IPCC Fourth Assessment Report: Climate Change 2007
Climate Change 2007: Working Group III: Mitigation of Climate Change

3.5.2 Linking emission scenarios to changes in global mean temperature, impacts and key vulnerabilities

In a risk-management framework, a first step to understanding the environmental consequences of mitigation strategies is to look at links between various stabilization levels for concentrations of greenhouse gases in the atmosphere, and the global mean temperature change relative to a particular baseline. A second step is to link levels of temperature change and key vulnerabilities. Climate models indicate significant uncertainty at both levels. Figure 3.38 shows CO2-eq concentrations that would limit warming at equilibrium below the temperatures indicated above pre-industrial levels, for ‘best estimate’ climate sensitivity, and for the likely range of climate sensitivity (see Meehl et al., 2007, Section 10.7, and Table 10.8; and the notes to Figure 3.38). It also shows the corresponding radiative forcing levels and their relationship to equilibrium temperature and CO2-eq concentrations. The table and the figure illustrate how lower temperature constraints require lower stabilization levels, and also that, if the potential for climate sensitivities is higher than the ‘best estimate’ and is taken into account, the constraint becomes more stringent. These more stringent constraints lower the risks of exceeding the threshold.

Figure 3.38 and Table 3.10 provide an overview of how emission scenarios (Section 3.3) relate to different stabilization targets and to the likelihood of staying below certain equilibrium warming levels. For example, respecting constraints of 2°C above pre-industrial levels, at equilibrium, is already outside the range of scenarios considered in this chapter, if the higher values of likely climate sensitivity are taken into account (red curve in Figure 3.38), whilst a constraint of respecting 3°C above pre-industrial levels implies the most stringent of the category I scenarios, with emissions peaking in no more than the next 10 years, again if the higher likely values of climate sensitivity are taken into account. Using the ‘best estimate’ of climate sensitivity (i.e. the estimated mode) as a guide for establishing targets, implies the need for less stringent emission constraints. This ‘best estimate’ assumption shows that the most stringent (category I) scenarios could limit global mean temperature increases to 2°C–2.4°C above pre-industrial levels, at equilibrium, requiring emissions to peak within 10 years. Similarly, limiting temperature increases to 2°C above pre-industrial levels can only be reached at the lowest end of the concentration interval found in the scenarios of category I (i.e. about 450 ppmv CO2-eq using ‘best estimate’ assumptions). By comparison, using the same ‘best estimate’ assumptions, category II scenarios could limit the increase to 2.8°C–3.2°C above pre-industrial levels at equilibrium, requiring emissions to peak within the next 25 years, whilst category IV scenarios could limit the increase to 3.2°C–4°C above pre-industrial at equilibrium requiring emissions to peak within the next 55 years. Note that Table 3.10 category IV scenarios could result in temperature increases as high as 6.1°C above pre-industrial levels, when the likely range for the value of climate sensitivity is taken into account. Hence, setting policy on the basis of a ‘best estimate’ climate sensitivity accepts a significant risk of exceeding the temperature thresholds, since the climate sensitivity could be higher than the best estimate.

Figure 3.38

Figure 3.38: Relationship between global mean equilibrium temperature change and stabilization concentration of greenhouse gases using: (i) ‘best estimate’ climate sensitivity of 3°C (black), (ii) upper boundary of likely range of climate sensitivity of 4.5°C (red), (iii) lower boundary of likely range of climate sensitivity of 2°C (blue) (see also Table 3.9).


1. IPCC (2007a) finds that the climate sensitivity is likely to be in the range 2°C –4.5°C, with a ‘best estimate’ of about 3°C, very unlikely to be less than 1.5°C and values substantially higher than 4.5°C ‘cannot be excluded’ (IPCC (2007a, SPM).

2. The simple relationship Teq = T2xCO2 x ln([CO2]/280)/ln(2) is used (see Meehl et al. (2007), Section 10.7, and Table 10.8), with upper and lower values of T2xCO2 of 2 and 4.5°C.

3. Non-linearities in the feedbacks (including e.g. ice cover and carbon cycle) may cause time dependence of the effective climate sensitivity, as well as leading to larger uncertainties for greater warming levels. This likelihood level is consistent with the IPCC Working Group I assessment of climate sensitivity, see Note 1, and drawn from additional consideration of Box 10.2, Figure 2, in IPCC (2007a).

