Scenario analysis 231 Overview

The scenario analysis presented within this paper is for the basket of six greenhouse gases only and relies, principally, on the scientific understanding contained within

AR4. The analysis does not take account of the following:

• the radiative forcing impacts of aerosols and non-CO2 aviation emissions (e.g. emissions of NOx in the upper troposphere, vapour trails and cirrus formation);5

• the most recent findings with respect to carbon sinks;6

• previously underestimated emission sources;7 and

• the implications of early emission peaks for 'overshooting' stabilization concentrations and the attendant risks of additional feedbacks.

3 CO2 data from the Carbon Dioxide Information Analysis Centre (CDIAC) including recent data from G. Marland (2006, personal communication); non-CO2 greenhouse gas data from the USA Environmental Protection Agency (EPA 2006) including the projection for 2005, and assuming deforestation emissions in 2005 to be 5.5 GtCO2 (1.5 GtC), with a 0.4 per cent growth in the preceding 5 years in line with data within the Global Forest Resources Assessment (FAO 2005).

4 FAO (2005) contains rates of tropical deforestation for the 1990s revised downward from those in the 2000 Global Forest Resources Assessment (FAO 2000; R. A. Houghton 2006, personal communication). An earlier estimate based on high-resolution satellite data over areas identified as 'hot spots' of deforestation, estimated the figure at nearer 3.7 GtCO2 (1 GtC) for 2000 (Achard et al. 2004). It is Houghton's more recent estimate that is used in this chapter.

5 There remains considerable uncertainty as to the actual level of radiative forcing associated with aerosols, exacerbated by their relatively short residence times in the atmosphere and uncertainty as to future aerosol emission pathways (Cranmer etal. 2001; Andreae etal. 2005). Similarly, there remain significant uncertainties as to the radiative forcing impact of non-CO2 emissions from aviation, particularly contrails and linear cirrus (e.g. Stordal et al. 2004; Mannstein & Schumann 2005).

6 For example, and in particular, the reduced uptake of CO2 in the Southern Ocean (Raupach et al. 2007) and the potential impact of low-level ozone on the uptake of CO2 in vegetation (Cranmer et al. 2001).

7 For example, significant uncertainties in the emissions estimates for international shipping (Corbett & Kohler 2003; Eyring et al. 2005).

While aerosols are most commonly associated with net global (or at least regional) cooling, the other factors outlined above are either net positive feedbacks or, as is the case for high peak-level emissions, increase the likelihood of net positive feedbacks. Consequently, the correlations between concentration and mitigation outlined in this analysis are, in time, liable to prove conservative.

The scenarios are for CO2e emission pathways during the twenty-first century, with empirical data used for the opening years of the century (in contrast to modelled or 'what if' data). The full scenario sets (AB1 and AB2) comprise different combinations of the following: (i) emissions of CO2 from deforestation, (ii) emissions of non-CO2 greenhouse gases and (iii) emissions of CO2 from energy and industrial processes.

For AB1

• Deforestation. Two low-emission scenarios for the twenty-first century.

• Non-CO2 greenhouse gases. Three scenarios peaking in 2015, 2020 and 2025 and subsequently reducing to 7.5 GtCO2e per year.

• Energy and process CO2. Three scenarios peaking in 2015, 2020 and 2025 and subsequently reducing to maintain the total cumulative emissions for the twenty-first century within the AR4 450 ppmv CO2e range (with carbon-cycle feedbacks).

For AB2

• Deforestation. Two low-emission scenarios for the twenty-first century.

• Non-CO2 greenhouse gases. One scenario peaking in 2020 subsequently reducing to 7.5 GtCO2e per year (as per AB1 with a 2020 peak).

• Energy and process CO2. Three scenarios, each following the same pathway to a 2020 peak, but subsequently reducing at different rates to maintain total annual CO2e reductions of 3, 5 and 7 per cent.

The following sections detail the deforestation and non-CO2 greenhouse gas emission scenarios used to derive the post-peak energy and process CO2 emission scenarios and ultimately the total global CO2e scenarios for the twenty-first century.

