30

A representative, but necessarily incomplete sample of the potential impacts highlighted in SAR WGII includes:

(a) Forests: Changes in temperature and water availability projected by general circulation models (GCMs) at equilibrium for doubled equivalent CO2 suggest that a substantial fraction (a global average of one-third, varying by region from one-seventh to two-thirds) of the existing forested area of the world will undergo major changes in broad vegetation types-with the greatest changes occurring in high latitudes and the least in the tropics. Climate change is expected to occur at a rapid rate relative to the speed at which forest species grow, reproduce, and re-establish themselves (SAR WGII: Summary for Policymakers (SPM) (Section 3.1) and Chapter 1). Multiple stresses to forests, including ozone and SO2 acidification, as well as climate and CO2 change, may have significant additional consequences;

(b) Mountain ecosystems: The altitudinal distribution of vegetation is projected to shift to higher elevation; some species with climatic ranges limited to mountain tops could become extinct because of disappearance of habitat or reduced migration potential (SAR WGII: SPM Section 3.1 and Chapter 5). The change in mountain ecosystems brings changes to the regulator function of altitudinal vegetation, altering the hydrological patterns in many regions;

(c) Aquatic and coastal ecosystems: The geographical distribution of wetlands is likely to shift with changes in temperature and precipitation. Some coastal ecosystems are particularly at risk, including saltwater marshes, mangrove ecosystems, coastal wetlands, sandy beaches, coral reefs, coral atolls and river deltas. Changes in these ecosystems would have major negative effects on tourism, freshwater supplies, fisheries and biodiversity (SAR WGII: SPM Section 3.1 and Chapters 6, 9 and 10);

(d) Hydrology and water resources management: Models project that between one-third and one-half of existing mountain glacier mass and a considerable area of permafrost could disappear over the next hundred years. The reduced extent of glaciers and depth of snow cover would also affect the seasonal distribution of river flow and water supply for hydroelectric generation and agriculture. Relatively small changes in temperature and precipitation, together with nonlinear effects on evapotranspiration and soil moisture, can generate relatively large changes in runoff, especially in semi-arid regions. The quantity and quality of water supplies already are serious problems today in many regions, including some low-lying coastal areas, deltas and small islands, which makes these regions particularly vulnerable to any additional reduction in indigenous water supplies (SAR WGII: SPM Section 3.2 and Chapters 7, 10 and 14);

(e) Food and fibre: Existing studies show that on the whole, global agricultural production could be maintained relative to baseline production in the face of climate change projected

under doubled equivalent CO2 equilibrium conditions. This conclusion takes into account the beneficial effects of CO2 fertilization but does not allow for changes in agricultural pests and the possible effects of changing climatic variability. However, there may be increased risk of bunger and famine in some locations; many of the world's poorest people particularly those living in subtropical and tropical areas and dependent on isolated agricultural systems in semi-arid regions are at the greatest risk (SAR WGII: SPM Section 3.3 and Chapters 13 and 16);

(Human infrastructure: Climate change will clearly increase the vulnerability of some coastal populations to flooding and erosional land loss. Some small island nations and other countries will confront greater vulnerability because their existing sea and coastal defence systems are less well established. Countries with higher population densities would be more vulnerable. Stormsurges and flooding could threaten entire cultures. For these countries, sea level rise could force internal or international migration (SAR WGII: SPM Section 3.4 and Chapters 9, 11, 12 and 17);

(g) Human health: Climate change is likely to have wide ranging and mostly adverse impacts on human health, with significant loss of life. Direct bealth effects include increases in mortality and illness (predominantly cardio-respiratory) due to an anticipated increase in the intensity and duration of heat waves. Temperature increases in colder regions should result in fewer cold-related deaths. Indirect effects of climate change, which are expected to predominate, include increases in the potential transmission of vector-borne infectious diseases (e.g., malaria, dengue, yellow fever and some viral encephalitis) resulting from extensions of the geographical range and season for vector organisms. Limitations on freshwater supplies and on nutritious food, as well as the aggravation of air pollution, will also have human health consequences (SAR WGII: SPM Section 3.5 and Chapter 18).

those impacts (e.g., value of a life, species loss, new species assemblages), and for combined market and non-market effects in some sectors (e.g., forest loss in lumber and public use value). Net climate change impacts include both market and non-market impacts, as far as they can be quantified, and in some cases include adaptation costs. Impacts are expressed in net terms to account for the fact that there may be some beneficial effects of climate change, even though this may obscure issues of distributional equity. The incomplete nature of the impact estimates presented here must be borne in mind when evaluating the full welfare implications of climate change.

