Climate Change 2001: Working Group III: Mitigation

	Other reports in this collection

2.5.2.3 Comparison of Quantified Stabilization Scenarios

Based on the storylines, 76 stabilization scenarios were quantified as shown in Table 2.6. The assessment of the post-SRES work in this section is restricted to the analysis of CO₂ emissions and energy use in the different model runs. The detailed comparison of macroeconomic costs of reducing CO₂ emissions costs is not dealt with here: Chapter 8 addresses this aspect of stabilization.

Figure 2.13 shows the CO₂ emission trajectories of the 76 post-SRES mitigation scenarios along with the ranges of SRES and other published scenarios. Quantifications differ with respect to the baseline scenario including assumed storyline, the stabilization target, and the model that was used. As shown in Figure 2.13, the post-SRES scenarios cover a very wide range of emission trajectories, but the range is relatively below the SRES range, and they are apparently classified into groups according to the different stabilization targets. The figure shows the WRE late-response scenario and WGI early-action scenario for 550ppmv stabilization to compare with post-SRES scenarios, and it shows that the post-SRES range covers a much wider range than that between WRE and WGI.

Figure 2.14 shows the comparison of SRES and post-SRES scenario ranges in total global CO₂ emissions. The post-SRES ranges are estimated based on the selected scenarios quantified by SRES participants in order to compare the formal SRES ranges in Nakicenovic et al. (2000). It is shown clearly in the figure that concentration stabilization requires much more reduction of CO₂ emissions under development paths with high emissions such as A1FI and A2 than under development paths such as B1 and B2. These differences in reduction requirements result in selection of different technology and/or policy measures and, as a consequence, different costs to stabilize concentrations even at the same level. In the A1 scenario family, with its different scenarios in technological development (A1B, A1FI, and A1T), technological change is also a key component in bringing down the costs of mitigation options and their contribution to the emissions reduction. The A1FI stabilization scenarios, which are based on the highest baseline emissions, require much larger emission reductions than the A1T stabilization scenarios. The role of technology has been found to be crucial in the A1 scenario variants.

Morita et al. (2000a) compared all the stabilization variants in detail and found several common characteristics among these scenarios. These findings are as follows:

• Comparing the CO₂ emissions reductions from SRES baselines, the models have many points in common, but there are also some clear differences. All models show an increase in CO₂ reduction over time. This reflects the strong constraint of atmospheric CO₂ concentration. There is a considerable range in reductions among models in early years. However, most models achieve a similar proportional reduction from the baseline over the observation period.
  
• For achieving stabilization at 550ppmv, the highest reductions in CO₂ emissions compared to the baseline are observed in the A2 family. B1 shows the lowest reductions. CO₂ reduction at the end of the 21^st century ranges in A2 between 75% and 80%, A1B between 50% and 75%, B2 between 40% and 70%, and B1 between 5% and 40%.
  
• The target stabilization level also significantly affects the CO₂ reduction, even when based on the same baseline scenario. In the 450ppmv stabilization case, the reduction reaches 70% to 100%¹⁴ of A1B baseline emissions at the end of the 21^st century.
  
• Energy consumption reductions are more complicated among models. There is no strong relationship between the level of energy consumption and the stabilization level.
  
• Different baselines lead to different macroeconomic costs in order to reach a stabilization target. In spite of the wide range among models, A2 would be the most expensive case while B1 would require the lowest cost for stabilization at 550ppmv. The GDP loss in B1 would be less than one-tenth that in the A1B case, and less than one-twentieth that in the A2 case.
  
• The CO₂ reduction and macroeconomic costs are also significantly affected by the target stabilization level, even when based on the same baseline scenario. The economic cost for 450ppmv stabilization would be around three times that for 550ppmv, and six to eight times that for 650ppmv. These relationships can be observed at both the global and regional levels.
  
• Different stabilization targets also require different timing for the introduction of reduction policies. The 450ppmv stabilization case requires drastic emission reductions that occur earlier than under the 650ppmv case. Very rapid increases in emission reduction over 20 to 30 years are also observed in the 450ppmv stabilization case.

In order to compare the scenarios in further detail, several indices were calculated for this review.

First, a CO₂ reduction index was compared among stabilization levels as well as among SRES worlds. This index is calculated by subtracting baseline emissions from mitigation scenario emissions. In general, the lower the stabilization level that is required, as well as the higher the level of baseline emissions caused by the selected development path, the larger the CO₂ divergence from the baseline that is needed in all the regions. However, it does not simply follow from the larger divergence in emissions that there is an earlier divergence from the baseline.