2.3.2 Deforestation emissions

A significant portion of the current global annual anthropogenic CO2 emissions are attributable to deforestation (in the region of 12-25 per cent). However, carbon mitigation policy, particularly in OECD nations, tends to focus on those emissions from energy and industrial processes (hereafter referred to as energy and process emissions), with less direct regard for emissions arising from deforestation. While the relatively high levels of uncertainty associated with deforestation emissions

Table 2.1 Deforestation emission scenario summary for two scenarios used to build the subsequent full CO2e scenarios (deforestation low, DL; deforestation high, Dh) and one for illustrative purposes only (deforestation very high, DVH)

2100 carbon stock

2000 emissions/year

remaining %

Emissions

(carbon stock)

Peak

(carbon stock)

2000-2100

Name

(GtCOa)

date

(GtCO2)

(GtCO2)

DL (developed for

5.5 (1060)

2015

80 (847)

213

this analysis)

Dh (Moutinho &

5.5 (1060)

2020

70 (741)

319

Schwartzman)

DVh (Moutinho &

5.5 (1060)

2036

55(583)

477

Schwartzman)

make their inclusion in global mitigation scenarios problematic, the scale of emissions is such that they must be included. Within this chapter two deforestation scenarios are developed; both assume climate change to be high on the political agenda and represent relatively optimistic reductions in the rate of, and hence the total emissions released from, deforestation.8 They both have a year 2000 baseline of 5.5 GtCO2, but post-2015 have different deforestation rates and hence different stocks of carbon remaining in 2100 (i.e. the amount of carbon stored in the remaining forest). The scenarios are illustrated numerically in Table 2.1 and graphically in Figure 2.1.

The scenarios are dependent not only on the baseline but also on estimates of the change in forestry carbon stocks between 2000 and 2100. The stock values used in the scenarios are taken from Moutinho & Schwartzman (2005) and based on their estimate of total forest carbon stock in 2000 of 1060 GtCO2. According to their assumptions, the carbon stock continues to be eroded at current rates until either 2012 or 2025, following which emissions from deforestation decline to zero by either 2100 or until they equate to 15 per cent of a particular nation's forest stock (compared with 2000). They estimate two values for the carbon stocks, released as CO2 emissions by 2100 as 319 and 477 GtCO2. This implies that within their scenarios, either 70 or 55 per cent of total carbon stocks remain globally. Given that this chapter and its accompanying AB1 and AB2 scenarios are premised on climate change being high on the international agenda, Moutinho & Schwartzman's

8 While the scenarios are at least as optimistic as those underpinning, for example, the 2005 Forest Resource Assessment (FAO 2005) and the 2006 Stern Report, it could be argued they are broadly in keeping with the high-profile deforestation gained during the 2007 United Nations Climate Change Conference in Bali.

Figure 2.1 Deforestation emission scenarios showing three CO2 emissions pathways based on varying levels of carbon stocks remaining in 2100.

55 per cent of total carbon stock value is considered too pessimistic within the context of this analysis, and although presented in Figure 2.1, is not included in the analysis from this point onwards. Moreover, to allow for a more stringent curtailment of deforestation, the scenario developed for a 70 per cent stock remaining estimate is complemented by one with 80 per cent remaining.

The Dl and DH curves both assume no increase in deforestation rates from current levels, with DL beginning to drop from the peak level of 5.5 GtCO2, 5 years prior to DH. This, combined with the higher level of forestry, and hence carbon stock remaining in 2100, gives the DL curve a faster rate of reduction in deforestation than is the case for the DH curve (typically, 7.4 and 4.8 per cent for Dl and DH, respectively).9

2.3.3 Non-CO2 greenhouse gas emissions

To estimate the percentage reductions required from energy and process CO2 emissions for both AB1 and AB2, it is necessary to consider a range of future emission scenarios for the non-CO2 greenhouse gases. Accordingly, three scenarios are developed assuming current US Environmental Protection Agency (EPA) estimates and projections of emissions from 2000 up to a range of peaking years, after which

9 Dl per cent change value is the mean for the period between 2030 and 2050, and Dh is the mean value for 2040-2060.

Table 2.2 Non-CO2 greenhouse gas emission scenario summary

Mean

Peak

Total

2000

growth

annual

2000-2100

emissions

Peak

to peak

emission

emissions

Name

(GtCO2)

year

(%)

(GtCO2e)

(GtCO2e)

Early action

9.5

2015

1.31

11.4

858

Mid-action

9.5

2020

1.51

12.2

883

Late action

9.5

2025

1.53

13.3

916

Figure 2.2 Three non-CO2 greenhouse gas emission scenarios with emission pathways peaking at different years but all achieving the same residual level by 2050.

emissions are assumed to decline towards the same long-term stable level. All the scenarios represent a long-term halving in emission intensity, with the difference between them arising from the range of cumulative emissions associated with each of the peaking dates. The scenarios are illustrated numerically in Table 2.2 and graphically in Figure 2.2.