The available studies reviewed in SAR WGIII estimate economic losses associated with a 2.5°C global warming (the mid-range estimate of equilibrium global temperature increase associated with a doubling of equivalent CO2 concentrations) on a world similar to today's (i.e., similar demographic characteristics, social structures, economic conditions) as follows:

(a) Developed country impact: 1-1.5 per cent of national GDP annually;

(b) Developing country impact: 2-9 per cent of national GDP annually.

The studies reviewed by WGIII aggregated these estimates in proportion to GDP, for a global total of 1.5-2 per cent GDP. These aggregated cost ranges are based on a large number of simplifying and controversial assumptions. They represent best-guess central estimates from relatively limited studies that attempt to include both market and non-market impacts, and in some cases also adaptation costs and they do not span the (large) range of uncertainty. The cost ranges are also imperfect in that GDP does not measure human and societal well-being accurately. Such aggregation faces numerous difficulties (SAR WGIII: Chapters 3 and 6) and was subject to severe reservations in the SAR WGIII Summary for Policymakers.

Existing estimates are rudimentary for several reasons. In addition to many of the problems that affect impact assessments in individual sectors, as noted above and in SAR WGII, additional uncertainties include:

(a) Estimates are predominantly for the United States and other Organisation for Economic Cooperation and Development (OECD) countries, and many regional and global estimates are based on extrapolations of these results. Material relating to other countries is sparse, although increasing. Hence, there is currently limited knowledge of regional and local impacts;

(b) Estimates of monetized impacts are for doubled equivalent CO2 concentration scenarios, usually based on the present day economy and expressed as a percentage of GDP. Simply projecting percentage losses is a somewhat unsatisfactory approximation, because future impacts will depend on economic, demographic and environmental

31

developments that will make future conditions very different from those of today. Some of the effects of climate change are likely to grow more than proportionately with GDP (e.g., the economic value of non-market goods) and others less than proportionately (e.g., agriculture);

(c) There are difficulties in measuring the economic value of impacts, even where the impacts are known. This is particularly the case for non-market impacts and the impacts in developing countries. Some regard monetary valuation of such impacts as essential to sound decision making, while others reject valuation of impacts, such as loss of human life or biodiversity, on ethical grounds;

(d) Calculating a global aggregate of impacts involves difficult questions about equity among countries, especially given income and other social differences. Simply aggregating GDP estimates means that equivalent impacts in two countries receive a different weight, based heavily on national economic product. The ethical issues involved in such aggregation raise difficulties of consistency that are not explicitly addressed in existing studies (SAR WGIII: Chapters 3 and 6);

(e) There are difficulties in setting discount rates, which are the analytical tool economists use to compare econom effects that occur at different points in time. This is important because climate change impacts are likely to impose costs on future generations.

The practical application of these estimates to climate change decision making is difficult, not only because of the uncertainty of the estimates themselves, but also because of the global and intergenerational nature of the problem. Some systems in some regions may benefit from climate change for some period of time, whereas many others will suffer adverse impacts; thus impacts will be distributed unequally. Climate change will affect an extremely diverse mix of human societies, some of which have less potential to adapt than others, and will thus suffer more than others. An evaluation entails trade-offs among impact categories, regions, nations, generations and individuals. Various techniques exist to make such trade-offs visible and manageable, but the actual decision regarding which impacts are most costly is a political one. Within well developed institutional/economic/political systems, mechanisms for making trade-offs and providing compensation to the losers exist. Internationally and inter-temporally, existing mechanisms are much weaker. Currently, the knowledge of climate change impacts is not sufficiently developed to make these trade-offs clear.

