2.4.6.8 Mitigation/stabilisation scenarios

Mitigation scenarios (also known as climate intervention or climate policy scenarios) are defined in the TAR (Morita et al., 2001), as scenarios that “(1) include explicit policies and/or measures, the primary goal of which is to reduce GHG emissions (e.g., carbon taxes) and/or (2) mention no climate policies and/or measures, but assume temporal changes in GHG emission sources or drivers required to achieve particular climate targets (e.g., GHG emission levels, GHG concentration levels, radiative forcing levels, temperature increase or sea level rise limits).” Stabilisation scenarios are an important subset of inverse mitigation scenarios, describing futures in which emissions reductions are undertaken so that GHG concentrations, radiative forcing, or global average temperature change do not exceed a prescribed limit.

Although a wide variety of mitigation scenarios have been developed, most focus on economic and technological aspects of emissions reductions (see Morita et al., 2001; van Vuuren et al., 2006; Nakićenović et al., 2007). The lack of detailed climate change projections derived from mitigation scenarios has hindered impact assessment. Simple climate models have been used to explore the implications for global mean temperature (see Box 2.8 and Nakićenović et al., 2007), but few AOGCM runs have been undertaken (see Meehl et al., 2007, for recent examples), with few direct applications in regional impact assessments (e.g., Parry et al., 2001). An alternative approach uses simple climate model projections of global warming under stabilisation to scale AOGCM patterns of climate change assuming unmitigated emissions, and then uses the resulting scenarios to assess regional impacts (e.g., Bakkenes et al., 2006).

Box 2.8. CO₂ stabilisation and global mean temperature response

Global mean annual temperature (GMAT) is the metric most commonly employed by the IPCC and adopted in the international policy arena to summarise future changes in global climate and their likely impacts (see Chapter 19, Box 19.2). Projections of global mean warming during the 21st century for the six SRES illustrative scenarios are presented by WG I (Meehl et al., 2007) and summarised in Figure 2.8. These are baseline scenarios assuming no explicit climate policy (see Box 2.2). A large number of impact studies reported by WG II have been conducted for projection periods centred on the 2020s, 2050s and 2080s¹⁰, but only best estimates of GMAT change for these periods were available for three SRES scenarios based on AOGCMs (coloured dots in the middle panel of Figure 2.8). Best estimates (red dots) and likely ranges (red bars) for all six SRES scenarios are reported only for the period 2090-2099. Ranges are based on a hierarchy of models, observational constraints and expert judgement (Meehl et al., 2007).

A more comprehensive set of projections for these earlier time periods as well as the 2090s is presented in the lower panel of Figure 2.8. These are based on a simple climate model (SCM) and are also reported in WG I (Meehl et al., 2007, Figure 10.26). Although SCM projections for 2090-2099 contributed to the composite information used to construct the likely ranges shown in the middle panel, the projections shown in the middle and lower panels should not be compared directly as they were constructed using different approaches. The SCM projections are included to assist the reader in interpreting how the timing and range of uncertainty in projections of warming can vary according to emissions scenario. They indicate that the rate of warming in the early 21st century is affected little by different emissions scenarios (brown bars in Figure 2.8), but by mid-century the choice of emissions scenario becomes more important for the magnitude of warming (blue bars). By late century, differences between scenarios are large (e.g. red bars in middle panel; orange and red bars in lower panel), and multi-model mean warming for the lowest emissions scenario (B1) is more than 2°C lower than for the highest (A1FI).

GHG mitigation is expected to reduce GMAT change relative to baseline emissions, which in turn could avoid some adverse impacts of climate change. To indicate the projected effect of mitigation on temperature during the 21st century, and in the absence of more recent, comparable estimates in the WG I report, results from the Third Assessment Report based on an earlier version of the SCM are reproduced in the upper panel of Figure 2.8 from the Third Assessment Report. These portray the GMAT response for four CO₂-stabilisation scenarios by three dates in the early (2025), mid (2055), and late (2085) 21st century. WG I does report estimates of equilibrium warming for CO₂-equivalent stabilisation (Meehl et al., 2007)¹¹. Note that equilibrium temperatures would not be reached until decades or centuries after greenhouse gas stabilisation.

