64. Modern Plastics. 1985. Materials '85. 62:61-71. Morikawa, T. 1978. Evolution of soot and polycyclic aromatic hydrocarbons in combustion. J. Comb. Toxicol. 5:349-360.
104 PHYSICALEFFECTS AND ENVIRONMENTAL CONSEQUENCES National Research Council (NRC). 1985. The effects on the atmosphere of a major nuclear exchange. Washington, D.C.: National Academy Press. Pittock, A. B., T. P. Ackerman, P. J. Crutzen, M. C. MacCracken, C. S. Shapiro, and R. P. Turco. 1986. Environmental Consequences of Nuclear War, Volume I: Physical and Atmospheric Effects. SCOPE 28. Chichester, U.K.: John Wiley & Sons. Quintiere, J. G. 1982. An assessment of correlations between laboratory and full-scale experiments for the FAA aircraft fire safety program. Part 1: Smoke. National Bureau of Standards Report NB SIR 82-2508. Washington, D.C.: National Bureau of Standards. Ramaswamy, V., and J. Kiehl. 1985. Sensitivities of the radiative forcing due to large loadings of smoke and dust. J. Geophys. Res. 90:5597-5613. Royal Society of Canada. 1985. Nuclear winter and associated effects: A Canadian appraisal of the environmental impact of nuclear war. Ottawa, Canada: Royal Society of Canada. Seader, J. D., and I. N. Einhon~. 1976. Some physical, chemical, toxicological and phys- iological aspects of fire smokes. Pp. 1423-1445 in Proceedings of the 16th International Symposium of the Combustion Institute, Pittsburgh, Pa. Small, R. D., and B. W. Bush. 1985. Smoke production from multiple nuclear explosions in nonurban areas. Science 229:465-469. Tewarson, A. 1982. Experimental evaluation of flammability parameters of polymeric materials. Pp. 97-153 in Flame Retardant Polymeric Materials, Vol. 3., M. Lewin, S. M. Atlas, and E. M. Pierce, eds. New York: Plenum. Thompson, S. L. 1985. Global interactive transport simulations of nuclear war smoke. Nature 317:35-39. Turco, R. P., O. B. Toon, T. P. Ackerman, J. B. Pollack, and C. Sagan. 1983. Nuclear winter: Global consequences of multiple nuclear explosions. Science 222:1283-1292. U.N. Statistical Yearbook, 1979/1980. 1981. New York: United Nations. U.S. Industrial Outlook, 1985. 1985. Washington, D.C.: Department of Commerce. U.S. Statistical Abstracts 1984. 1985. Washington, I).C.: Department of Commerce. APPENDIX SCOPE/ENUWAR Executive Summary The executive summary of the SCOPE (Scientific Committee on Prob- lems of the Environment) ENUWAR (Environmental Effects of Nuclear War) Report, Volume I: Physical and Atmospheric Effects (Pittock et al., 1986), reviews the most important findings of recent detailed scientific studies of the nuclear winter phenomenon (SCOPE 28, Volumes I and II, John Wiley & Sons, Ltd., Chichester-available also from Wiley, Inc. for North America). The summary is reproduced here in its entirety by permission of SCOPE. Executive Summary This volume presents the results of an assessment of the climatic and atmospheric effects of a large nuclear war. The chapters in the volume follow a logical sequence of development, starting with discussions of nuclear weapons effects and possible characteristics of a nuclear war. The report continues with a treatment of the consequent fires, smoke emissions, and dust injections and their effects on the physical and chemical processes
THE ENVIRONMENTAL CONSEQUENCES OF NUCLEAR WAR 105 of the atmosphere. This is followed by a chapter dealing with long-term radiological doses. The concluding chapter contains recommendations for future research and study. In assessments of this type, a variety of procedural options are available, including, for example, "worst case" analyses, risk analyses, and "most probable" analyses. All of these approaches have relevance for the subject addressed here due to the large uncertainties which surround many aspects of the problem. Some of these uncertainties are inherent in studies of nuclear war and some are simply the result of limited information about natural physical processes. In general, in making assumptions about scen- arios, models, and magnitudes of injections, and in estimating their at- mospheric effects, an attempt has been made to avoid "minimum" and "worst case" analyses in favor of a "middle ground" that encompasses, with reasonable probability, the atmospheric and climatic consequences of a major nuclear exchange. The principal results of this assessment, arranged roughly in the same order as the more detailed discussions contained in the body of this volume, are summarized below. The Executive Summary of Volume II (Harwell and Hutchinson, 1985), which describes the ecological and agricultural consequences of a nuclear war, is included as Appendix 1 at the end of this volume. A Glossary is included as Appendix 2 and a list of participants in the study is included as Appendix 3. 