strong Northern Hemisphere sink containing both oceanic and terrestrial biospheric components, and a weak Southern Hemisphere (SH) sink. The previous IPCC global estimate for an ocean sink of 2.0±0.8 GtC per year is still a reasonable one. The terrestrial biospheric processes which are suggested as contributing to the sinks are sequestration due to forest regeneration, and fertilization arising from the effects of both CO2 and nitrogen (N), but none of these can be adequately quantified. This implies that the imbalance (of order 1-2 GtC/yr) between sources and sinks, i.e., "the missing sink", has not yet been resolved. This fact has significant consequences for estimates of future atmospheric CO2 concentrations and the analysis of the concept of the Greenhouse Warming Potential.

Sources of Methane.—A total (anthropogenic plus natural) annual emission of CH. of about 500Tg can be deduced from the magnitude of its sinks combined with its rate of accumulation in the atmosphere. While the sum of the individual sources is consistent with a total of 500Tg CH4, there are still many uncertainties in accurately quantifying the magnitude of emissions from individual sources. Significant new information includes a revised rate of removal of CH, by atmospheric hydroxyl (OH) radicals (because of a lower rate constant), a new evaluation of some of the sources (e.g., from rice fields) and the addition of new sources (e.g., animal and domestic waste). Recent CH4 isotopic studies suggest that approximately 100Tg CH, (20% of the total CH source) is of fossil origin, largely from the coal, oil, and natural gas industries. Recent studies of CH. emissions from rice agriculture, in particular Japan, India, Australia, Thailand and China, show that the emissions depend on growing conditions, particularly soil characteristics, and vary significantly. While the overall uncertainty in the magnitude of global emissions from rice agriculture remains large, a detailed analysis now suggests significantly lower annual emissions than reported in IPCC 1990. The latest estimate of the atmospheric lifetime of CH. is about 11 years.

Sources of Nitrous Oxide.-Adipic acid (nylon) production, nitric acid production and automobiles with three-way catalysts have been identified as possibly significant anthropogenic global sources of nitrous oxide. However, the sum of all known anthropogenic and natural sources is still barely sufficient to balance the calculated atmospheric sink or to explain the observed increase in the atmospheric abundance of N2O.

Sources of Halogenated Species.-The worldwide consumption of CFCs 11, 12, and 113 is now 40% below 1986 levels, substantially below the amounts permitted under the Montreal Protocol. Further reductions are mandated by the 1990 London Amendments to the Montreal Protocol. As CFCs are phased out, HCFCs and HFCs will substitute, but at lower emission rates.

Atmospheric Concentrations and Trends of Other Gases that Influence the Radiative Budget

Ozone (O3) is an effective greenhouse gas both in the stratosphere and in the troposphere. Significant decreases have been observed during the last one to two decades in total column O, at all latitudes except the tropics-in spring, summer and winter. The downward trends were larger during the 1980s than in the 1970s. These decreases have occurred predominantly in the lower stratosphere (below 25 km), where the rate of decrease has been up to 10% per decade depending on altitude. In addition, there is evidence to indicate that O, levels in the troposphere up to 10 km altitude above the few existing ozonesonde stations at northern middle latitudes have increased by about 10% per decade over the past two decades. Also, the abundance of carbon monoxide (CO) appears to be increasing in the NH at about 1% per year. However, there is little new information on the global trends of other tropospheric O, precursors, (non-methane hydrocarbons (NMHC) and oxides of nitrogen (NO,)).

Stratospheric Ozone Depletion.-Even if the control measures of the 1990 London amendments to the Montreal Protocol were to be implemented by all nations, the abundance of stratospheric chlorine and bromine will increase over the next several years. The Antarctic ozone hole, caused by industrial halocarbons, will therefore recur each spring. In addition, as the weight of evidence suggests that these gases are also responsible for the observed reductions in middle- and high latitude stratospheric Os, the depletion at these latitudes is predicted to continue unabated through the 1990s.