32

expected to have larger impacts on natural and human systems than trajectories that assume a slower accumulation of forcing and lower stabilized concentrations. It is not currently possible, however, to determine how the impacts that may be associated with one stabilization target or emissions trajectory may differ from those associated with another target or trajectory. For many reasons, there is not a simple relationship between emissions and atmospheric concentrations of greenhouse gases and aerosols, on the one hand, and potential impacts on the other. The reasons include:

(a) Altered patterns of radiative forcing and global mean changes in climate will have different effects on climate conditions in different regions. These local and regional conditions, including changes in the length of growing seasons, the availability of water, and the incidence of disturbance regimes (extreme high temperature events, floods, droughts, fires, and pest outbreaks) have important impacts on the structure and function of both natural and human-made environments;

(b) Some systems are more vulnerable to changes in regional climate than others e.g., human systems are more adaptive, bence on average less vulnerable, than natural systems; forested systems require longer periods than grassland systems to establish themselves and hence are less likely to be able to migrate to new locations with suitable conditions, as temperature and precipitation patterns shift;

(c) The relative vulnerability of individual regions is likely to vary. Typically, systems are more vulnerable in developing countries, where economic and institutional circunstances are less favourable than in developed countries. People who live in semi-arid regions, upland regions, low-lying coastal areas, water-limited or flood-prone areas, or on small islands are particularly vulnerable to climate change. Sensitive areas such as river flood plains and coastal plains have become more vulnerable to hazards such as storms, floods and droughts as a result of increasing population density and economic activity;

(d) Impacts are not a linear function of the magnitude and rate of change; for some species (and hence systems), thresholds of change in temperature, precipitation or other factors may exist which, once exceeded, lead to discontinuous changes in viability, structure or function. This suggests that small changes in local climates may have a disproportionately large impacts;

(e) Most existing studies are limited to analysis of impacts that would result from changes associated with a doubled equivalent CO2 equilibrium climate; very few studies have considered dynamic responses to steadily increasing concentrations of greenhouse gases or stabilization scenarios, and fewer still have examined the consequences of increases beyond a doubling of equivalent atmospheric CO2 concentrations. Even fewer studies have assessed the implications of multiple stress factors, such as O3, SO2 acidification, or other pollutant stressors in the presence of climate and CO2 change.

The costs of a carbon constraint depend on the emissions "baseline". i.e., how emissions are projected to grow in the absence of policy intervention. The higher the baseline, the more carbon must be removed to meet a particular stabilization target, thus the greater the need for intervention. Figure 16a shows anthropogenic CO2 emissions for the six IS92 baseline scenarios. The differences in emissions are generated by different assumptions about population, economic growth, the cost and availability of energy supply- and demand-side alternatives, and other factors.

333

emissions faster than reductions in energy intensity and fuelswitching to less carbon intensive sources reduce emissions.

The rising baseline does not imply that there are no economically-attractive alternatives to fossil fuels on either the supply-side or demand-side of the energy system. Such options typically are included in sizeable quantities in most economic analyses. A growing baseline only means that these options are not implemented at a rate sufficient to arrest the growth in carbon emissions. This may be due to an insufficient supply of no-regrets options.

Figure 16h translates the emission scenarios into CO2 concentrations. None of the six scenarios leads to stable concentrations before 2100, although IS92c leads to a very slow growth in CO2 concentration after 2050. IS92a, b, e and fall double the preindustrial CO2 concentration before 2070.

The Stabilization Target

The costs of a carbon constraint are also sensitive to the concentration stabilization target. As a first approximation, a stabilization target defines an amount of carbon that can be emitted between now and the date at which the target is to be achieved (the "carbon budget"). Table 5 shows the "carbon budgets" to the year 2100 associated with the 450, 550, 650, 750 and 1 000 ppmv stabilization profiles (see Figure 6 for the cumulative emissions from which the carbon budgets were derived). The lower the stabilization target, the smaller the carbon budget (i.e.. the smaller the cumulative emissions amount).

