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.
Chapter 1 CURRENT STATUS INTRODUCTION This chapter reviews recent changes in the state of understanding of the chemical and physical processes that determine the effect of human activities on concentrations of stratospheric ozone. The report is motivated by a continuing need to assess the potential effects on stratospheric ozone of chlorofluorocarbons (CFCs) and other chemicals, as prescribed in the Clean Air Act, as amended (42 USC 7450). The topic has been the subject of intense study during the past decade; our report builds on that work, most notably on previous studies by the National Research Council (NRC 1975, 1976b, 1977, 1978, 1979b) and the National Aeronautics and Space Administration (NASA) (Hudson and Reed 1979). To prepare our assessment, we relied on our professional knowledge, on a concurrent technical review prepared under the auspices of NASA, the Federal Aviation Administration, the National Oceanic and Atmospheric Administration, and the World Meteorological Organization (WMO) (Hudson et al. 1982), and on a series of topical reviews prepared at our request by technical consultants. The consultants' reports are contained in Appendixes A to F. PROCESSES DETERMINING OZONE CONCENTRATIONS Ozone (O3) is formed in the stratosphere by reaction of atomic oxygen (O) with diatomic molecular oxygen (O2). The process is initiated by photolysis of O2, that is, the dissociation of O2 into atomic oxygen by absorption of solar ultraviolet radiation at wavelengths below 240 nanometers (nm). Photolysis of O2 occurs mainly at altitudes above 25 km. 15
16 According to current understanding, approximately 1 percent of the ozone created in the stratosphere is removed by transport to the troposphere; the remaining 99 percent is destroyed by chemical reactions in the stratosphere that re-form ozone into O2. The net effect of these chemical reactions is either the combination of ozone with atomic oxygen to form O2, represented by the equation O + O3 ~ 2o2 r or the combination of two ozone molecules represented by O3 + O3 ~ 3O2 I II These equations represent the net results of a number of complex sets of reactions catalyzed by a variety of gases and chemical radicals present in the stratosphere in trace amounts. Important examples of sets of reactions summarized by process I are C1 + O3 + C10 + O2 (la) C10 + O ~ C1 + O2, (lb) NO + O3 ~ NO2 + O2 NO2 + O ~ NO + O2, OH + O3 ~ HO2 ~ O2 0 + HO2 + OH + O2 Process I may also proceed by the direct path O + O3 ~ 2O2 (2a) (2b) (3a) (3b) (4) These reactions are limited by the availability of oxygen atoms and therefore occur mainly at altitudes above 25 km. The reactions that limit the rates at which chains 1, 2, and 3 proceed are (lb), (2b), and (3b), respectively. Process II summarizes reaction schemes in which atomic oxygen is not limiting, for example, OH + O3 ~ HO2 + O2 HO2 + O3 ~ OH + 2O2. (5a) (5b)
17 Reactions (5) account for most of the ozone lost below 25 km in current models. The chemistry of the lower strato- sphere is complex, however (Appendix A), and one cannot exclude additional reaction schemes involving oxides of nitrogen and chlorine (NOX, ClOX) and oxidation products of hydrocarbons such as methane (CH4)~ Ozone removed from the stratosphere by transport to the troposphere is ultimately lost by chemical reactions in the gas phase or at the earth's surface. The spatial and temporal distribution of the concentration of ozone reflects a dynamic balance among the processes that form and remove ozone (Figure 1.1). According to current understanding, photolysis of O2 provides a global source of ozone of 50,000 million metric tons per year, with more than 90 percent of this amount formed above 25 km. Most of this ozone is removed by reactions represented by process I. At altitudes between 25 km and 45 km, reaction (2b) accounts for roughly 45 percent of the ozone removed while reactions (lb) and (4) each account for about 20 percent and reaction (3b) for 10 percent (S.C. Wofsy, Harvard University, private communication, 1982). About 1 percent of stratospheric ozone, 600 million metric tons per year, is removed below 25 km by process II, with a similar amount