1. DIRECT EFFECTS OF NUCLEAR EXPLOSIONS The two comparatively small detonations of nuclear weapons in Japan in 1945 and the subsequent higher yield atmospheric nuclear tests pre- ceding the atmospheric test ban treaty of 1963 have provided some in- formation on the direct effects of nuclear explosions. Typical modern weapons carried by today's missiles and aircraft have yields of hundreds ~ . . . . · . . . .. ~ . Of kilotons or more. ll detonated, such explosions WOU1O nave the ~o~- lowing effects: · In each explosion, thermal (heat) radiation and blast waves would result in death and devastation over an area of up to 500 km2 per megaton of yield, an area typical of a major city. The extent of these direct effects depends on the yield of the explosion, height of burst, and state of the local environment. The destruction of Hiroshima and Nagasaki by atomic bombs near the end of World War II provides examples of the effects of relatively small nuclear explosions. · Nuclear weapons are extremely efficient incendiary devices. The thermal radiation emitted by the nuclear fireball, in combination with the accidental ignitions caused by the blast, would ignite fires in urban/in- dustrial areas and wildlands of a size unprecedented in history. These fires
106 PHYSICAL EFFECTS AND ENVIRONMENTAL CONSEQUENCES would generate massive plumes of smoke and toxic chemicals. The newly recognized atmospheric effects of the smoke from a large number of such fires are the major focus of this report. ~ For nuclear explosions that contact land surfaces (surface bursts), large amounts (of the order of 100,000 tonne per megaton of yield) of dust, soil, and debris are drawn up with the fireball. The larger dust particles, carrying about half of the bomb's radioactivity, fall back to the surface mostly within the first day, thereby contaminating hundreds of square kilometers near and downwind of the explosion site. This local fallout can exceed the lethal dose level. · All of the radioactivity from nuclear explosions well above the surface (airbursts) and about half of the radioactivity from surface bursts would be lofted on particles into the upper troposphere or stratosphere by the rising fireballs and contribute to longer term radioactive fallout on a global scale. · Nuclear explosions high in the atmosphere, or in space, would gen- erate an intense electromagnetic pulse capable of inducing strong electric currents that could damage electronic equipment and communications networks over continent-size regions. 2. STRATEGIES AND SCENARIOS FOR A NUCLEAR WAR In the forty years since the first nuclear explosion, the five nuclear powers, but primarily the U.S. and the U.S.S.R., have accumulated very large arsenals of nuclear weapons. It is impossible to forecast in detail the evolution of potential military conflicts. Nevertheless, enough of the general principles of strategic planning have been discussed that plausible scenarios for the development and immediate consequences of a large- scale nuclear war can be derived for analysis. ~ NATO and Warsaw Pact nuclear arsenals include about 24,000 stra- tegic and theatre nuclear warheads totaling about 12,000 megatons. The arsenals now contain the equivalent explosive power of about one million "Hiroshima-size" bombs. · A plausible scenario for a global nuclear war could involve on the order of 6000 Mt divided between more than 12,000 warheads. Because of its obvious importance, the potential environmental consequences of an exchange of roughly this size are examined. The smoke-induced at- mospheric consequences discussed in this volume are, however, more dependent on the number of nuclear explosions occurring over cities and industrial centers than on any of the other assumptions of the particular exchange.