The size of this "carbon budget" is an important determinant of mitigation costs. Lower stabilization targets require smaller carbon budgets, which require a greater degree of intervention. Table 5 compares the carbon budget for the stabilization level and paths from Figures 5 and 6 to the accumulated anthropogenic CO, emissions for the IS92 emission scenarios.

3.2.1.3

Figure 16. (a) Total anthropogenic CO2 emissions under the IS92 emissions scenarios: (b) The deduced atmospheric CO, cogentrations for the IS92 emissions scenarios calculated using the Bern carbon cycle model (see SAR WGI (Section 2.1)) (taken from SAR WGI: Technical Summary).

Cost Differential Between Fossil Fuels and Carbon-free Alternatives

The cost of stabilizing CO2 concentrations also depends on the cost of fossil fuels relative to carbon-free alternatives. For a given energy demand, the cost of reducing energy-related CO2 emissions depends on the cost difference between the available fossil fuels and the carbon-free alternatives at the time when global CO, emissions are reduced.

The cost differential between conventional fossil fuels (e.g.. conventional crude oil, natural gas, and coal) and carbon-free alternatives is forecast to narrow, although how much remains uncertain and widely debated. During the next hundred years. the cost of conventional fossil fuels should increase as these resources are exploited, and the least expensive and most accessible coal deposits are mined. At the same time, improvement in

34

basic science, engineering, and institutional arrangements should reduce the cost of carbon-free technologies (and unconventional fossil fuels).

The degree to which cumulative emissions exceed conventional crude oil and natural gas resources gives some indication of the contribution these fuels make to total energy consumption (see Table 9 of the IPCC Technical Paper on Technologies, Policies and Measures for Mitigating Climate Change (IPCC TP P&M. 1997) for estimates of global energy reserves and resources). If cumulative emissions associated with a stabilization target are equal to or lower than the cumulative emissions that would result from the combustion of conventional oil and gas resources, these fuels will probably be an important component of total energy supply during the transition period to carbonfree alternatives. On the other hand, if cumulative emissions associated with a stabilization target are significantly greater than the cumulative emissions that would result from the combustion of conventional crude oil and natural gas resources, these fuels will probably be a relatively small component of total energy supply during the transition period. The cost difference between fossil fuels and carbon-free alternatives will be smaller in the latter case. While the cost premium for carbonfree alternatives is likely to be smaller for higher stabilization levels, total energy demand is higher so the net effect on transition costs is not clear.

However, we cannot predict how the absolute level of the cost differential between unconventional fossil fuels and carbon-free alternatives will change over time. Technical change will probably reduce the costs of unconventional fossil fuels and carbon-free alternatives, but the rate of technical change is

likely to differ. Technical gains that reduce the costs of unconventional fossil fuels relative to carbon-free alternatives will increase transition costs by increasing the cost differential between fossil fuels and carbon-free alternatives, whereas technical changes that reduce the costs of carbon-free technologies have the opposite impact.

Differences between the costs of available fossil fuels affect transition costs in a similar manner.

3.2.1.4 The Emissions Pathway

As indicated in Figure 5 and described in Section 2.2.1.2, the same concentration target (see Figure 4) can be achieved through several emission pathways. Emissions in the near-term can be balanced against emissions in the long-term. On the other hand, higher early emissions decrease the options to adjust emissions later on. In Figure 5, the dashed lines (the WRE profiles) show higher emissions in the early years. although a more rapid transition from increasing to decreasing emissions. The pathways associated with the solid lines (the S profiles) allow higher emissions later on, but have lower emissions in the early years. Thus, as explained in Section 2.2.1.2, for a given stabilization level, there is a "budget" of allowable accumulated carbon emissions and the choice of pathway to stabilization can be viewed as a problem of how to best (i.e.. with the greatest economic efficiency and least damaging impacts) allocate this carbon budget over time.

The differences in the emission paths for the same stabilization level are important because costs differ among pathways. SAR WGIII identifies the following factors that affect the costs of alternative pathways: (a) the treatment of existing and future

both known and unknown, that remain to be combusted.

discount rate; and (ɗ) the carbon budget.