being lost by physical transport to the troposphere. Only 30 percent of global ozone is stored at altitudes above 25 km, reflecting the relatively short chemical lifetime of ozone at high altitudes. The rest is contained in the region below 25 km, and more than 70 percent of the amount below 25 km is found at latitudes above 30°. The abundance of ozone below 25 Am is determined by the balance between transport from the chemically more active region at higher altitudes and losses to the troposphere; its distribution is regulated by atmospheric motions. Adding to the stratosphere substances that destroy ozone has the effect of creating a new balance between production and removal processes in which the total abundance of ozone is reduced. For example, stratospheric concentrations of chlorine monoxide (C1O) and nitrogen dioxide (NO2) may be increased as a result of emissions of CFCs and nitrous oxide (N2O) from human activities. The effects are persistent. A typical CFC molecule, CF2C12 for example, survives for approximately 75 years in the atmosphere before it is decomposed by sunlight releasing its constituent chlorine atoms in the
18 / ////.4 I ~ o i' I .~°, ~ o ° o ~ on' I 1 1 1 1 to I o I ~ = 1 ~ - ~gag \ \ C ~ \ ~ o \ lo, £= \ \ i\ C To ',o:, \ / ,Y,0~t / : ~/ / / / Ads' :~ 1 v. l ~ I a' ° Cot O I,,, -o ]'~° ·~'C '° \ ~ \~ \ .O , \ \ \ \ o At, \ Who 0~/ o' As, By O. - A _& Cal o Ct Cal .= o o Cal o o ·_ Cal hi o hi hi id c) o a, o o ~ - ct C/3 c:
19 stratosphere. _~ _" A chlorine atom can affect recombination of between 104 and 105 ozone molecules during its lifetime in the stratosphere (on the order of two years) before it returns to the troposphere, mainly as hydro- chloric acid (HC1). A similar situation holds for N2O. Approximately 10 percent of N2O molecules released to the atmosphere decompose by paths leading to production of stratospheric nitric oxide (NO), and subsequently NO2, by reaction (2a). The average NOX molecule also removes between 104 and 105 ozone molecules before it returns to the troposphere, after its typical two-year residence in the stratosphere. Current theoretical models lead us to conclude that the dependence of ozone concentration on altitude will also change, the net effect being a redistribution of ozone from higher to lower altitudes. Quantitative estimates of these effects have varied somewhat over the past decade (Appendix A). Perturbations by Chlorine Currently, approximately 3 parts per billion (ppb) of the lower stratosphere consists of chlorine bound in organic molecules such as methyl chloride (CH3C1), carbon tetrachloride (CC14), and CFCs (Hudson and Reed 1979, Hudson et al. 1982). Table 1.1 indicates the abundances of the more prevalent species; only methyl chloride is known to have natural origins. The table also shows estimates of current rates of release of man-made compounds found in the lower stratosphere. Halocarbons decompose under the influence of sunlight at altitudes above 20 km; the fractional abundances (mixing ratios) of halocarbons (in ppb) are observed to decrease with increasing altitude (Appendix C). The chlorine produced by decomposition of halocarbons is converted to inorganic species, including HC1, chlorine nitrate (ClNO3), C1O, and atomic chlorine (C1). Hydro- chloric acid is the major reservoir for chlorine at altitudes above 25 km (Appendix C). Concentrations of C1, C1O, and HC1 have been observed in the stratosphere; observations and predictions of theoretical models are in general agreement, although some difficulties remain (Appendix D), as we shall see. Computer calculations using current understanding and incorporating new data on rates of several important reactions (Appendixes C and D) suggest that continued release of the CFCs, CF2C12 and CFC13, at rates _ , , ~. ~ ~ ~ \