THE ENVIRONMENTAL CONSEQUENCES OF NUCLEAR WAR 107 · Many targets of nuclear warheads, such as missile silos and some military bases, are isolated geographically from population centers. Never- theless, enough important military and strategic targets are located near or within cities so that collateral damage in urban and industrial centers from a counterforce nuclear strike could be extensive. As a result, even relatively limited nuclear attacks directed at military-related targets could cause large fires and smoke production. · Current strategic deterrence policies imply that, in an escalating nu- clear conflict, many warheads might be used directly against urban and industrial centers. Such targeting would have far-reaching implications because of the potential for fires, smoke production, and climatic change. 3. THE EXTENT OF FIRES AND GENERATION OF SMOKE During World War II, intense city fires covering areas as large as 10 to 30 square kilometers were ignited by massive incendiary bombing raids, as well as by the relatively small nuclear explosions over Hiroshima and Nagasaki. Because these fires were distributed over many months, the total atmospheric accumulation of smoke generated by these fires was small. Today, in a major nuclear conflict, thousands of very intense fires, each covering up to a few hundred kilometers, could be ignited simul- taneously in urban areas, fossil fuel processing plants and storage depots, wildlands, and other locations. Because there have never been fires as large and as intense as may be expected, no appropriate smoke emission measurements have been made. Estimates of emissions from such fires rely upon extrapolation from data on much smaller fires. This procedure may introduce considerable error in quantifying smoke emissions, espe- cially in making estimates for intense fire situations. · About 70% of the populations of Europe, North America and the Soviet Union live in urban and suburban areas covering a few hundred thousand square kilometers and containing more than ten thousand million tonne of combustible wood and paper. If about 25-30% of this were to be ignited, in just a few hours or days, tens of millions to more than a hundred million tonne of smoke could be generated. About a quarter to a third of the emitted smoke from the flaming combustion of this material would be amorphous elemental carbon, which is black and efficiently absorbs sunlight. · Fossil fuels (e.g., oil, gasoline, and kerosene) and fossil fuel-derived products (including plastics, rubber, asphalt, roofing materials, and or- ganochemicals) are heavily concentrated in cities and industrial areas; flaming combustion of a small fraction (~25-30%) of the few thousand
108 PHYSICAL EFFECTS AND ENVIRONMENTAL CONSEQUENCES million tonne of such materials currently available could generate 50-150 million tonne of very sooty smoke containing a large fraction (50% or greater) of amorphous elemental carbon. The burning of 25-30% of the combustible materials of the developed world could occur with near total burnout of less than one hundred of the largest industrialized urban areas. · Fires ignited in forests and other wildlands could consume tens to hundreds of thousands of square kilometers of vegetation over days to weeks, depending on the state of the vegetation, and the extent of fire- spread. These fires could produce tens of millions of tonne of smoke in the summer half of the year, but considerably less in the winter half of the year. Because wildland fire smoke contains only about 10% amorphous elemental carbon, it would be of secondary importance compared to the smoke created by urban and industrial fires, although its effects would not be negligible. · The several tens of millions of tonne of sub-micron dust particles that could be lofted to stratospheric altitudes by surface bursts could reside in the atmosphere for a year or more. The potential climatic effects of the dust emissions, although substantially less than those of the smoke, also must be considered. 4. THE EVOLUTION AND RADIATIVE EFFECTS OF THE SMOKE The sooty smoke particles rising in the hot plumes of large fires would consist of a mixture of amorphous elemental carbon, condensed hydro- carbons, debris particles, and other substances. The amount of elemental carbon in particles with effective spherical diameters on the order of 0.1 ,um to perhaps 1.0 Am would be of most importance in calculating the potential effect on solar radiation. Such particles can be spread globally by the winds and remain suspended for days to months. · Large hot fires create converging surface winds and rapidly rising fire plumes which, within minutes, can carry smoke particles, ash and other fire products, windblown debris, and water from combustion and the surrounding air to as high as 10-15 kilometers. The mass of particles deposited aloft would depend on the rate of smoke generation, the intensity of the fire, local weather conditions, and the effectiveness of scavenging processes in the convective column. · As smoke-laden, heated air from over the fire rises, adiabatic ex- pansion and entrainment would cause cooling and condensation of water vapor that could lead, in some cases, to the formation of a cumulonimbus cloud system. Condensation-induced latent heating of the rising air parcels
