years away from the present. As an example, average sea level rise is projected by current models to be about 30 centimeters, or one foot, at that time. This rate also seems to be consistent with the early results from the TOPEX satellite observations. I should be wanting in this recital if I did not direct attention to the large difference between this result and earlier predictions of order of magnitude larger changes. The number that surfaces most often in the discussion of climate is the change in the average global surface temperature change. The data for the past one hundred years has been painfully and methodically pieced together and shows rise of about 0.6 degrees Centigrade. Unfortunately, this is the least interesting aspect of climate change. What is crucial is knowledge of the change in the statistical behavior of regional quantities such as rainfall, storm frequency and intensity, flooding, coastal storm surges and so on. By contrast we can visualizee an induced change that increases cloud coverage to the extent that the temperature rises very little but clearly this is a definite climate change. The weaknesses of the models is most clearly demonstrated by the historical observation that, since the original Charney report of the NRC in 1978, the NAS report of 1983, the IPCC report of 1990 and today, the spread of the temperature rise between the fifteen or so climate models worldwide remains between 1.5 to 4.5 degrees centigrade for an anticipated doubling of atmospheric CO2. This variation between the models is also reflected in lack of agreement in many of the predicted regional climate changes which are so important. The primary reason for these disagreements is the poorly represented internal atmospheric radiation transfer properties and the overall effect of water, in clouds, cloud formation and water vapor. This is in recognition that water in the form of clouds and vapor is the primary greenhouse gas. Changes in CO2 concentration indirectly influence climate via a feedback effect of the water content of the atmosphere. This has been recognized for years, going back at least to the 1983 NAS report but it is only recently that serious field programs have been undertaken to attempt to formulate a better parametrization of these phenomena. The ARM and CHAMPP programs are in this category. The other element of time that is crucial is the lifetime of the excess anthropogenic CO2 in the atmosphere. This is a most important number for policy considerations. Here the advent of the coupled ocean-atmosphere model has made definite impact. What is being discussed here is how quickly would the atmosphere relax to a steady state or even its original unimpacted state if the anthropologicall input of CO2 was severely reduced or even curtailed. The base assumption is that, one way or another, fossil fuel reserves will be exhausted in several hundred years leading to a cutoff in the injection of CO2 into the atmosphere. At the time of the 1983 report the then current calculation in the literature proposed a behavior based on a 1000 thousand year exponential lifetime for the disappearance of the excess CO2 in the atmosphere. That is, that concentration would be reduced by one third each thousand years thereafter. This was serious. This meant that, even if the flow of CO2 were cutoff, the climate effects, for better or worse, would persist for that length of time. It is even worse. It would mean that modest reductions in emission rates would have negligible effects on the peak value of the perturbation (although it would delay its onset). It would take a massive change in emissions to have an appreciable effect. This creates policy situation that is almost a crisis one in that any climate change must be treated as hostile even in the absence of specific knowledge and the ability to make accurate climate change predictions. The coupled ocean-atmosphere model has given us better insight into the mechanism of this decay and has greatly reduced its value. For today we can take it as ranging between 50 and 160 years. The actual choice is a complicated one and is an approximation to the true time variation depending on how long is the extent and the shape of the emission curve. At any rate, it is big change from the prior 1000 year value and completely alters the policy picture. It can be interpreted in either of two ways or combination of both. On the one hand, those who feel that corrective measures must be taken immediately can now take comfort in the fact that any reasonably applied change in emissions would be reflected in a proportionate reduction in the peak CO2 concentration. On the other, those who take a more conservative view may argue that one can now safely wait until the climate changes become clearer and more definitely negative before taking action since the effect of the mitigating actions would show up in reasonable time. Taking into account this argument and the currently accepted gestation time of one century, I incline to the latter view. That inclination is strengthened by lack of certainty in the inevitability of all climate changes being harmful. One highly speculative example is based on the recently noted fact that the one degree Fahr enheit increase in temperature is primarily in the night time temperature variation. The daytime temperature is essentially flat. (The models, so far, do not exhibit this behavior.) Some argue that this phenomenon, if continued, would imply better agriculture. I repeat that this is highly speculative but does make the point that all change is not necessarily bad. The other concern is related to the frequency and severity of storms. There has been considerable arm-waving on this subject with differing conclusions. We now do have a sensitive model calculation (Report No. 139 from the Max-Planck-Institutfur Meteorologie). The authors conclude that, while the intensity of storms will remain the same, their number will decrease. This is still another indication that we have much to learn respecting the effects of anthropologically induced climate change. There are other important