Sources of Precursors of Tropospheric Ozone.-Little new information is available regarding the tropospheric ozone precursors (CO, NMHC, and NO,), all of which have significant natural and anthropogenic sources. Their detailed budgets there fore remain uncertain.

Source of Aerosols.-Industrial activity, biomass burning, volcanic eruptions, and sub-sonic aircraft contribute substantially to the formation of tropospheric and stratospheric aerosols. Industrial activities are concentrated in the Northern Hemisphere where their impact on tropospheric sulphate aerosols is greatest. Sulphur emissions, which are due in large part to combustion effluents, have a similar emissions history to that of anthropogenic CO2. Estimates of emissions of natural sulphur compounds have been reduced from previous figures, thereby placing more emphasis on the anthropogenic contribution.

(iii) Scenarios of Future Emissions

Scenarios of net greenhouse gas and aerosol precursor emissions for the next 100 years or more are necessary to support study of potential anthropogenic impacts on the climate system. The scenarios provide inputs to climate models and assist in the examination of the relative importance of relevant trace gases and aerosol precursors in changing atmospheric composition and climate. Scenarios can also help in improving the understanding of key relationships among factors that drive future emissions.

Scenario outputs are not predictions of the future, and should not be used as such; they illustrate the effect of a wide range of economic, demographic and policy assumptions. They are inherently controversial because they reflect different views of the future. The results of scenarios can vary considerably from actual outcomes even over short time horizons. Confidence in scenario outputs decreases as the time horizon increases, because the basis for the underlying assumptions becomes increasingly speculative. Considerable uncertainties surround the evolution of the types and levels of human activities (including economic growth and structure), technological advances, and human responses to possible environmental, economic and institutional constraints. Consequently, emission scenarios must be constructed carefully and used with great caution.

Since completion of the 1990 IPCC Scenario A (SA90) events and new information have emerged which relate to that scenario's underlying assumptions. These developments include: the London Amendments to the Montreal Protocol; revision of population forecasts by the World Bank and United Nations; publication of the IPCC Energy and Industry Sub-group scenario of greenhouse gas emissions to 2025; political events and economic changes in the former USSR, Eastern Europe and the Middle East; re-estimation of sources and sinks of greenhouse gases (reviewed in this Assessment); revision of preliminary FAO data on tropical deforestation; and new scientific studies on forest biomass. There has also been recognition of considerable uncertainty regarding other important factors that drive future emissions. These factors have led to an update of the SA90. Six alternative IPCC Scenarios (IS92 a-f) now embody a wide array of assumptions, summarized in Table 1, affecting how future greenhouse gas emissions might evolve in the absence of climate policies beyond those already adopted. This constitutes a significant improvement over the previous methodology. However, the probability of any of the resulting emission paths has not been analyzed. IPCC WGI does not prefer any individual scenario. Other combinations of assumptions could illustrate a broader variety of emission trajectories. The different worlds which the new scenarios imply, in terms of economic, social and environmental conditions, vary widely. The current exercise provides an interim view and lays a basis for a more complete study of future emissions of greenhouse gas and aerosol precursors.

Scenario Results

The range of possible greenhouse gas futures is very wide, as the Figure below illustrates (showing only CO2). All six scenarios can be compared to SA9O. IS92a is slightly lower than SA90 due to modest and largely offsetting changes in the underlying assumptions. (For example, compared to SA9O, higher population forecasts increase the emission estimates, while phaseout of halocarbons and more optimistic renewable energy costs reduce them.) The highest greenhouse gas levels result from the new scenario IS92e which combines, among other assumptions, moderate population growth, high economic growth, high fossil fuel availability and eventual hypothetical phaseout of nuclear power. The lowest greenhouse gas levels result from IS92c which assumes that population grows, then declines by the middle of the next century, that economic growth is low and that there are severe constraints on fossil fuel supplies. The results of all six scenarios appear in Table 2. Overall, the scenarios indicate that greenhouse gas omissions could rise substantially over the coming century in the absence of new measures explicitly intended to reduce their emission. However, IS92c has a CO2, emission path which eventually falls below its starting 1990 level. IS92b, a modification of İS92a, suggests that current commitments by many OECD Member countries to stabilize or reduce CO2 might have a small