THE ENVIRONMENTAL CONSEQUENCES OF NUCLEAR WAR 109 would help to loft the smoke particles to higher altitudes than expected from the heat of combustion alone. ~ Although much of the water vapor drawn up from the boundary layer would condense, precipitation might form for only a fraction of the fire plumes. In the rising fire columns of such fires, soot particles would tend to be collected inefficiently by the water in the cloud. Smoke particles however, are generally composed of a mixture of substances and might, at least partially, be incorporated in water droplets or ice particles by processes not now well understood. Smoke particles that are captured could again be released to the atmosphere as the ice or water particles evaporate in the cloud anvils or in the environment surrounding the con- vective clouds. Altogether, an unknown fraction of the smoke entering the cloud would be captured in droplets and promptly removed from the atmosphere by precipitation. · Not all fires would, however, induce strong convective activity. This depends on fuel loading characteristics and meteorological conditions. It is assumed in current studies that 30-50% of the smoke injected into the atmosphere from all fires would be removed by precipitation within the first day, and not be available to affect longer-term large-scale, meteor- ological processes. This assumption is a major uncertainty in all current assessments. For the fire and smoke assumptions made in this study, the net input of smoke to the atmosphere after early scavenging is estimated to range from 50 to 150 million tonne, containing about 30 million tonne of amorphous elemental carbon. · Smoke particles generated by urban and fossil fuel fires would be strong absorbers of solar radiation, but would be likely to have compar- atively limited effects on terrestrial longwave radiation, except perhaps under some special circumstances. If 30 million tonne of amorphous el- emental carbon were produced by urban/industrial fires and spread over Northern Hemisphere mid-latitudes, the insolation at the ground would be reduced by at least 90%. The larger quantities of smoke that are possible in a major nuclear exchange could reduce light levels under dense patches to less than 1% of normal, and, on a daily average, to just several percent of normal, even after the smoke has spread widely. ~ Because of the large numbers of particles in the rising smoke plumes and the very dense patches of smoke lasting several days thereafter, co- agulation (adhering collisions) would lead to formation of fewer, but somewhat larger, particles. Coagulation of the particles could also occur as a result of coalescence and subsequent evaporation of rain droplets or ice particles. Because optical properties of aerosols are dependent on particle size and morphology, the aggregated aerosols may have different optical properties than the initial smoke particles, but the details, and even
110 PHYSICAL EFFECTS AND ENVIRONMENTAL CONSEQUENCES the sign, of such changes are poorly understood. The optical properties of fluffy soot aggregates that may be formed in dense oil plumes, however, seem to be relatively insensitive to their size. This is not the case for more consolidated particle agglomerates. · Little consideration has yet been given to the possible role of me- teorological processes on domains between fire plume and continental scales. Mesoscale and synoptic-scale motions might significantly alter, mix, or remove the smoke particles during the first several days. Studies to examine quantitatively the microphysical evolution of smoke particles during this period are needed. While changes in detailed understanding are expected, a significant fraction of the injected smoke particles is likely to remain in the atmosphere and affect the large-scale weather and climate. 5. SMOKE-INDUCED ATMOSPHERIC PERTURBATIONS In a major nuclear war, continental scale smoke clouds could be gen- erated within a few days over North America, Europe, and much of Asia. Careful analysis and a hierarchy of numerical models (ranging from one- dimensional gIobal-average to ~ree-dimensional gIobal-scale models) have been used to estimate the transport, transformation, and removal of the smoke particles and the effects of the smoke on temperature, precipitation, winds, and other important atmospheric properties. All of the simulations indicate a strong potential for large-scale weather disruptions as a result of the smoke injected by extensive post-nuclear fires. These models, however, still have important simplifications and uncertainties that may affect the fidelity and the details of their predictions. Nonetheless, these uncertainties probably do not affect the general character of the calculated atmospheric response. · For large smoke injections reaching altitudes of several kilometers or more and occurring from spring through early fall in the Northern Hemisphere, average land surface temperatures beneath dense smoke patches could decrease by 20-40°C below normal in continental areas within a few days, depending on the duration of the dense smoke pall and the particular meteorological state of the atmosphere. Some of these patches could be carried long distances and create episodic cooling. During this initial period of smoke dispersion, anomalies could be spatially and tem- porally quite variable while patchy smoke clouds strongly modulate the insolation reaching the surface. · Smoke particles would be spread throughout much of the Northern Hemisphere within a few weeks, although the smoke layer would still be