Table 3.10: Properties of emissions pathways for alternative ranges of CO2 and CO2-eq stabilization targets. Post-TAR stabilization scenarios in the scenario database (see also Sections 3.2 and 3.3); data source: after Nakicenovic et al., 2006 and Hanaoka et al., 2006)

Class Anthropogenic addition to radiative forcing at stabilization (W/m2Multi-gas concentration level (ppmv CO2-eq) Stabilization level for CO2 only, consistent with multi-gas level (ppmv CO2Number of scenario studies Global mean temperature C increase above pre-industrial at equilibrium, using best estimate of climate sensitivityc) Likely range of global mean temperature C increase above pre-industrial at equilibriuma) Peaking year for CO2 emissionsb) Change in global emissions in 2050 (% of 2000 emissions)b) 
2.5-3.0 445-490 350-400 2.0-2.4 1.4-3.6 2000-2015 -85 to -50 
II 3.0-3.5 490-535 400-440 18 2.4-2.8 1.6-4.2 2000-2020 -60 to -30 
III 3.5-4.0 535-590 440-485 21 2.8-3.2 1.9-4.9 2010-2030 -30 to +5 
IV 4.0-5.0 590-710 485-570 118 3.2-4.0 2.2-6.1 2020-2060 +10 to +60 
5.0-6.0 710-855 570-660 4.0-4.9 2.7-7.3 2050-2080 +25 to +85 
VI 6.0-7.5 855-1130 660-790 4.9-6.1 3.2-8.5 2060-2090 +90 to +140 


a. Warming for each stabilization class is calculated based on the variation of climate sensitivity between 2ºC –4.5ºC, which corresponds to the likely range of climate sensitivity as defined by Meehl et al. (2007,Chapter 10).

b. Ranges correspond to the 70% percentile of the post-TAR scenario distribution.

c. ‘Best estimate’ refers to the most likely value of climate sensitivity, i.e. the mode (see Meehl et al. (2007,Chapter 10) and Table 3.9

Table 3.11 highlights a number of climate change impacts and key vulnerabilities organized as a function of global mean temperature rise (IPCC, 2007b, Chapter 19). The table highlights a selection of key vulnerabilities representative of categories covered in Chapter 19 (Table 19.1) in IPCC (2007b). The italic text in Table 3.11 highlights examples of avoided impacts derived from ensuring that temperatures are constrained to any particular temperature range compared to a higher one. For example, significant benefits result from constraining temperature change to not more than 1.6°C–2.6°C above pre-industrial levels. These benefits would include lowering (with different levels of confidence) the risk of: widespread deglaciation of the Greenland Ice Sheet; avoiding large-scale transformation of ecosystems and degradation of coral reefs; preventing terrestrial vegetation becoming a carbon source; constraining species extinction to between 10–40%; preserving many unique habitats (see IPCC, 2007b, Chapter 4, Table 4.1 and Figure 4.5) including much of the Arctic; reducing increases in flooding, drought, and fire; reducing water quality declines, and preventing global net declines in food production. Other benefits of this constraint, not shown in the Table 3.11, include reducing the risks of extreme weather events, and of at least partial deglaciation of the West Antarctic Ice Sheet (WAIS), see also IPCC, 2007b, Section 19.3.7. By comparison, for ‘best guess’ climate sensitivity, attaining these benefits becomes unlikely if emission reductions are postponed beyond the next 15 years to a time period between the next 15–55 years. Such postponement also results in increasing risks of a breakdown of the Meridional Overturning Circulation (IPCC, 2007b, Table 19.1).

Even for a 2.6°C –3.6°C temperature rise above pre-industrial levels there is also medium confidence in net negative impacts in many developed countries (IPCC, 2007b, Section 19.3.7). For emission-reduction scenarios resulting in likely temperature increases in excess of 3.6°C above pre-industrial levels, successively more severe impacts result. Low temperature constraints are necessary to avoid significant increases in the impacts in less developed regions of the world and in polar regions, since many market sectors in developing countries are already affected below 2.6°C above pre-industrial levels (IPCC, 2007b, Section 19.3.7), and indigenous populations in high latitude areas already face significant adverse impacts.

It is possible to use stablization metrics (i.e. global mean temperature increase, concentrations in ppmv CO2-eq or radiative forcing in W/m2) in combination with the mitigation scenarios literature to assess the cost of alternative mitigation pathways that respect a given equilibrium temperature, key vulnerability (KV) or impact threshold. Whatever the target, both early and delayed-action mitigation pathways are possible, including ‘overshoot’ pathways that temporarily exceed this level. A delayed mitigation response leads to lower discounted costs of mitigation, but accelerates the rate of change and the risk of transiently overshooting pre-determined targets (IPCC, 2007b, Section 19.4.2).

A strict comparison between mitigation scenarios and KVs is not feasible as the KVs in Table 3.11 refer to realized transient temperatures in the 21st century rather than equilibrium temperatures, but a less rigorous comparison is still useful. Avoidance of many KVs requires temperature change in 2100 to be below 2°C above 1990 levels (or 2.6°C above pre-industrial levels). Using equilibrium temperature as a guide, impacts or KV could be less than expected, for example if impacts do not occur until the 22nd century, because there is more time for adaptation. Or they might be greater than expected, as temperatures in the 21st century may transiently overshoot the equilibrium, or stocks at risk (such as human populations) might be larger. Some studies explore the link between transient and equilibrium temperature change for alternative emission pathways (O’Neill and Oppenheimer, 2004; Schneider and Mastrandrea, 2005; Meinshausen, 2006).