20 TABLE 1.1 Concentration in the Lower Stratosphere and Release Rates of Major Sources of Chlorine in the Stratosphere Rate of Release Concentration (ppb) a (million metric . Compound Molecular Chlorine tons of C1 per year) Methyl chloride (CH3 C1) 0.62 0.62 2b F-12 (CF2C12) 0.30 0.60 o.lgc Fell (CFC13) 0.18 0.54 o.2oc Carbon tetrachloride (CC14) 0.13 0.52 0.0s3d Methyl chloroform (CH3CC13) 0.1 1 0.33 0 3Se aHudson et al. (1982). bAbout 85 to 90 percent of CH3 C1 is naturally produced, the remainder being attr~b uted to industrial sources (Cicerone 1981). The total release rate varies slowly in time because of the large contributions of natural sources. C1980 release rate from "World Production and Release of Chlorofluorocarbons 1 1 and 12 through 1980," Chemical Manufacturers Association Fluorocarbon Program Panel, July 29, 1981. Release rate has decreased by about 20 percent from the peak rate of 1974. dl 976 release rate (NRC 1 979b). The release rate is apparently relatively constant, although somewhat uncertain. eNeely and Plonka (1978). prevalent in 1977 would ultimately cause a net decrease of total global ozone roughly between 5 percent and 9 percent assuming no other perturbations (Hudson et al. 1982). We regard a representative result to be 7 percent tAPpendix C). This would result in a smaller steady state reduction in ozone than reported in NRC (1979b), which was 16.5 percent with a 95 percent probability that the true value lies between 5 percent and 28 percent. bother models current in 1979 gave reductions ranging from 15 percent to 18 percent tHudson and Reed 1979). Estimates have fluctuated between roughly 5 percent and 20 percent over the past eight years as models have been refined (Appendix A).) The steady state reduction would be reached asymptotically in times on the order of a century. Calculations now indicate that the reduction would occur almost entirely at altitudes above 35 km, in the region of the stratosphere where the ozone concentra- tion is determined primarily by chemical processes, with a smaller, partially compensating increase in ozone concentrations at lower altitudes. The current result obtains for both 1- and 2-dimensional models and further differs from that prevalent in 1979 in that earlier calculations showed regions of reduction both above and below 35 km.
21 The differences between current findings and those reported in 1979 are the result of refinements in the values for the rates of several reactions affecting the concentration of the hydroxyl radical (OH) (Appendixes C and D). The refinements are the result of improved laboratory measurements (Hudson et al. 1982). OH is important because the concentration of C1O in the lower stratosphere is particularly sensitive to it. Results of model calculations using current values for these reaction rates are in good agreement with observations of C1O for altitudes below 35 km (Appendixes C and D), whereas models using the reaction rates favored in 1979 give concentrations of C1O a factor of 3 higher than observed values in this range. The new reaction rates have not changed greatly the results of calculations for altitudes above 35 km, however, so that the amounts of reduction in ozone above 35 km obtained in the 1979 and current models are about the same. The models continue to indicate lower concentrations of C1O in the stratosphere above 40 km than are observed. We shall return to this discrepancy. Increased attention to effects of releases of methyl chloroform (CH3CC13) on stratospheric ozone is warranted because of the growing use of this compound, an industrial solvent. The release rate increased by a factor of about 50 between 1958 and 1978 (Neely and Plonka 1978). Perturbations by Oxides of Nitrogen The chemically active oxides of nitrogen in the strato- sphere (such as NO2) are thought to arise mainly from photooxidation of N2O. N2O is formed naturally by bacteria in soil and water. As indicated earlier, reactions involving NO2 account for about 45 percent of the ozone removed in the stratosphere between 25 km and, 45 km. The human influence on the global cycle of fixed nitrogen is thought to be significant and increasing (NRC 1978). The global atmospheric concentration of N2O appears to have increased by 2.7 percent (from 292 ppb in 1964 to 300 ppb in 1980) over the past 16 years (Weiss 1981, Weiss and Craig 1976). The concentration of N2O in the atmosphere is likely to continue to increase with increases in emissions associated with agricultural practices, disposal of human and animal wastes, and possibly combustion; but we cannot say how or on what time scale.