THE ENVIRONMENTAL CONSEQUENCES OF NUCLEAR WAR 111 far from homogeneous. For spring to early fall injections, solar heating of the particles could rapidly warm the smoke layer and lead to a net upward motion of a substantial fraction of the smoke into the upper tro- posphere and stratosphere. The warming of these elevated layers could stabilize the atmosphere and suppress vertical movement of the air below these layers, thereby extending the lifetime of the particles from days to perhaps several months or more. · Average summertime land surface temperatures in the Northern Hemisphere mid-latitudes could drop to levels typical of fall or early winter for periods of weeks or more with convective precipitation being essen- tially eliminated, except possibly at the southern edge of the smoke pall. Cold, near-surface air layers might lead initially to fog and drizzle, es- pecially in coastal regions, lowland areas, and river valleys. In continental interiors, periods of very cold, mid-winter-like temperatures are possible. In winter, light levels would be strongly reduced, but the initial temper- ature and precipitation perturbations would be much less pronounced and might be essentially indistinguishable in many areas from severe winters currently experienced from time to time. However, such conditions would occur simultaneously over a large fraction of the mid-latitude region of the Northern Hemisphere and freezing cold air outbreaks could penetrate southward into regions that rarely or never experience frost conditions. · In Northern Hemisphere subtropical latitudes, temperatures in any season could drop well below typical cool season conditions for large smoke injections. Temperatures could be near or below freezing in regions where temperatures are not typically strongly moderated by warming in- fluence from the oceans. The connectively driven monsoon circulation, which is of critical importance to subtropical ecosystems, agriculture, and is the main source of water in these regions, could be essentially elimi- nated. Smaller scale, coastal precipitation might, however, be initiated. · Strong solar heating of smoke injected into the Northern Hemisphere between April and September would carry the smoke upwards and equa- torward, strongly augmenting the normal high altitude flow to the Southern Hemisphere (where induced downward motions might tend to slightly suppress precipitation). Within one or two weeks, thin, extended smoke layers could appear in the low to mid-latitude regions of the Southern Hemisphere as a precursor to the development of a more uniform veil of smoke with a significant optical depth (although substantially smaller than in the Northern Hemisphere). The smoke could induce modest cooling of land areas not well buffered by air masses warmed over nearby ocean areas. Since mid-latitudes in the Southern Hemisphere would already be experiencing their cool season, temperature reductions would not likely
112 PHYSICAL EFFECTS AND ENVIRONMENTAL CONSEQUENCES be more than several degrees. In more severe, but less probable, smoke injection scenarios, climatic effects in the Southern Hemisphere could be enhanced significantly, particularly during the following austral spring and summer. · Much less analysis has been made of the atmospheric perturbations following the several week, acute climatic phase subsequent to a nuclear war involving large smoke injections. Significant uncertainties remain concerning processes governing the longer-term removal of smoke par- ticles by precipitation scavenging, chemical oxidation, and other physical and chemical factors. The ultimate fate of smoke particles in the perturbed atmospheric circulation is also uncertain, both for particles in the sunlit and stabilized upper troposphere and stratosphere and in the winter polar regions, where cooling could result in subsidence that could move particles downward from the stratosphere to altitudes where they could later be scavenged by precipitation. O Present estimates suggest that smoke lofted to levels (either directly by fire plumes or under the influence of solar heating) which are, or become, stabilized, could remain in the atmosphere for a year or more and induce long-term (months to years) global-scale cooling of several degrees, especially after the oceans have cooled significantly. For such conditions, precipitation could also be reduced significantly. Reduction of the intensity of the summer monsoon over Asia and Africa could be a particular concern. Decreased ocean temperatures, climatic feedback mechanisms (e.g., ice-albedo feedback), and concurrent ecological changes could also prolong the period of meteorological disturbances. 6. ATMOSPHERIC CHEMISTRY IN A POST-NUCLEAR-WAR ENVIRONMENT Nuclear explosions and the resultant fires could generate large quantities of chemical compounds that might themselves be toxic. In addition, the chemicals could alter the atmospheric composition and radiative fluxes in ways that could affect human health, the biosphere, and the climate. ~ Nitrogen oxides (NOx) created by nuclear explosions of greater than several hundred kilotons would be lofted into the stratosphere. Depending on the total number of high yield weapons exploded, the NOX would catalyze chemical reactions that, within a few months time, could reduce Northern Hemisphere stratospheric ozone concentrations by 10 to 30% in an atmosphere free of aerosols. Recovery would take several years. How