The impact on the timing of emission reduction of both the stabilization level and the baseline level of emissions is further elaborated in Figure 2.15. This figure shows when the reduction in energy-related CO₂ emissions in each stabilization scenario would reach 20% of baseline emissions. This figure indicates that more stringent stabilization targets require earlier emission reductions from baseline levels. Higher emission worlds such as A1F1 and A2 also require earlier reduction than lower emission worlds such as A1T and B1.

A key policy question is what kind of emission reductions would be needed in the medium term, after the commitment period of the Kyoto Protocol (assuming that it will be implemented). Figure 2.16 shows the percent reduction in energy-related CO₂ emissions in Annex I countries from 1990 for the various stabilization cases. Since the first commitment period of the Kyoto Protocol ends in 2012, this can give some indication of the extent to which emission reduction commitments after 2012 would be needed to achieve the various stabilization levels. It should be noted that about two thirds of the scenarios assume that developing countries have already diverged from their baseline emission trajectories in 2020. Another point is that the post-SRES scenarios were not developed specifically to include the Kyoto targets, so there is a range of Annex I emission reductions (from 1990 levels) in 2010, 2020 and 2030. The mid-course scenarios are indicated in Figure 2.16 as the range between the 25^th and 75^th percentiles of the frequency distribution of the scenarios.

Figure 2.16 shows that:

• In the 450ppmv stabilization scenarios, the middle range (between the 25^th and 75^th percentiles) of Annex I emissions in 2010 lies between the Kyoto target and a 19% reduction from 1990 levels. This range increases after 2010, as does the decrease in Annex I emissions that would be needed to achieve stabilization at 450 ppmv. The percent reduction from 1990 levels in the middle range of scenarios is 13%–34% in 2020 and 11%–52% in 2030;
  
• In the 550ppmv stabilization scenarios, the middle range of Annex I emissions in 2010 is around the Kyoto target (from an 11% decrease to a 5% increase from 1990 levels); in 2020, the middle range of emissions lies between a 17% decrease and an 8% increase from 1990 levels; and in 2030, it lies between an 18% decrease and an 8% increase from 1990 levels. The average level of emissions slightly decreases after 2010; and
  
• The 650 or 750ppmv stabilization scenarios show similar changes in emission levels in 2010 compared to 1990, and few of them show any additional reduction in Annex I emissions after 2010. The middle range of emissions lies between an increase of 1%–17% from 1990 levels in 2020, and an increase of 4%–21% from 1990 levels in 2030.

This suggests that achievement of stabilization at 450ppmv will require emissions reductions in Annex I countries by 2020 that go significantly beyond their Kyoto Protocol commitments for 2008 to 2012.¹⁵ It also suggests that it would not be necessary to go much beyond the Kyoto commitments for Annex I countries (assuming as indicated that developing countries diverge from their baselines by 2020) to achieve stabilization at 550ppmv or higher. However, it should be recognized that several scenarios do indicate the need for significant emission reductions by 2020 in order to achieve these stabilization levels. These findings should be interpreted in light of the facts that CO₂ concentrations are assumed to reach one of the alternative fixed target levels in the year 2150, and unlike “emission corridor” analyses, these scenarios do not introduce other conditions such as a constraint on the rate of temperature increase.

Another important policy question concerns the participation of developing countries in emission mitigation. As a first step in addressing this question, the post-SRES scenarios were evaluated according to when per capita CO₂ emissions in Annex I countries would fall below per capita emissions in non-Annex I countries, assuming that all CO₂ emission reduction necessary for stabilization would occur in Annex I countries and that non-Annex I countries would emit CO₂ without any controls. This hypothetical assumption permits the analysis of one of the determinants of when non-Annex I emissions might begin to diverge from baseline levels. The results are shown in Figure 2.17 for each stabilization level and for three groups of SRES baselines.

Figure 2.17 shows that:

• Assuming that all the CO₂ reductions for concentration stabilization are undertaken in Annex I countries, most of the post-SRES scenarios indicate that per capita Annex I emissions would fall below per capita non-Annex I emissions in the 21^st century. This situation occurs before 2050 in two-thirds of the scenarios. Only in the A1T or B1 worlds would per capita CO₂ emissions in developing countries remain below those of developed countries in the 21^st century.
  