Anthropogenic non-CO2 greenhouse gas emissions are dominated by methane and nitrous oxide and, along with the other non-CO2 greenhouse gases, accounted for approximately 9.5 GtCO2e in 2000 (EPA 2006; similar figures are used within the Stern Review), equivalent to 23 per cent of global CO2e emissions. Understanding how this significant portion of emissions may change in the future is key to exploring the scope for future emissions reduction from all the greenhouse gases.

The three non-CO2 greenhouse gas scenarios presented here are broadly consistent with a global drive to alleviate climate change. The principal difference between the scenarios is the date at which emissions are assumed to peak, with the range chosen to match that for the total CO2e emissions, namely an early-action scenario where emissions peak in 2015, a mid-action peak of 2020 and finally a late-action peak in 2025. All three scenarios have a growth rate from the year 2000 up until a few years prior to the peak, equivalent to that projected by the EPA (2006),10 and broadly in keeping with recent trend data. The scenarios all contain a smooth transition through the period of peak emissions and on to a pathway leading towards a post-2050 value of 7.5 GtCO2e. This value is again specifically chosen to reflect a genuine global commitment to tackle climate change. It is approximately 25 per cent lower than the current level and consistent with a number of other 450 ppmv scenarios.11 Given that the majority of the non-CO2 greenhouse gas emissions are associated with food production, it is not possible, with our current understanding of the issues, to envisage how emissions could tend to zero while there remains a significant human population. The 7.5 GtCO2e figure used in this paper, assuming a global population in 2050 of 9 billion (thereafter remaining stable), is equivalent to approximately halving the emission intensity of current food production. While a reduction of this magnitude may be considered ambitious in a sector with little overall emission elasticity, such improvements are necessary if global CO2e concentrations are to be maintained within any reasonable bounds.

The non-CO2 greenhouse gas scenarios have similar growth rates from 2000 to their respective peak values, and ultimately all have the same post-2050 emission level (7.5 GtCO2e). The rate of reduction in emissions from the respective peaks demonstrates the importance of timely action to curtail the current rise in annual emissions: the early-action scenario is required to reduce at 1.35 per cent per year, while the mid- and late-action scenario values are at 2 and 3 per cent, respectively. Similarly, Table 2.2 and Figure 2.2 demonstrate the importance for cumulative values of non-CO2 greenhouse gas emissions not rising much higher than today and that the post-peak reduction rate achieves the long-term residual emission level as soon as is possible (7.5 GtCO2e by 2050). If the year in which emissions reach the residual level had been 2100 rather than 2050, the modest differences in cumulative emissions between the early-, mid- and late-action scenarios would have been substantially increased. Given that the cumulative value of non-CO2 greenhouse

10 EPA values for global warming potential of the basket of six gases are slightly different from those used in

IPCC. The difference, though noted here, does not significantly alter the analysis or results.

11 For example, in Stern (2006, p. 233), for both his 450 ppmv CO2e and 500-450 ppmv overshoot curve.

gas emissions is a significant proportion of total cumulative CO2e emissions, any delay in achieving the residual value would have significant implications for the reduction rate of energy and process CO2 emissions necessary to meet the AB1 and AB2 criteria.

2.3.4 CO2e emission scenarios for the twenty-first century

Having developed the deforestation and non-CO2 greenhouse gas scenarios, this section presents the complete greenhouse gas emission scenarios, AB1 and AB2, for the twenty-first century. The emissions released from the year 2000 until the peak dates are discussed here in relation to both AB1 and AB2, before the post-peak scenarios for each of the scenario sets are presented.

AB1 and AB2: emissions from 2000 to the peak years

By combining the deforestation and non-CO2 greenhouse gas scenarios with assumptions about energy and process CO2, scenarios for all greenhouse gas emissions up until the three peaking dates are developed. Energy and process CO2 emissions for the years 2000-2005 are taken from the Carbon Dioxide Information Analysis Centre (CDIAC), with estimates for 2006-2007 based on BP inventories (BP 2007). From 2007 to the three peaking dates of 2015 (early action), 2020 (mid-action) and 2025 (late-action) emissions of energy and process CO2 grow at 3 per cent per year until 5 years prior to peaking. Beyond this point, emission growth gradually slows to zero at the peak year before reversing thereafter. The 3 per cent emission growth rate chosen for CO2 is consistent with recent historical trends. Between 2000 and 2005, CDIAC data show a mean annual growth in energy and process CO2 emissions of 3.2 per cent; this includes the slow growth years following the events of 11 September 2001.