22 An increase in N2O concentrations of about 30 percent in the absence of other perturbations could cause a reduction in global ozone of an amount comparable with the 7 percent reduction currently estimated due to continued emissions of CF2C12 and CFC13 at 1977 rates, also taken as the sole perturbation. This estimate is based on current model calculations that indicate that, should the concentration of N2O double in the absence of other perturbations, total global ozone would decline between 10 percent and 16 percent (Hudson et al. 1982). Early attention to human influences on the strato- sphere focused on effects of NOX released by high- flying aircraft (NRC 1975). Models then and now suggest that an input of NOk at altitudes above about 20 km should lead to reduction in stratospheric ozone. A source of NOX at lower altitude, associated for example with subsonic commercial aviation, can modify local chemistry such as to cause an increase in tropospheric ozone. It has been suggested that reductions in the column of ozone above the earth's surface due to reductions in stratospheric ozone may be masked to some extent by increases in tropospheric ozone attributable to subsonic jets and urban smog. Assessment of the impact on stratospheric ozone due to a combination of perturbations requires investigation of specific cases since the effects are not simply additive. Hudson et al. (1982) report the results of several studies of the effects of doubling atmospheric N2O concentrations and continuing releases of CFCs at 1977 rates, both separately and in combination. The Lawrence Livermore National Laboratory (LLNL) model, for example, indicated a reduction of 12.5 percent due to doubling N2O with a reduction of 12.9 percent due to the combination of perturbations. The LLNL model gives a reduction of 5.0 percent for CFC releases alone. Another model, from Atmospheric and Environmental Research, Inc., gives reductions of 9.5 percent for doubling N2O, 6.1 percent for continuing CFC releases, and 13.0 percent for the combination. The results may be misleading, however, since current trends suggest a considerably longer time scale for doubling atmospheric concentrations of N2O than for reaching the steady state reduction due to continued emissions of CFCs.
23 Perturbations by Other Species Stratospheric ozone may be affected by human activity in a number of other ways. Of greatest potential concern are changes in concentrations of carbon dioxide (CO2), water vapor (H2O), and perhaps methane (CH4). The well-documented increase in atmospheric concentrations of CO2 is directly attributable to combustion of fossil fuels and wood. This increase is expected to lead to a global warming of the atmosphere near the surface of the earth but is expected to cause a reduction in the temperature of the stratosphere (Fels et al. 1980). Lower stratospheric temperatures would have at least two effects. First, the chemical removal processes affecting ozone that were described earlier are sensitive functions of temperature, being less efficient at lower temperature. Consequently, with lower temperature the equilibrium concentration of ozone would be higher. Current models incorporating this effect suggest that the steady state reduction in total ozone due to continuing emissions of CFCs at 1977 rates would change from 5 percent to 9 percent to between 4 percent and 6 percent if global CO2 were doubled concurrently (Hudson et al. 1982). (Global CO2 has increased by about 6 percent in the past 22 years.) The possible second effect of lower statospheric temperatures resulting from increased CO2 is a thermally driven change in stratospheric water vapor (H2O) caused by a change in the temperature of the tropical tropopause. Dissociation of H2O provides the source of hydrogen radicals, and these radicals play a key role in strato- spheric chemistry regulating abundances of both active NOk and C1X species in addition to their contributions to reactions (3) and (5). A complete model for strato- spheric chemistry should include a description of H2O interactions, a requirement beyond current capability. Stratospheric ozone may also vary in response to changes in concentrations of CH4, which plays an important role in reaction (lb) by regulating the partitioning of chlorine between HC1 and C1O (Hudson et al. 1982). Recent reports (Rasmussen and Khalil 1981) suggest increases in global concentrations of CH4, but likely future changes and their consequences are unknown.