THE ENVIRONMENTAL CONSEQUENCES OF NUCLEAR WAR 113 ever, if the atmosphere were highly perturbed due to smoke heating and by injection of gaseous products from fires, the long-term ozone changes could be enhanced substantially in ways that cannot yet be predicted. · Ozone reductions of tens of percent could increase surface intensities of biologically-active ultraviolet (UV) radiation by percentages of up to a few times as much. The presence of smoke would initially prevent UV- radiation from reaching the surface by absorbing it. The smoke, however, might also prolong and further augment the long-term ozone reduction as a result of smoke-induced lofting of soot and reactive chemicals, conse- quent heating of the stratosphere, and the occurrence of additional chemical reactions. · Large amounts of carbon monoxide, hydrocarbons, nitrogen and sul- fur oxides, hydrochloric acid, pyrotoxins, heavy metals, asbestos, and other materials would be injected into the lower atmosphere near the surface by flaming and smoldering combustion of several thousand million tonne of cellulosic and fossil fuel products and wind-blown debris. Before deposition or removal, these substances, some of which are toxic, could be directly and/or indirectly harmful to many forms of life. In addition, numerous toxic chemical compounds could be released directly into the environment by blast and spillage, contaminating both soil and water. This complex and potentially very serious subject has so far received only cursory consideration. · If the hydrocarbons and nitrogen oxides were injected into an oth- erwise unperturbed troposphere, they could enhance average background ozone concentrations several-fold. Such ozone increases would not sig- nif~cantly offset the stratospheric ozone decrease, which also would be longer lasting. It is highly questionable, however, whether such large ozone increases could indeed occur in the presence of smoke because ozone generation in the troposphere requires sunlight as well as oxides of nitrogen. It is possible that, in the smoke perturbed atmosphere, the fire- generated oxides of nitrogen could be removed before photochemical ozone production could take place. · Precipitation scavenging of nitrogen, sulfur, and chlorine compounds dispersed by the fire plumes throughout the troposphere could increase rainfall acidity by about an order of magnitude over large regions for up to several months. This increased acidity could be neutralized to some degree by alkaline dust or other basic (as opposed to acidic) compounds. · Rapid smoke-induced cooling of the surface under dense smoke clouds could induce the formation of shallow, stable cold layers that might trap chemical emissions from prolonged smoldering fires near the ground. In such layers, concentrations of CO, HC1, pyrotoxins, and acid fogs could
114 PHYSICAL EFFECTS AND EN1fIRONMENTAL CONSEQUENCES reach dangerous levels. The potential for local and regional effects in areas such as populated lowland areas and river valleys merits close attention. 7. RADIOLOGICAL DOSE Near the site of an explosion, the health effects of prompt ionizing radiation from strategic nuclear warheads would be overshadowed by the effects of the blast and thermal radiation. However, because nuclear ex- plosions create highly radioactive fission products and the emitted neutrons may also induce radioactivity in initially inert material near the detonation, radiological doses would be delivered to survivors both just downwind (local fallout) and out to hemispheric and global scales (global fallout). · Local fallout of relatively large radioactive particles lofted by the number of surface explosions in the scenario postulated in this study could lead to lethal external gamma-ray doses (assuming no protective action is taken) during the first few days over about 7 percent of the land areas of the NATO and Warsaw Pact countries. Areas downwind of missile silos and other hardened targets would suffer especially high exposures. Sur- vivors outside of lethal fallout zones could still receive debilitating radia- tion doses (exposure at half the lethal level can induce severe radiation sickness). In combination with other injures or stresses, such doses could increase mortality. If large populations could be mobilized to move from highly radioactive zones or take substantial protective measures, the human impact of fallout could be greatly reduced. · The uncertainty in these calculations of local fallout is large. Doses and areas for single nuclear explosions could vary by factors of 2-4 depending on meteorological conditions and assumptions in the models. A detailed treatment of overlapping fallout plumes from multiple explo- sions could increase the areas considerably (by a factor of 3 in one sample case). Results are also sensitive to variations in the detonation scenario. · Global fallout following the gradual deposition of the relatively small radioactive particles created by strategic air and surface bursts could lead to average Northern Hemisphere lifetime external gamma ray doses on the order of 10 to 20 reds. The peak values would lie in the northern mid- latitudes where the average doses for the scenarios considered would be about 20 to 60 reds. Such doses, in the absence of other stresses, would be expected to have relatively minor carcinogenic and mutagenic effects (i.e., increase incidence at most a few percent above current levels). Smoke-induced perturbations that tend to stabilize the atmosphere and slow deposition of radioactive particles might reduce these estimated av- erage doses by perhaps 15%.