AB1: emissions from peak years to 2100

From the peak years onwards, AB1 (summarized in Table 2.3) takes the approach that to remain within the bounds of a 450 ppmv CO2e stabilization target, the cumulative emissions between 2000 and 2100 must not exceed the range presented within the latest IPCC report in which carbon-cycle feedbacks are included (IPCC 2007b).

AB1 final scenarios

The emission pathways for the full greenhouse gas AB1 scenarios from 2000 to 2100 are presented in Figure 2.3. The plots comprise the earlier deforestation and non-CO2 greenhouse gas scenarios with growing energy and process CO2

Table 2.3 Summary of the core components of scenario set AB1

Characteristic

2015-2100

2020-2100

2025-2100

Deforestation" Non-CO2 greenhouse gases" Approximate peaking value

(GtCO2e) Cumulative emissions (GtCO2e) IPCC AR4

2100 residual emissions (GtCO2e)

Dh and Dl early action 54

low: 1376 medium: 1798 high: 2202 7.5

Dh and Dl mid-action

low: 1376 medium: 1798 high: 2202 7.5

Dh and Dl late action 64

low: 1376 medium: 1798 high: 2202 7.5

Deforestation and non-CO2 greenhouse gas scenarios as in Tables 2.1 and 2.2.

Figure 2.3 Greenhouse gas emission scenarios for AB1 with emissions peaking in (a) 2015, (b) 2020 and (c) 2025 (see also colour plate).

Table 2.4 Scenarios assessed in relation to their practical feasibility. (X denotes a scenario rejected on the basis of being quantitatively impossible or with prolonged percentage annual reduction rates greater than 15%. The percentage reductions given illustrate typical sustained annual emission reductions required to remain within budget.)

Table 2.4 Scenarios assessed in relation to their practical feasibility. (X denotes a scenario rejected on the basis of being quantitatively impossible or with prolonged percentage annual reduction rates greater than 15%. The percentage reductions given illustrate typical sustained annual emission reductions required to remain within budget.)

Peak date

Deforestation DL

Deforestation DK

Low

Medium

High

Low

Medium

High

2015

X

13%

4%

X

X

4%

2020

X

X

8%

X

X

11%

2025

X

X

X

X

X

X

emissions up to the peaking year, and all have total twenty-first century cumulative values of CO2e matching the 450 ppmv figures within AR4.

It is evident from the data underpinning Figure 2.3 that 10 of the 18 proposed pathways cannot be quantitatively reconciled with the cumulative CO2e emissions budgets for 450 ppmv provided within AR4. Table 2.4 identifies the 'impossible' scenarios (including three with prolonged annual reduction rates greater than 15%) and illustrates the post-peak level of sustained emission reduction necessary to remain within budget.

AB1: implications for energy and process CO2 The constraints on the greenhouse gas emission pathways of achieving 450 ppmv CO2e render most of the AB1 scenarios impossible to achieve. Having established which scenarios are at least quantitatively possible and subtracting the respective non-CO2 greenhouse gas and deforestation emissions, the energy and process emissions associated with each of the scenarios that remain feasible (Figure 2.4) can be derived.

Figure 2.4 illustrates that complete decarbonization of the energy and process system is necessary by between 2027 and 2063, if the total greenhouse gas emissions are to remain within the IPCC's 450 ppmv CO2e budgets. Moreover, in combination with Table 2.5, it is evident that the only meaningful opportunity for stabilizing at 450 ppmv CO2e occurs if the highest of the IPCC's cumulative emissions range is used and if emissions peak by 2015.

AB2: emissions from 2020 (peak year) to 2100

The AB2 scenario set complements the AB1 scenario set by exploring the implications for CO2e budgets of three post-peak annual emission reduction rates (3, 5 and 7 per cent). Only one peaking year is considered within this analysis with

Table 2.5 Twenty-year sustained post-peak per cent reductions in energy and process CO2 emissions (from 5 years following the peak year). (X denotes a scenario rejected on the basis of being quantitatively impossible, with prolonged per cent annual reduction rates greater than 15% or scenarios where full decarbonization is necessary within 20 years.)