impact on greenhouse gas emissions over the next few decades, but would not offset substantial growth in possible emissions in the long run. IS92b does not take into account that such commitments could accelerate development and diffusion of low greenhouse gas technologies, nor possible resulting shifts in industrial mix.

Carbon Dioxide.-The new emissions scenarios for CO2 from the energy sector span a broad range of futures. Population and economic growth, structural changes in economies, energy prices, technological advance, fossil fuel supplies, nuclear and renewable energy availability are among the factors which could exert major influence on future levels of CO2 emissions. Developments such as those in the republics of the former Soviet Union and in Eastern Europe, now incorporated into all the scenarios, have important implications for future fossil fuel carbon emissions, by affecting the levels of economic activities and the efficiency of energy production and use. Biotic carbon emissions in the early decades of the scenarios are higher than SA90, reflecting higher preliminary FAŎ estimates of current rates of deforestation in many-though not all-parts of the world, and higher estimates of forest bio

mass.

Halocarbons.-The revised scenarios for CFCs and other substances which deplete stratospheric ozone are much lower than in SA90. This is consistent with wide participation in the controls under the 1990 London Amendments to the Montreal Protocol. However, the future production and composition of CFC substitutes (HCFCs and HFCs) could significantly affect the levels of radiative forcing from these compounds.

Methane, Nitrous Oxide, Ozone Precursors and Sulphur Gases.-The distribution of CH4 and N2O emissions from the different sources has changed from the SA90 case. Methane from rice paddies are lower, and emissions from animal waste and domestic sewage have been added. N2O emission factors for stationary sources and biomass burning have been revised downwards. Adipic and nitric acid have been included as additional sources of N2O. Preliminary analysis of the emissions of volatile organic compounds and sulphur dioxide suggests that the global emissions of these substances are likely to grow in the coming century if no new limitation strategies are implemented.

(iv) Relationship Between Emissions and Atmospheric Concentrations and the Influence on the Radiative Budget

A key issue is to relate emissions of greenhouse gases, greenhouse gas precursors and aerosol precursors to future concentrations of greenhouse gases and aerosols in order to assess their impact on the radiative balance. A number of different types of model have been developed.

Carbon Cycle Models

While there is a variety of carbon cycle models (including 3-D ocean-atmosphere models, 1-D ocean-atmosphere box-diffusion models, and box models that incorporate a tenestrial biospheric sink) all such models are subject to considerable uncertainty because of an inauate understanding of the processes controlling the uptake and release of CO2 from the oceans and terrestrial ecosystems. Some models assume a net neutral terrestrial biosphere, balancing fossil fuel emissions of CO2 by oceanic uptake and atmospheric accumulation, others achieve balance by invoking additional assumptions regarding the effect of CO2 fertilization on the different parts of the biosphere. However even models that balance the past and contemporary carbon cycle may not predict future atmospheric concentrations accurately because they do not necessarily represent the proper mix of processes on land and in the oceans. The differences in predicted changes in CO2 concentrations are up to 30%. This does not represent the major uncertainty in the prediction of future climate change compared with uncertainties in estimating future patterns of trace gas emissions, and in quantifying climate feedback processes. A simple empirical estimate can be based on the assumption that the fraction of emissions which remains in the atmosphere is the same as that observed over the last decade; i.e., 46±7%.