• These timings are significantly affected by the time series of emission reductions in the scenarios, and consequently they diverge in the scenarios. However, comparison within individual models suggests that the lower the stabilization level, the earlier that Annex I per capita emissions fall below non-Annex I per capita emissions. Stabilization scenarios based on higher emission worlds such as A1FI and A2 also tend to show earlier timing for Annex I to fall below non-Annex I per capita emissions compared to scenarios based on the lower emission worlds of B1 or A1T. This suggests that the stabilization target and the baseline emission level are both important determinants of the timing when developing countries’ emissions might need to diverge from their baseline emissions.

In order to assess priority setting in energy intensity reduction or in carbon intensity reduction, a “response index” was calculated for all stabilization variants of post-SRES scenarios for the years 2020, 2050, and 2100, as shown in Figure 2.18. This index relates the impact on CO₂ emission reduction of switching towards low-carbon or carbon free energy to the impact of energy intensity reduction. The response index is the ratio of the change in carbon intensity to the change in primary energy intensity¹⁶.

When energy intensity reduction is relatively larger than carbon intensity reduction, the index shows more than 1.0, and less than 1.0 in the opposite case.

It is clear from Figure 2.18 that the priority of response to reduce CO₂ emissions would change over time. Energy intensity reduction would be relatively larger than carbon intensity reduction in the beginning of 21^st century, but these would be of equal weight by the middle of the century. The impact of energy intensity reduction would be saturated towards the end of the 21^st century, and the use of low-carbon or carbon-free energy sources would become relatively much larger. This pattern is generally consistent across the stabilization levels. The lower the stabilization target, the higher the relative importance of energy intensity reduction in the beginning of the 21^st century, and the higher the relative importance of low-carbon or carbon free energy towards the end of the 21^st century.

These trends are important, but it is necessary at the same time to understand the model assumptions behind them. Most of the models do not accommodate very well structural and consumption-pattern-related efficiency measures (e.g., advanced dematerialization, major structural change, and changes in consumption patterns and lifestyles). A few cases which incorporate drastic changes in social structure (e.g., some of the scenarios based on AIM and WorldScan) give relatively high priority to energy efficiency improvement even in the latter half of 21^st century.

A per capita final energy index was calculated in order to analyze equity between North and South. Since one of the weak points of quantified scenario analysis concerns equity or “burden sharing”, the comparison of this kind of index is very important. Even though the per capita income is the most popular index to analyze equity, this index was not estimated by all the modelling teams. Therefore, final energy consumption per person in each region was adopted as an appropriate index for the equity analysis, because this index is closely related to per capita economic welfare. Figure 2.19 shows this index (in GJ/capita) among the OECD, EFSU, ASIA and ALM regions¹⁷ for all post-SRES and SRES variants over the period 1990 to 2100.

As shown in this figure, some interesting trends can be observed:

• In the development-emphasized worlds (A1B and A2) climate policy would reduce per capita final energy in both the Annex I and non-Annex I countries, while in the environment-emphasized worlds (B1 and B2) climate policy would have little effect on energy use. These impacts would slightly improve equity in per capita final energy use between the Annex I and non-Annex I countries, because the reduction in energy use caused by climate policies would be larger in Annex I than in non-Annex I.
  
• However, the impact of climate policies on equity in per capita final energy use would be much smaller than that of the future development path. The differences among the various SRES baseline conditions have the largest impact upon whether per capita energy use values converge between Annex I and non-Annex I countries, with the highest degree of convergence occurring in the A1B and B1 worlds. This can be seen in Figure 2.19 by comparing the (smaller) change in energy use between regions within each of the four columns (i.e., between the baseline and the 55ppmv stabilization scenario for each world) with the (much larger) change between regions across each of the two rows (i.e., across the baseline or across the 550ppmv stabilization scenario).

Though the analyses described above mainly focus on CO₂ emissions from energy consumption, it is also important to consider non-CO₂ emissions as well as non-energy-related CO₂ emissions. However, very few scenarios that include these emissions have been quantified and therefore it was not possible to include this additional review in this report. Some of the nine modelling approaches used here do include other radiatively active gases. However, the mitigation and/or stabilization scenarios include explicit limitations only on CO₂ emissions, and hence the reductions in other gases are indirect results (or ancillary benefits) of the CO₂ reduction measures.