factors that need more field research effort. One is on the nature, distribution and lifetimes of aerosols in the atmosphere. I am assuming that other speakers will discuss this subject. The other and very difficult area is biota. There is a marked budgetary imbalance in the circulation of carbon between the atmosphere, land and the ocean. It is suspected, with some evidence, that it is caused by the growth of forests, a conclusion opposite to the generally held intuitive one. Until this is numerically fixed a certain degree of uncertainty will remain in model results, particularly in estimating residence times. One of the purposes of The Bridge is to bring a diversity of virus to the attention of Academy members and the public. This issue of the magazine presents several papers that address differ ent aspects of climate warming. William Nierenberg argues that little scientific progress has been made in understanding and predicting dimate warning caused by greenhouse gas emissions. He correctly points out that current estimates of future warming are no more precise than those of a decade ago. Still, notes Eric Baron, research over the last 10 years has substantially improved our ability to account for important atmospheric, ocean, and surface effects on climate. That the resulting climate models seem unable to better estimate global warming reflects the complexities introduced by more sophisticated modeling, not a failure of the underlying science, he believes. Neither point of view may have much relevance to public perceptions of global warming, however Through various types of questionnaires, Granger Morgan and his colleagues have revealed a public woefully uninformed about even the most basic elements of the global warming debate. Many noneqperts, it appears, have trouble differentiating between the specific causes of climate change and more general environmental pollution. A substantial proportion, in one survey, indicated they thought that the space program and nuclear power are in part to blame for global warming. The implications of these flaved mental models for personal, business, and political decision mak ing cannot be ignored These papers are timely given the expected publication this summer of a new report by the Intergovernmental Panel on Ch mate Change (IPCC). This international group of scientists, which operates under the aegis of the World Meteorological Organization and the United Nations Environment Program, periodically makes systematic appraisals of dimate change. The IPCC is regarded by many international organizations, includ ing the Conference of Parties on the Climate Warming Treaty, to be the authority on global warming assessment. While it is not possible to predict at this time exactly what the IPCC will conclude, many expect the group to substantially ret erte past findings that the average global surface temperature will rise 2°C to 4°C by the middle of the next century, when the greenhouse gas content of the atmosphere reaches double its current amount. It is also anticipated that the IPCC will present for the first time modified estimates of greenhouse gas warming that take into account the cooling effects of aerosols. The edent of mid-century warming under this scenario is likely to be less than that predicted by the standard model. William A. Nierenberg is director emeritus of the Scripps Institution of Oreanography in La Jolla, California. Progress and William A. Nierenberg SUMMER 1995 5 Recent data on the longevity of excess human-generated CO, in the atmosphere argue for only modest reductions in greenhouse gas emissions N early 7 years ago in these sume pages, I gave my views on the state of research on global warming (Nierenberg, 1988). The purpose of this article is to update that earlier assessment. In the time since my 1988 article was published in The Bridge, the United States has spent some $6 billion on research related to climate change and global warming. The nation currently invests about $1.6 billion annually on such studies. Given this level of support, what have been the results? What is known now that was not known in the mid1980s? Accepting that increasing CO2 concentrations will induce climate changes, when and where will these changes occur, and what will be their extent? Most important, what are the implications of the data that have been gathered, both for science and public policy? Atmospheric CO2 The one clear piece of information to emerge over the last decade is that CO, concentrations in the atmo‐ sphere are increasing steadily. This supports the findings of C.D. Keeling and his colleagues (Keeling and Bacastow, 1977), who studied the impact of industrial gases on climate nearly 20 years ago. The rate of increase slowed between 1980 and 1990, but this sort of fluctuation is not unusual. Most believe the upward trend will continue to parallel global population growth. Reductions in CO2 output by industrial nations through their use of energy efficient technolo gies will probably be offset by increases in greenhousegas production by transitional and developing countries as their standards of living improve. Scientists understand much less about the time dependence of the anthropogenic excess of CO2 in the atmosphere. Most observers agree that the pattern of rising global average surface temperatures over the last 100 years does not conclusively point to CO2 as the culprit; "normal" climate fluctuation may be to blame. Still, the near irreversibility of the predicted phe early efforts to calculate the longevity of excess atmospheric CO came up with estimates as high as 1,000 years (Keeling and Bacastow, 1977). That is, it would take a millennium for the CO2 to be absorbed into the total global system essentially, the world's oceans. The implications of this estimate are important. If true, there would be no clear path to take to reverse CO-induced climatic changes. And since it will take 200 to 300 years to exhaust the world's supply of fossil carbon fuels, modest reductions (on the order of 20 percent) in CO2 emissions would probably have little effect on