24 CURRENT STATUS OF MODELS OF THE STRATOSPHERE Theoretical models of stratospheric chemistry cannot be validated by measurements of total ozone only, owing to the diversity of factors, natural and man-made, that may affect ozone concentrations. Comparison of calculated and observed values for the concentrations of important trace species and radicals--such as OH, ClO, NO2, and atomic oxygen--must play a central role in any orderly strategy for validating models. In general terms, agreement in detail between the predictions of theoretical models and observations is excellent. For example, changes in reaction rates since 1979 have resulted in substantial agreement between theory and observation for ClO below 35 km. There are, however, three areas in which discrepancies remain. The discrepancies may or may not point to significant difficulties in modeling. Similarly, agreement between modeling results and observation of C10, while encouraging, need not imply validity of the model at lower altitudes. The improved agreement between observed and calculated concentrations of C10 in the lower stratosphere may be attributed mainly to changes in rate constants for reactions affecting OH. Concentrations of OH in current models are lower than values obtained in 1979, with the result that a larger fraction of Clx is now found as HC1. The chemistry of the lower stratosphere is complex, however. Agreement between model and observed values of C10 in the lower stratosphere should be considered necessary but not sufficient for validation. A more extensive and demanding test would require comparison of theoretical and observed profiles of other radicals, particularly OH. To improve understanding of stratospheric chemistry also requires that attention be directed to the assumptions of the models and to the measurements against which models are tested. High-quality measurements are obviously prerequisite to validation of models. Confidence in observations of critical species is enhanced by using a number of independent, inter- calibrated techniques, each relying on different physical properties. Validation of measurement technique is difficult since concentrations of the important atmospheric species may vary in time and space on scales that are not well understood. Validation procedures involve coordinated studies in the field requiring considerable logistical support.
25 The assumptions of models are of two types: (1) input ~ ~ conditions, reaction rates, and (2) the reaction schemes incor data on environmental other parameters, and porated into the model. There are still uncertainties about the appropriateness of some assumptions common in current models. For example, the rate for reaction of OH with HO2, an important path for removal of hydrogen radicals, remains uncertain despite extensive and continuing efforts in the laboratory. There are other reactions in need of similar clarification. Models are sensitive to assumptions about the abundance and distribution of stratospheric H2O; the underlying physical and chemical processes that regulate this key parameter are not well understood. It is difficult to rule out the possibility of an important role for species not now included in models, and, if history is a guide, there may well be future surprises in this area. Models for the stratosphere have been adjusted over the past decade in just this manner to include gases such as C10 (1974), ClNO3 (1976), and HOC1 (1978) (Appendix A, Figure A.1), and there is current discussion of a possible participation of sodium (Kolb and Elgin 1976, Murad et al. 1981). Progress in recognition of missing species or reactions occurs through a combination of laboratory, field, and theoretical studies, the normal practice of validating models and resolving discrepancies. As was noted earlier, there is reasonable agreement between model calculations and observations for C10 in the lower stratosphere. Currently, however, there is a discrepancy between theory and observation for C10 in the region above 35 km, where chlorine-mediated catalysis is most important. The average value for the concentration of C10 measured by Anderson and co-workers (see Appendix D) near 40 km is almost a factor of 2 larger than the value calculated from models. Furthermore, theory and experiment give different dependences of the concentration of C10 on altitude in the upper stratosphere. The C10 discrepancy is particularly important because it occurs at altitudes where ozone is most sensitive to perturba- tions caused by CFCs (Appendix D, Figure D.52). Extensive ground-based observations of NO2 have been made over a range of latitudes by J. Noxon (see Appendix C), revealing a sharp spatial discontinuity in concentra- tion in the winter with very low concentrations poleward of the discontinuity. Thus far, no theroretical model has been able to explain this phenomenon.