Deforestation DL Deforestation DH

Table 2.5 Twenty-year sustained post-peak per cent reductions in energy and process CO2 emissions (from 5 years following the peak year). (X denotes a scenario rejected on the basis of being quantitatively impossible, with prolonged per cent annual reduction rates greater than 15% or scenarios where full decarbonization is necessary within 20 years.)

Deforestation DL Deforestation DH

Peak date

Low

Medium

High

Low

Medium

High

2015

X

X

-6%

X

X

-8%

2020

X

X

X

X

X

X

2025

X

X

X

X

X

X

Figure 2.4 Energy and process CO2 emissions derived by subtracting the non-CO2 emissions and deforestation emissions from the total greenhouse gas emissions over the period of 2000-2100, for the AB1 scenarios.

2020 chosen as arguably the most 'realistic' of the three dates in terms of both the 'practicality' of being achieved and of the respective scope for remaining within 'reasonable' bounds of CO2e concentrations. Table 2.6 summarizes the data underpinning Figure 2.5.

Table 2.6 Summary of the core components of the AB2 scenarios

Characteristic 2020-2100

Deforestation" DH and DL

Non-CO2 greenhouse gasesa mid-action

Approximate peaking value (GtCO2e) 60

Post-2020 CO2e reductions (%) 3, 5 and 7

2100 residual emissions (GtCO2e) 7.5

Deforestation and non-CO2 greenhouse gas scenarios as in Tables 2.1 and 2.2

7% reduction DL 7% reduction DH 5% reduction DL 5% reduction DH 3% reduction DL 3% reduction DH

2020

2040

2060

2080

2000

2020

2040

2060

2080

2100

Year

Figure 2.5 Greenhouse gas emission scenarios peaking in 2020, with sustained percentage emission reductions of 3, 5 and 7 per cent. The 3 and 5 per cent DH scenarios are so similar to the 3 and 5 per cent DL that they are hidden behind those profiles.

The pathways within Figure 2.5 equate to a range in cumulative CO2e emissions for 2000-2100 of 2.4 TtCO2e, 2.6 TtCO2e and 3 TtCO2e for 7, 5 and 3 per cent reductions, respectively. According to the cumulative emissions data contained within the Stern Review (Stern 2006: figure 8.1, p. 222), the first two values approximate to a CO2e concentration of approximately 550 ppmv with the latter being closer to 650 ppmv.

Table 2.7 Post-peak (2020) per cent reduction in energy and process CO2 emissions

Annual reduction Deforestation DL (%) Deforestation DH (%)

Energy and process CO2 3 6 9 4 7 12

7% reduction DL 7% reduction DH 5% reduction DL 5% reduction DH 3% reduction DL 3% reduction DH

7% reduction DL 7% reduction DH 5% reduction DL 5% reduction DH 3% reduction DL 3% reduction DH

2000

2020

2040

2060

2080

Year

Figure 2.6 CO2 emissions derived by removing the non-CO2 greenhouse gas emissions and deforestation emissions from the total greenhouse gas emissions over the period of 2000-2100 for the AB2 scenarios.

2000

2020

2040

2060

2080

2100

Year

Figure 2.6 CO2 emissions derived by removing the non-CO2 greenhouse gas emissions and deforestation emissions from the total greenhouse gas emissions over the period of 2000-2100 for the AB2 scenarios.

AB2: implications for energy and process CO2 Having developed the total CO2e pathways for AB2, and given the deforestation and non-CO2 greenhouse gas emission scenarios outlined earlier, the associated energy and process CO2 scenarios can be derived (Figure 2.6). Table 2.7 indicates typical post-peak annual reduction rates in energy and process CO2 emissions for the families of 3, 5 and 7 per cent CO2e scenarios.

According to these results, the 3, 5 and 7 per cent CO2e annual reduction rates comprising the AB2 scenarios correspond with energy and process decarbonization rates of 3-4, 6-7 and 9-12 per cent, respectively. While the latter two ranges correlate broadly with stabilization at 550 ppmv CO2e, the former, although arguably offering less unacceptable rates of reduction, correlates with stabilization nearer 650 ppmv CO2e.

Was this article helpful?

0 0

Post a comment