Figure 2.8. Projected ranges of global mean annual temperature change during the 21st century for CO₂-stabilisation scenarios (upper panel, based on the TAR) and for the six illustrative SRES scenarios (middle and lower panels, based on the WG I Fourth Assessment). Different approaches have been used to obtain the estimates shown in the three panels, which are not therefore directly comparable. Upper panel. Projections for four CO₂-stabilisation profiles using a simple climate model (SCM) tuned to seven AOGCMs (IPCC, 2001c, Figure SPM-6; IPCC, 2001a, Figure 9.17). Broken bars indicate the projected mean (tick mark) and range of warming across the AOGCM tunings by the 2020s (brown), 2050s (blue) and 2080s (orange) relative to 1990. Time periods are based on calculations for 2025, 2055 and 2085. Approximate CO₂-equivalent values – including non-CO₂ greenhouse gases – at the time of CO₂-stabilisation (ppm) are also shown. Middle panel. Best estimates (red dots) and likely range (red bars) of warming by 2090-2099 relative to 1980-1999 for all six illustrative SRES scenarios and best estimates (coloured dots) for SRES B1, A1B and A2 by 2020-2029, 2050-2059 and 2080-2089 (IPCC, 2007, Figure SPM.5). Lower panel. Estimates based on an SCM tuned to 19 AOGCMs for 2025 (representing the 2020s), 2055 (2050s) and 2085 (2080s). Coloured dots represent the mean for the 19 model tunings and medium carbon cycle feedback settings. Coloured bars depict the range between estimates calculated assuming low carbon cycle feedbacks (mean - 1 SD) and those assuming high carbon cycle feedbacks (mean + 1 SD), approximating the range reported by Friedlingstein et al., 2006. Note that the ensemble average of the tuned versions of the SCM gives about 10% greater warming over the 21st century than the mean of the corresponding AOGCMs. (Meehl et al., 2007, Figure 10.26 and Appendix 10.A.1). To express temperature changes relative to 1850-1899, add 0.5°C.

The scarcity of regional socio-economic, land-use and other detail commensurate with a mitigated future has also hindered impact assessment (see discussion in Arnell et al., 2002). Alternative approaches include using SRES scenarios as surrogates for some stabilisation scenarios (Swart et al., 2002; see Table 2.4), for example to assess impacts on ecosystems (Leemans and Eickhout, 2004) and coastal regions (Nicholls and Lowe, 2004), demonstrating that socio-economic assumptions are a key determinant of vulnerability. Note that WG I reports AOGCM experiments forced by the SRES A1B and B1 emissions pathways up to 2100 followed by stabilisation of concentrations at roughly 715 and 550 ppm CO₂ (equated to 835 and 590 ppm equivalent CO₂, accounting for other GHGs: see Meehl et al., 2007).

A second approach associates impacts with particular levels or rates of climate change and may also determine the emissions and concentration paths that would avoid these outcomes. Climate change and impact outcomes have been identified based on criteria for dangerous interference with the climate system (Mastrandrea and Schneider, 2004; O’Neill and Oppenheimer, 2004; Wigley, 2004; Harvey, 2007) or on meta-analysis of the literature (Hitz and Smith, 2004). A limitation of these types of analyses is that they are not based on consistent assumptions about socio-economic conditions, adaptation and sectoral interactions, and regional climate change.

A third approach constructs a single set of scenario assumptions by drawing on information from a variety of different sources. For example, one set of analyses combines climate change projections from the HadCM2 model based on the S750 and S550 CO₂-stabilisation scenarios with socio-economic information from the IS92a reference scenario in order to assess coastal flooding and loss of coastal wetlands from long-term sea level rise (Nicholls, 2004; Hall et al., 2005) and to estimate global impacts on natural vegetation, water resources, crop yield and food security, and malaria (Parry et al., 2001; Arnell et al., 2002).