26 A third area of discrepancy between current models and observations is in concentrations of CFC13 and CF2C12 at altitudes above about 20 km (see Appendix C, Figure C.ll). Observed values are substantially lower than predicted values. The difference could be due to errors in model simulation of ultraviolet radiance in the lower stratosphere, which, if true, would imply that the CFCs have shorter residence time in the lower strato- sphere. The issue is not resolved and requires continuing attention. Nevertheless, the extent of agreement between measurement and theory is encouragingly good. MONITORING AND ASSESSMENT OF TRENDS Measurements of the total amount of ozone above a unit area of the earth's surface (called total column ozone) are essential for assessing the human influence on ozone (as well as the potential effects of changes in ozone on humans and other organisms). As detailed in Appendix A, total column ozone fluctuates on a variety of spatial and temporal scales owing to natural causes; these fluctua- tions tend to mask possible systematic changes due to man-made perturbations. For example, current models for single and combined perturbations predict a reduction of total column ozone over the past decade of less than 1 percent, but a change of this magnitude cannot be distinguished from fluctuations due to other causes (Appendixes D and E). Models of the stratosphere predict that the largest reductions in ozone due to releases of CFCs should occur near 40 km. Reductions should therefore be most readily detectable at this altitude. Current models suggest that ozone concentrations at 40 km should have decreased by several percent over the past decade. There have been reports in the press that an effect of this order has been detected in data from satellite experiments (see, for example, Science, Sept. 4, 1981, pp. 1088-1089). The community of atmospheric scientists has not yet had the opportunity to scrutinize this evidence, which must therefore be regarded as preliminary (Appendix F). Our ability to detect trends in ozone in the future will depend on the availability of consistent, high- quality data taken over long time intervals. Improvements in the current monitoring systems are feasible and clearly needed. For example, it is vitally important to improve
27 and enhance systems for monitoring ozone profiles in the upper stratosphere that could provide a valuable early indication of systematic changes In ozone aue co emend ~ Ural of CFCs or N2O, but existing data in the upper strato- sphere are inadequate for this purpose. It is also imperative to continue, and desirable to expand, the high-quality monitoring of total ozone by Dobson spectrophotometers. THE QUESTION OF EARLY DETECTION A notable feature of the ozone issue is that a reduction due to increases in the tropospheric concentrations of CFCs or N2O, once it has taken place, is expected to persist for more than 100 years even if the practices that caused it are stopped immediately. It is therefore important to detect an anthropogenic effect at the Three methods currently exist , earliest possible time. for this purpose. 1. Measurement of total ozone. Relying on measurement of total ozone has the following advantages (Appendix E): There exists a relatively long historical base (30 to 50 years) of data. Ground-based instrumentation is available and may be readily complemented by observations from satellites. Finally, total ozone is most directly related to one of the consequences of depletion that is of concern, the possibility of enhanced exposure to ultraviolet radiation at the ground (Part II). Since, however, the reduction due to CFCs is expected to be concentrated at high altitudes, measurements of total column ozone are less sensitive indicators of an anthropogenic effect than are measurements of ozone profiles. 2. Measurement of ozone at high altitudes. The advantages of this method derive from the theoretical result that changes in ozone due to CFCs are predicted to be largest at high altitudes. Changes in the spatial distribution of ozone may be important for understanding the second major consequence of depletion that is of concern, the possibility of climate change (Appendixes and C). The disadvantages stem from the difficulty of making the measurements, whose quality and stability are inferior to those of total ozone (Hudson and Reed 1979, Hudson et al. 1982). Satellite data are particularly subject to changes in calibration of instruments, which -- B
28 cannot be refurbished; ground-based measurements by the Umkehr method give poor height resolution and are subject to perturbations by hazes and stratospheric particulate matter. Partly because of these difficulties, the data base is relatively small and somewhat fragmented (Appendix F). Ozone measurements using satellites would have the desirable attribute of obtaining temporal and spatial distributions that would be useful in validating 2- and 3-dimensional models. 