Atmospheric Gas Phase Chemistry Models

Current tropospheric models exhibit substantial differences in their predictions of changes in O3, in the hydroxyl radical (OH) and in other chemically active gases due to emissions of CH., nonmethane hydrocarbons, CO and, in particular, NO,. These arise from uncertainties in the knowledge of background chemical composition and our inability to represent small-scale processes occurring within the atmosphere. These deficiencies limit the accuracy of predicted changes in the abundance and distribution of tropospheric O,, and in the lifetimes of a number of other greenhouse gases, including the HCFCs and HFCs, all of which depend upon the abundance of the OH radical. Increases in CH., NMHCs, and CO all lead to increases in O,, and

decreases in OH, thus leading to an increase in radiative forcing. On the other hand because increases in NO, lead to an increase in both O, and OH, the net effect on radiative forcing is uncertain.

Atmospheric Sulphate Aerosol Models

The atmospheric chemistry of sulphate aerosols and their precursors has been extensively studied in relation to the acid rain issue. While our understanding of processes related to chemical transformations has increased significantly in recent years, substantial uncertainties remain, especially regarding the microphysics of aerosol formation, interaction of aerosols with clouds, and the removal of aerosol particles by precipitation.

(v) How has our Understanding of Changes in Radiative Forcing Changed?

Since IPCC (1990), there have been significant advances in our understanding of the impact of ozone depletion and sulphate aerosols on radiative forcing and of the limitations of the concept of the Global Warming Potential.

Radiative Forcing due to Changes in Stratospheric Ozone

For the first time observed global depletions of O3 in the lower stratosphere have been used to calculate changes in the radiative balance of the atmosphere. Although the results are sensitive to atmospheric adjustments, and no GCM studies of the implications of the O, changes on surface temperature have been performed, the radiative balance calculations indicate that the O, reductions observed during the 1980s have caused reductions in the radiative forcing of the surfacetroposphere system at mid- and high-latitudes. This reduction in radiative forcing resulting from O3 depletion could, averaged on a global scale and over the last decade, be approximately equal in magnitude and opposite in sign to the enhanced radiative forcing due to increased CFCs during the same time period. The effect at high latitudes is particularly pronounced and, because of these large variations with latitude and region, studies using GCMs are urgently required to further test these findings.

Radiative Forcing due to Changes in Tropospheric Ozone

While there are consistent observations of an increase in tropospheric ozone (up to 10% per decade) at a limited number of locations in Europe, there is not an adequate global set of observations to quantify the magnitude of the increase in radiative forcing. However, it has been calculated that a 10% uniform global increase in tropospheric ozone would increase radiative forcing by about a tenth of a wart per square metre.

Radiative Effects of Sulphur Emissions

Emissions of sulphur compounds from anthropogenlc sources lead to the presence of sulphate aerosols which reflect solar radiation. This is likely to have a cooling influence on the Northern Hemisphere (there is negligible effect in the Southern Hemisphere). For clear-sky conditions alone, the cooling caused by current rates of emissions has been estimated to be about 1Wm2 averaged over the Northern Hemisphere, a value which should be compared with the estimate of 2Wm-2 for the heating due to anthropogenic greenhouse gas emissions up to the present. The non-uniform distribution of anthropogenic sulphate aerosols coupled with their relatively short atmospheric residence time produce large regional variations in their effects. In addition, sulphate aerosols may affect the radiation budget through changes in cloud optical properties.

Global Warming Potentials

Gases can exert a radiative forcing both directly and indirectly: direct forcing occurs when the gas itself is a greenhouse gas; indirect forcing occurs when chemical transformation of the original gas produces a gas or gases which themselves are greenhouse gases. The concept of the Global Warming Potential (GWP) has been developed for policymakers as a measure of the possible warming effect on the surface-troposphere system arising from the emission of each gas relative to CO2. The indices are calculated for the contemporary atmosphere and do not take into account possible changes in chemical composition of the atmosphere. Changes in radiative forcing due to CO2, on a kg basis, are non-linear with changes in the atmospheric CO2 concentrations. Hence, as CO2 levels increase from present values, the GWPS of the non-CO2 gases would be higher than those evaluated here. For the concept to be most useful, both the direct and indirect components of the GWP need to be quantified.