It is transient climate change, rather than equilibrium change, that will drive impacts. More research is required to address the question of emission pathways and transient climate changes and their links to impacts.[25] In the meantime, equilibrium temperature change may be interpreted as a gross indicator of change, and given the caveats above, as a rough guide for policymakers’ consideration of KV and mitigation options to avoid KV.

GMT range relative to 1990 (pre-industrial) Geophysical systems Example: Greenland ice sheeta (IPCC, 2007b: 6.3;; IPCC, 2007a: 4.7.4;;; Global biological systems Example: terrestrial ecosystemsb (IPCC, 2007b: 4.4.11; 1.3.4; 1.3.5) Global social systems Example: waterc (IPCC, 2007b: 3 ES; 3.4.3; 13.4.3) Global social systems Example: food supplyc (IPCC, 2007b: 5.6.1; 5.6.4) Regional systems Example: Polar Regionsd (IPCC, 2007b: 15.4.1; 15.4.2; 15.4.6; 15.4.7) Extreme events Example: fire riske (IPCC, 2007a: 7.3; IPCC, 2007b; 1.3.6) 
>4 (>4-6) Near-total deglaciation**  Large-scale transformation of ecosystems and ecosystem services** At least 35% of species committed to extinction (3°C)** Severity of floods, droughts, erosion, water quality deterioration will increase with increasing climate change*** Further declines in global food production o/* Continued warming likely to lead to further loss of ice cover and permafrost**. Arctic ecosystems further threatened**, although net ecosystem productivity estimated to increase (o) Frequency and intensity likely to be greater, especially in boreal forests and dry peat lands after melting of permafrost** 
3-4 (3.6-4.6) Commitment to widespread** to near-total deglaciation* 2-7 m sea level rise over centuries to millennia Global vegetation becomes net source of C above 2-3ºC */** Sea level rise will extend areas of salinization of ground water, decreasing freshwater availability in coastal areas***   While some economic opportunities will open up (e.g. shipping), traditional ways of life will be disrupted**   
2-3 (2.6-3.6) Lowers risk of near-total deglaciation Widespread disturbance, sensitive to rate of climate change and land use*** 20 to 50% species committed to extinction* Avoids widespread disturbance to ecosystems and their services***, and constrains species losses Hundreds of millions people would face reduced water supplies (**Global food production peaks and begins to decrease o/* (1-3ºC) Lowers risk of further declines in global food production associated with higher temperatures*     
1-2 (1.6-2.6) Localized deglaciation (already observed due to local warming), extent would increase with temperature**  10-40% of species committed to extinction* Reduces extinctions to below 20-50%*, prevents vegetation becoming carbon source*/** Many ecosystems already affected*** Increased flooding and drought severity** Lowers risk of floods, droughts, deteriorating water quality*** and reduced water supplied for hundreds of millions of people** Reduced low latitude production*. Increased high latitude production* (1-3ºC) Climate change is already having substantial impacts on societal and ecological systems*** Increased fire frequency and intensity in many areas, particularly where drought increases** 
0-1 (0.6-1.6) Lowers risk of widespread** to near-total deglaciation* Reduces extinctions to below 10-30%*; reduces disturbance levels***   Increased global production o/* Lowers risk of decrease in global food production and reduces regional losses (or gains) o/* Reduced loss of ice cover and permafrost; limits risk to Arctic ecosystems and limits disruption of traditional ways of life** Lowers risk of more frequent and more intense fires in many areas** 


Plain text shows predicted vulnerabilities in various temperature ranges for global annual mean temperature rise relative to 1990. Italic text shows benefits (or damage avoided) upon constraining temperature increase to lower compared to higher temperature ranges.

Confidence symbol legend: o low confidence; * medium confidence; ** high confidence; *** very high confidence

Excerpts from IPCC, 2007b, Table 19.1;

a. Refer to IPCC (2007b, Table 19.1) for further information, also concerning bio-geochemical cycles, West Antarctic Ice Sheet and Meridional Overturning Circulation

b. Refer to IPCC (2007b, Table 19.1) for further information, also concerning marine and freshwater ecosystems

c. Refer to IPCC (2007b, Table 19.1) for further information, also concerning infrastructure, health, migration and conflict, and aggregate market impacts

d. Refer to IPCC (2007b, Table 19.1) for further information and also for other regions

e. Refer to IPCC (2007b, Table 19.1) for further information, also concerning tropical cyclones, flooding, extreme heat, and drought


  1. ^  See IPCC (2007b, Section 19.4, Figure 19.2) and Meehl et al. (2007, Section 10.7) for further discussion of equilibrium and transient temperature increases in relation to stabilization pathways