3. Measurement of key radicals involved in chemical removal processes. Measurements~of spatial and temporal profiles of important species such as C10 and OH may be combined with chemical models for assessment of trends and their causes, such that the dependence on specific models can be relatively slight. This method is in principle the most sensitive, but it is also the least direct. The last approach is regarded by many experts as having already shown the effect of chlorine of human origin, mainly connected with emissions of CFCS. But this conclusion would be more firmly established with more direct confirmation, as discussed in the previous section. Ideally, all three types of measurement should be integrated (with due regard to their sensitivity) in a strategy for early detection of anthropogenic effects. UNCERTAINTY Quantitative estimates of the uncertainties inherent in current estimates of reductions in ozone due to emissions of CFCs and N2O are difficult to obtain. The ability to make quantitative estimates of uncertainty depends both on what we know and on what we do not know. Such estimates employ professional judgments about the importance of various factors and the sensitivity of the results to potential changes in understanding. Our major concern in estimating uncertainties in our understanding of stratospheric ozone is with the possibility that some key process or processes may be missing from current models. In an orderly scientific strategy, continuing development of models on the basis of an ongoing comparison with observational data is expected. Progress is stimulated by the existence of discrepancies or uncertainties and tends to occur in more or less discrete steps rather than uniformly. Our
29 understanding of the lower stratosphere has improved over the past two years as a result of developments that may be attributed at least in part to efforts to resolve earlier (and larger) discrepancies between observed and computed values for the concentration of C1O. Agreement between observed and computed values of C1O is now satisfactory below 35 km, but, as noted earlier, there continues to be a serious discrepancy at higher altitudes. This disagreement illustrates the difficulty of estimating limits of uncertainty for current estimates for reduction in ozone due to CFCs. For example, observed values of C1O at higher altitudes are larger than calculated values, suggesting that the long-term reduction in ozone could be correspondingly larger. One can, however, conceive of speculative chemical schemes that could suggest a stratosphere less vulnerable to perturbations. In circumstances such as this, the usual ways of estimating uncertainty (using mathematically rigorous procedures) are not applicable. professional judgment. Instead we rely on The predictions of the current chemical scheme have been cross-checked against observed atmospheric data in many ways, and the agreement in general is quite good. As stated earlier, a representa- tive estimate of potential steady state reduction of global ozone due to continued releases of CFCs at the 1977 rate in the absence of other perturbations is 7 percent. There continue to be, however, important discrepancies between theory and observation. Our opinions are divided on whether there are sufficient scientific grounds to estimate the effect of resolving one of the discrepancies, that of C1O in the upper stratosphere, on calculations of ozone reduction. We agree that we do not know enough at this time to make a quantitative judgment of the uncertainty associated with the other major discrepancies, NO2 at high latitudes and lifetime of CFCs in the stratosphere above 20 km. Those of us who believe there are grounds to judge the effect of resolving the C1O issue conclude that our estimate of ozone reduction from CFC emissions should not change by more than a factor of 2. Those of us unwilling to offer quantitative estimates of uncertainty hold the conviction that no rigorous scientific basis exists for such statements. We are concerned by implications of the discrepancies noted earlier. These discrepancies should be resolved in the
30 next few years by orderly application of the scientific method with appropriate interaction between theory and observation. We see no reason to prejudge the result of this process. Research during the past several years has enhanced our understanding of the factors affecting stratospheric ozone. Development of the field is progressing rapidly. We anticipate further developments in both observation and modeling in the next few years that will result in considerable improvement in our understanding, both clarifying and reducing uncertainties. FINDINGS 1. Our understanding of the stratosphere has advanced considerably in the past two years. Progress is significant in all areas with improvements in our ability to model the system in more than 1 dimension, with impressive achievements in techniques for measurement of chemical reactions in the laboratory, and with major advances in our ability to measure concentrations of important trace species in the atmosphere. We note her that the success of the research is due in no small part to the breadth of the scientific effort involving scientists from many countries with support from both private and governmental sources. We expect continued improvement in understanding of the chemistry and dynamics of ozone reduction to result from research currently under way, planned, and proposed. 2. The concern regarding the possibility of reduction e in stratospheric ozone due to CFCS remains, although current estimates for the effect are lower than results given in NRC (1979b). The change in estimates of ozone reduction reflects improvements in our understanding of chemical processes in the stratosphere below 35 km. There has been no significant change in results obtained by models for the stratosphere above 35 km. The major impact of CFCs is predicted for the height range of 35 km to 45 km. 3. The chlorine species C1 and C10 participate in a series of chemical reactions that destroy ozone. The radical C10 has been measured in the stratosphere in significant amounts and is believed to be primarily of human origin. Our current understanding indicates that if-production of CFCs continues into the future at the rate existing in 1977, the steady state reduction in