Direct Global Warming Potentials.-The direct components of the Global Warming Potentials (GWPs) have been recalculated, taking into account revised estimated lifetimes, for a set of time horizons ranging from 20 to 500 years, with CO2 as the

reference gas. The same ocean-atmosphere carbon cycle model as in IPCC (1990) has been used to relate CO2 emission to concentrations. Table 3 shows values for a selected set of key gases for the 100 year time horizon. While in most cases the values are similar to the previous IPCC (1990) values, the GWPS for some of the HCFCs and HFCs have increased by 20 to 50% because of revised estimates of their lifetimes. The direct GWP of CH, has been adjusted upward, correcting an error in the previous IPCC report. The carbon cycle model used in these calculations probably underestimates both the direct and indirect GWP values for all non-CO2 gases. The magnitude of the bias depends on the atmospheric lifetime of the gas, and the GWP time horizon.

Indirect Global Warming Potentials.—Because of our incomplete understanding of chemical processes, most of the indirect GWPs reported in IPCC 1990) are likely to be in substantial error, and none of them can be recommended. Although we are not yet in a position to recommend revised numerical values, we know, however, that the indirect GWP for methane is positive and could be comparable in magnitude to its value. In contrast, based on the sub-section above, the indirect GWPS for chlorine and bromine halocarbons are likely to be negative. The concept of a GWP for short-lived, inhomogeneously distributed constituents, such as CO, NMHC, and NO, may prove inapplicable, although, as noted above, we know that these constituents will affect the radiative balance of the atmosphere through changes in tropospheric ozone and OH. Similarly, a GWP for SO2 is viewed to be inapplicable because of the non-uniform distribution of sulphate aerosols.

PART III: SUMMARY

It is clear in the 1992 IPCC supplement that there are significant scientific uncertainties associated with our assessment of global climate change. Resolution of these scientific uncertainties will take years to decades even with comprehensive research programs such as the U.S. Global Change Research Program (USGCRP) that are focussed on resolving the key IPCC scientific uncertainties; i.e:

clouds, which control the magnitude of climate change

• sources and sinks of greenhouse gases and aerosols, which control future atmospheric concentrations, hence the magnitude of climate change

⚫ oceans, which control the timing and regional patterns of climate change

• land-surface hydrological processes, which control water availability and regional patterns of climate change

• cryosphere, which affects sea level rise and regional patterns of climate change • ecological systems, which affect, and are affected by climate change

The USGCRP will provide the scientific information needed to formulate mitigation and/or adaptation policies.

Consequently, near-term national and international policies will have to be formulated recognizing that these scientific uncertainties exist and that the magnitude and rate of global warming may have been overestimated or underestimated. The scientific evidence for global warming, coupled with the long atmospheric lifetimes of greenhouse gases like carbon dioxide, suggests that the U.S. approach of identifying and implementing energy conservation, energy efficiency, and reforestation measures that reduce net greenhouse gas emissions is prudent.

The CHAIRMAN. Thank you very much, Dr. Watson.

Next we will hear from Dr. Richard Lindzen, who is Professor of Meteorology at Massachusetts Institute of Technology and who is also on the Scientific Advisory Board of the Marshall Institute. Dr. Lindzen.

STATEMENT OF DR. RICHARD S. LINDZEN, ALFRED P. SLOAN PROFESSOR OF METEOROLOGY, MASSACHUSETTS INSTITUTE OF TECHNOLOGY, CAMBRIDGE, MA

Dr. LINDZEN. Mr. Chairman, members of the committee, I am pleased to be here.

I am on the Scientific Advisory Board of the Marshall Institute. I must say immediately that I have seen this document no sooner than the rest of you and was not an author. I have just read it. I must say that in connection with this-and I will only say a few words I tend to agree with Bob Watson with one possible ex