THE ENVIRONMENTaL CONSEQUENCES OF NUCLEAR WAR 115 · Intermediate time scale and long term global fallout would be de- posited unevenly, largely because of meteorological effects, leading to "hotshots" of several hundred thousand square kilometers in which av- erage doses could be as high as 100 reds, and, consequently, large areas where doses would be lower than the average value. · In the Southern Hemisphere and tropical latitudes, global fallout would produce much smaller, relatively insignificant, radiological doses about one-twentieth those in the Northern Hemisphere, even if cross- equatorial transport were accelerated by the smoke clouds. Additional local fallout would be important only within a few hundred kilometers downwind of any surface burst in the Southern Hemisphere. · Additional considerations not factored into the above estimates are possible from several sources. Doses from ingestion or inhalation of ra- dioactive particles could be important, especially over the longer term. Beta radiation could have a significant effect on the biota coming into contact with the local fallout. Fission fractions of smaller modern weapons could be twice the assumed value of 0.5; adding these to the scenario mix could cause a 20% increase in areas of lethal fallout. General tactical and theater nuclear weapons, ignored in these calculations, could also cause a 20% increase in lethal local fallout areas in certain geographical regions, particularly in Europe. The injection into the atmosphere of radionuclides created and stored by the civilian nuclear power industry and military reactors, a possibility considered remote by some, could increase estimates of long-term local and global radiological doses to several times those estimated for weapons alone. 8. TASKS FOR THE FUTURE Extensive research and careful assessment over the past few years have indicated that nuclear war has the potential to modify the physical envi- ronment in ways that would dramatically impair biological processes. The perturbations could impact agriculture, the proper functioning of natural ecosystems, the purity of essential air and water resources, and other important elements of the global biosphere. Because current scientific conclusions concerning the response of the atmosphere to the effects of nuclear war include uncertainties, research can and should be undertaken to reduce those uncertainties that are accessible to investigation. ~ Laboratory and field experiments are needed to improve estimates of the amount and physical characteristics of the smoke particles that would be produced by large fires, particularly by the combustion of fossil fuels and fossil fuel-derived products present in urban and industrial regions. . .
116 PHYSICAL EFFECTS AND ENVIRONMENTAL CONSEQUENCES Experimental conditions should be designed to emulate as much as possible the effects of large-scale fires. · Laboratory, field, and theoretical studies are needed to determine the potential scavenging rates of smoke particles in the convective plumes of large fires and the scavenging processes that operate on intermediate and global scales as the particles disperse. · Further theoretical calculations of the seasonal response of the at- mosphere to smoke emissions from large fires are needed, particularly of the extent of the perturbation to be expected at early times, when the smoke is freshly injected and patchy. Simulations must be made for later times from months to a year or more, when the atmosphere has been highly perturbed and a substantial fraction of the smoke may have been lofted to high altitudes. Closer attention should be paid to He possible effects in low latitudes and in the Southern Hemisphere, where the climatic effects are likely to be much more important than the direct effects of the nuclear detonations, which are expected to be confined largely to the Northern Hemisphere. ~ · Laboratory and theoretical studies are needed of the potential chem- ical alterations of the atmosphere on global and local scales, and of the extent that smoke particles could affect and might be removed by chemical reactions high in the atmosphere. · Radiological calculations should be undertaken using models that more realistically treat the overlap of fallout plumes, complex meteoro- logical conditions, and that consider both external and internal doses. Patterns of land use and likely targeting strategy should be used in esti- mating the potential significance of various scenarios. The question of the possible release of radioactivity from nuclear fuel cycle facilities in a nuclear war should be explored more thoroughly.