31 total ozone, in the absence of other perturbations, would be between 5 percent and 9 percent. Previous estimates fluctuate between roughly 5 percent and 20 percent, with those current in 1979 ranging from 15 percent to 18 percent. Latest results also suggest that CFC releases to date should have reduced the total ozone column by less than 1 percent. 4. According to current understanding, increases of N2O in the stratosphere would result in reductions in total ozone. with the largest effects occurring in the lower stratosphere. Although concentrations of N2O in the stratosphere appear to be increasing, we cannot reliably project the future course of N2O sources. If, however, the concentrations of N2O in the atmosphere were to double, in the absence of other perturbations, current models suggest that the steady state reduction in the total ozone would be between 10 percent and 16 percent. 5. On the whole, there have been substantial improvements in the agreement between model predictions and observed profiles of trace species in the past several years. Three exceptions are still a cause for concern: Above 40 km, more C1O is observed than is predicted by current theory; the behavior of NOX in winter at near-polar latitudes is unexplained; and concentrations of CFCS in the stratosphere above 20 km are lower than predicted by the models. 6. Examination of historical data (extending back 30 to 50 years) has not yet shown a significant trend in total ozone '~hat can be ascribed to human activities. Current models of combinations of pollutants suggest that a reduction of total ozone to date from human activities would be less than 1 percent. No detectable trend would be expected on the basis of these results. 7. Data on total ozone should not be used alone to guide decisions on whether to take action to prevent future changes in stratospheric ozone. Although an important guide, analysis of trends in total ozone cannot by itself reveal causes of ozone reductions or increases. Such analysis, together with measurement of altitude profiles of trace species and ozone and theoretical modeling, offers promise of understanding causes of ozone changes and the consequences of alternative actions in response. 8. The impact of CFCs should be assessed in the context of a broad understanding of the variety of ways in which human activity can alter stratospheric
32 composition. Ozone may be reduced by increasing levels of CFCs and N2O, but reductions might be offset in part by higher concentrations of CO2 and perhaps CH4. Human activities have already increased the amounts of CO2 and CFCS in the atmosphere, and from the known release rates, further increases can be confidently expected. In addition, there is evidence that N2O and CH4 concentrations are also increasing. A special reason for concern about perturbations potentially caused by CFCs and N2O is the long lifetime of these gases in the atmosphere, of the order of 50 to 150 years. Even if the releases of these gases were reduced, the atmosphere would not recover until far in the future. RECOMMENDATIONS In light of our findings, we believe it is important to maintain a competent, broadly based research program that includes a long-term commitment to monitoring programs. The research effort should extend over at least two solar cycles (of 11 years each) to distinguish between changes induced by variations in the sun from those associated with man. Accordingly, we make the following recommendations: 1. The national research program, including atmospheric observations, laboratory measurements, and theoretical modeling, should maintain a broad perspective with some focus on areas of discrepancy between theory and observation. A coordinated research program to understand the spatial and temporal distributions of key species and radicals merits highest priority. Observations should be extended to include studies of the equatorial and polar regions. 2. The global monitoring effort should include both ground-based and satellite observations of total ozone and of concentrations of ozone above 35 km, where theory indicates the largest reductions might occur. We also need data to define the variability of stratospheric temperature and water vapor. We regard sound, satellite-based systems for stratospheric observations as essential. 3. Potential emissions of a number of relevant gases, in addition to CFCs and N2O, and their consequences for stratospheric ozone should be thoroughly evaluated and assessed. It is important that we understand current and
33 potential rates of emissions of these compounds and the effects these emissions might have on ozone in addition to understanding emissions and effects of CECs. There is observational evidence that atmospheric concentrations of N2O and CO2 are increasing. Models should be developed to describe the combined effects on strato- spheric ozone of future changes in releases of all relevant gases, such as CFCs, N2O, CO2, CB4, CH3C1, and CH3CC13.