The stabilization scenarios that were estimated based on the above baselines also have a very wide range, as shown in Figure 2.3. This wide range is caused by several factors, including differences in emission time-paths for the stabilization, differences in timing of the stabilization at 550 ppmv, and different carbon cycle models used to assess the stabilization.
The divergence in reduction time-path has been discussed based on two sets of popular scenarios. One is a set of IPCC Working Group (WG) I scenarios (Houghton et al., 1996) which is sometimes referred to as early action scenarios and denoted as WGI; the other is a set of scenarios published by Wigley et al. (1996), sometimes referred to as delayed action scenarios and denoted WRE. Chapter 8 explains that these terms are misleading, since WRE scenarios may not assume early emissions reductions, but do assume early actions to facilitate such reductions later. Figure 2.3 compares the 550 ppmv stabilization scenarios of these two scenario sets with the reviewed scenarios, and it shows that scenarios reviewed here cover a wider range than that of the WGI and WRE scenarios. While the RICE and MERGE scenarios show late reduction (WRE type) trajectories, the CETA, MARIA and MIT scenarios show more severe reduction (WGI type) trajectories.7 A few scenarios, for example ICAM2, show no drastic reduction even in the latter half of the 21st century. Most of the scenarios have emissions trajectories that lie in between.
The reduction time-path of emissions is a controversial point, which is closely related to the intergenerational equity issue. However, no conclusion can be drawn from such global trajectories, since behind them lies a distribution between countries and the political, technical, economic, and social acceptability of this distribution would depend on how the equity concerns are sorted out.
Figures 2.6 and 2.7 show energy-related CO2 reduction at the global and the non-OECD levels, respectively, which were estimated for each scenario source by subtracting stabilization scenario emissions (Figure 2.3) from baseline scenario emissions (Figure 2.2). These figures show that the range of reduced CO2 emissions for 550ppmv stabilization is also very wide both at the global and the non-OECD levels. This wide range is apparently caused by the divergent baseline scenarios shown in Figure 2.2, while other factors such as differences in emission time-path, in timing of stabilization and in the carbon cycle model used also tend to increase the range.
Figures 2.6 and 2.7 show the simulation results of models, assuming that non-OECD countries would participate in mitigation. The distribution of mitigation among the countries is based on different approaches, such as the introduction of emission caps, or the assumption of the same rate of emission reduction for all countries, or global emission trading. The results show that emission trading may lower the mitigation cost, and could lead to more mitigation in the non-OECD countries.
The regional allocation of reductions is a controversial and highly political issue from the equity viewpoint. Mostly, modellers do not explicitly state the burden-sharing rule. Nevertheless, the emission reduction from baseline by the non-Annex I countries is a good indicator of when it is assumed that these countries start sharing the reductions. The data set used in this analysis is limited in the sense that models have different regional specifications; it was therefore difficult to obtain a large number of data points to analyze non-Annex I emissions. As a proxy, emission reduction from the baseline by the non-OECD region is used, which includes Russia and Eastern Europe. This is shown in Figure 2.7. In part of the AIM, MiniCAM, FUND, and PEF scenarios, introduction of climate policy in the non-OECD region is assumed not to begin by 2010. Although Russia and Eastern European countries are included in the Kyoto Protocol, the models do assume that because of the decreased emissions in these countries since 1990, actual climate policies would not be needed until 2010. Some scenarios show that non-OECD regions may not have to significantly reduce emissions before 2030. However, there are still other scenarios that show an opposite picture. The RICE, MERGE, MIT, and MARIA scenarios show a very steep increase in emission reduction from baseline levels in the non-OECD region starting very early in the 21st century.
One of the ways to explain this divergence in reduction time series is to differentiate the assumptions about trade in these scenarios. Some scenarios assume trade in emission credits, which are allotted initially to each country or region. This allows some countries to purchase emission rights from other countries to minimize the cost of meeting their emission targets. The dotted lines in Figure 2.7 show the scenarios that assume trade in emission credits between the Annex I and non-Annex I countries. The scenarios that show an early reduction of emissions in the non-OECD region are included in the trade scenarios, and they assume the OECD region would transfer funds to the non-OECD region via emission credit trading. Most of the other scenarios assume that the non-OECD region would start to introduce reduction policies after 2010.
With regard to overall mitigation, the range of assumed policies is very wide, resulting in a wide range of emission reductions. The additional increase in energy efficiency improvement from the baseline ranges between minus 0.04 and 1.56% per year within the sampled data, while the additional reduction in carbon intensity from the baseline is between zero and 3.76% per year. Although it is difficult to identify detailed policy assumptions from the database, the range of these factors suggests divergent policy options among scenarios. These policy options are dependent not only on the level of CO2 reduction, but also on the baseline scenarios that have been used for 550 ppmv stabilization quantification.
Figure 2.8 (a) shows the relationship between the effects of efficiency improvement policy in mitigation scenarios and the energy intensity reduction assumption in baseline scenarios. This figure suggests an inverse relationship between them. The implication of this is that scenarios in which there is an assumed adoption of high-efficiency measures in the baseline usually would have less scope for further introduction of efficiency measures in the mitigation scenarios, as compared to scenarios that have a lower level of efficiency improvement in their baseline.8 As a result, the additional reduction of energy intensity in mitigation scenarios over the base cases would be lower when the assumed energy intensity reduction is high in the base case, and vice versa. In the case of unanticipated technological breakthroughs, of course, this relationship may not hold and one could expect further energy efficiency improvements, even when the baseline has a fair amount of energy efficiency built into it.
Figure 2.8 (b) shows the relationship between the effects of decarbonization policies and the carbon intensity reductions assumed in the baseline scenarios. This figure suggests that baseline scenarios with high carbon intensity reductions show larger carbon intensity reductions in their mitigation scenarios, while those with low carbon intensity reductions in the base case show smaller reductions in carbon intensity in their corresponding stabilization cases. This is somewhat counterintuitive and difficult to explain simply on the basis of the results available. One might expect that high carbon intensity reductions in the base case might use up decarbonization potential, giving rise to lower additional reduction of carbon intensity in mitigation scenarios. On the other hand, increased investment in low-carbon energy technology in the base case could increase the resource base of low-carbon energy, thereby providing more opportunity to reduce CO2 emissions in the stabilization case. The mitigation potential in this direction depends not only on the technology but also, and perhaps more, on the economics and social acceptance of the technology. A closer and more careful analysis of which particular mitigation policies were assumed in constructing the scenario than was possible on the basis of the available information, would reveal the underlying reasons for such a pattern.
Finally, Figure 2.8 (c) shows the relationship between macroeconomic costs9 in the mitigation scenarios and GDP growth assumptions in the baseline scenarios. No clear relationship is visible, but it can be observed that macroeconomic costs for the world as a whole are estimated to range between 0% and 3.5% of GDP in 2100, while a few simple models estimate more increase in the second half of the 21st century. The GDP loss may or may not be related to the GDP growth assumptions in baselines. For instance, high baseline economic growth would lead to higher emissions of GHGs, which would lead to increased GHG reduction costs compared to the corresponding mitigation scenario for a low-growth baseline. On the other hand, high economic growth could provide increased funds for research and development (R&D) of advanced technologies, which would decrease the cost of GHG reduction. The net cost would depend on the relative strengths of these effects. Another aspect is that the costs are also dependent upon the structure of economies, i.e., economies with high fossil fuel dependence, via either exports or domestic consumption, are likely to experience higher costs compared with economies with relatively lower fossil fuel dependence.
Other reports in this collection