climate. Reductions on the order of 80 percent would be necessary. That would be a daunting challenge. More recently, with the development of reasonable ocean models coupled to atmospheric climate models, researchers have arrived at greatly reduced estimates of the decay time of atmospheric CO2. Research now suggests that the decay curve is really a sum of exponentials. Current estimates indicate that it takes between 50 and 150 years for anthropogenic CO2 to be absorbed into the oceans (O'Neill et al., 1994).' The amount of CO2 in the atmosphere may be able to be "tailored" in some optimum fashion. From a policy standpoint, these findings raise two important issues. First, given these decay times, there may be no urgent need to take action to reduce emissions until the severity of the climate effects are either firmly predicted or clearly evident. If the climatechange "signal" does emerge from the natural background fluctuations and is unfavorable, there is ample time at that point to take whatever remedial action, including emissions controls, may be needed. Second, in contrast to the case if CO2 were expected to stay in the atmosphere for 1,000 years, the new estimates mean that reducing emissions will result in a corresponding reduction of the peak CO extess. Therefore, not only is there considerable elasticity in atmospheric response to CO, but the amount of CO2 in the atmosphere may be able to be “tailored” in some opti The BRIDGE 6 Modeling Climate Change One aspect of the climate warming issue that has not changed is the uncertainty surrounding how much the global average surface temperature will rise in response to a doubling of the concentration of greenhouse gases. There are approximately 15 computer facilities around the world devoted to modeling the hydrodynamics and thermodynamics of the atmosphere and oceans. In 1983, a committee assembled by the National Academy of Sciences (NAS) predicted that a twofold jump in greenhouse gas amounts would boost the glob al average surface temperature by between 1.5 and 4.5 degrees Celsius (Carbon Dioxide Assessment Committec, 1983). This temperature range was similar to values reported in an earlier NAS report (Assembly of Mathematical and Physical Sciences, 1979). Estimates of how much the seas will rise due to increasing concentrations of CO2 in the atmosphere have improved considerably. Today, despite continuing efforts to refine these models, the imprecision in the temperature-change predictions remains. This poses a dilemma, since the observed temperature rise in the last 100 years, if attributable to increased CO2 concentrations, lies just at the bottom limit of the range proposed by the 1983 NAS panel. Furthermore, the shape of the curve is not consonant with the predicted model behavior. It is unlikely that researchers will be able to better pin down CO associated temperature fluctuations in the foreseeable future. The models they use are mathematically complex, often embodying a large number of empirically determined physical and biological phenomena (such as the effect of soil moisture, the water retention of soils, ice formation and breakup, cloud formation in height, thickness, and density, as well the distribution of water droplets and aerosols). Some models also try to account for feedback phenomena, such as increased photosynthetic activity and oxidation of clathrates, while others attempt to incorporate changes in the rate of growth of other atmospheric greenhouse gases like methane, whose sources are poorly understood. Differences in the computers, software, numerical analyses, and phenomenologies used by the various research groups add to the problem. Another key weakness of the models, apparent even a decade ago, is their inability to capture adequately the interactions between water and solar radiation. Water vapor is the principal greenhouse gas, and its effects far exceed those of CO. Increasing CO2 concentrations actually enhance the thermodynamic behavior of atmospheric water. Although our understanding of these interactions has improved, we are still far from resolving their complexities. Thus, even as the individual elements of the models are better understood, the necessary introduction of new real-world climate variables will continue to hinder precise predictions of temperature change. Efforts to study climate have placed great emphasis on global average surface temperature. In the 1988 Bridge article, I pointed out that this number is a surrogate measure of regional climate changes that can be of serious consequence, such as droughts, floods, and storms, as well as of long-term ecological changes. While the average global temperature may vary by only a small amount, alterations in certain aspects of regional climates may be appreciable. Unfortunately, current models of climate change rarely yield consistent results in this area. Over the past 10 years, researchers have also learned that the global increase of 0.6 degree Celsius recorded during the past 100 years is attributable largely to a shift in nighttime temperatures. The fluctuation of daytime temperatures has been minimal. Estimates of how much the seas will rise due to increasing concentrations of CO2 in the atmosphere have improved considerably. The 1983 NAS report predicted that a doubling of CO would raise ocean heights about 2 feet. Now, however, scientists believe the correct amount is closer to 1 foot, a figure that agrees with satellite measurements of the current rate of change in sea level. This is supported by the very small change in the thickness of the principal ice caps. Research on Climate, Temperature, and CO2 During the last 10 years, long-awaited data from several research projects have shed light on some impor tant climate questions. Ice cores drilled in the Antarctic and Greenland ice caps, for example, provided a historical picture of the Earth's climate going back 100,000 years. One of the first bits of information to be extracted from trapped gases inside the cores was the atmospheric CO2 concentration in the period immediately preceding the Industrial Revolution. Despite |