The uncertainty of climate sensitivity and its implication for the Paris negotiation
- First Online:
- Cite this article as:
- Kaya, Y., Yamaguchi, M. & Akimoto, K. Sustain Sci (2016) 11: 515. doi:10.1007/s11625-015-0339-z
Uncertainty of climate sensitivity is one of the critical issues that may affect climate response strategies. Whereas the equilibrium climate sensitivity (ECS) was specified as 2–4.5 °C with the best estimate of 3 °C in the 4th Assessment Report of IPCC, it was revised to 1.5–4.5 °C in the 5th Assessment Report. The authors examined the impact of a difference in ECS assuming a best estimate of 2.5 °C, instead of 3 °C. The current pledges of several countries including the U.S., EU and China on emission reductions beyond 2020 are not on track for the 2 °C target with an ECS of 3 °C but are compatible with the target with an ECS of 2.5 °C. It is critically important for policymakers in Paris to know that they are in a position to make decisions under large uncertainty of ECS.
KeywordsUncertainty Equilibrium climate sensitivity Paris climate conference Intended nationally determined contributions 2 °C target
By the end of June, 2015, the United States, the European Union, China and several other countries submitted their intended nationally determined contributions (INDCs) to the UNFCCC secretariat. This is a good start toward the coming Paris climate conference (COP 21). However, according to our estimate based on our global energy systems model DNE21+ (Akimoto 2008) and a simple climate change model MAGICC (Meinshausen et al. 2011), these pledges are nowhere near sufficient to limit the temperature increase to less than 2 °C since pre-industrialization if we apply 3 °C as the best estimate1 of the equilibrium climate sensitivity (ECS).
ECS is defined as an increase in global mean surface temperature caused by a doubling of the atmospheric CO2 concentration. The uncertainty of climate sensitivity poses one of the greatest challenges in planning strategies on how and to what extent we should cope with risks of climate change.
Throughout IPCC’s 1st to 3rd Assessment Report, the likely range of ECS was estimated as 1.5–4.5 °C with its best estimate at 2.5 °C. The 4th Assessment Report (AR4) specified a likely range (greater than 66 % probability) of ECS as 2–4.5 °C with its “most likely value” or “best estimate” of 3 °C, but the 5th Assessment Report (AR5) lowered the figure to 1.5–4.5 °C. In addition, no best estimate was given by AR5 because of the difference of methodologies of estimating ECS as explained below (IPCC 2013).
According to the estimates of atmosphere–ocean general circulation models (AOGCMs) in the AR5, the mean value of ECS is 3.2 °C and the ranges is 2.0–4.5 °C, close to that in the AR4. Climate sensitivity can also be estimated from observations of surface temperature and climate forcing data. The values of ECS thus estimated are rather lower as pointed out by Rogelj et al. and others (IPCC 2013, Lewis and Curry, 2014, Otto et al. 2013) and Lewis and Curry (2014) estimated the likely range of climate sensitivity as 1.25–2.45 °C with its median estimate at 1.64 °C. There are several criticisms of the observation-based methods, though, including one arguing that the observed warming is likely biased low (Durack et al. 2014).
We would like to examine the effect of uncertainty of ECS on emissions targets. Rogelj et al. (2014) admitting recent estimates based on the observed warming trends tend to show lower values of climate sensitivity, argues that “[T]here are several climate policy implications that can be drawn from recent ECS estimates. The most important, however, is that they do not change the big picture if all available evidence is taken into account. (…) Even the lowest ECS estimate assumed in this study only results in a delay of less than a decade in the timing of when the 2 °C threshold would be crossed when emission trends from the past 10 years are continued.”
This conclusion, what we found, comes from the Supplementary Material of the paper. There, two ECS distributions, among others, were used for comparison; one was named as IPCC AR4 consistent and the other as IPCC AR5 consistent. By comparing those two, Rogelj et al. concluded that the differences of climate sensitivity ‘do not change the picture’. What matters is that the median values of those two were almost the same, i.e., 3 °C for the former and 3.1 °C for the latter, though distribution itself is a little bit flat in AR5 consistent.
As pointed out previously, the likely range of ECS was lowered to 1.5–4.5 °C (in AR5) from 2 to 4.5 °C (in AR4), and experts were unable to agree on the value of the best estimate in AR5 though it was agreed as 3 °C in AR4. In addition, the value of 2.5 °C had been used as best estimate (most likely value) throughout IPCC’s 1st to 3rd assessment reports where the likely range of climate sensitivity had been 1.5–4.5 °C. Under the above situation, it is only natural to assume the best estimate (median) for AR5 will be lower than 3 °C. Therefore, we chose the best estimate value of 2.5 °C for the purpose of comparison to explore the impact of difference in ECS on climate negotiations. The point at issue here is whether INDCs submitted by major countries are consistent with the 2 °C target under different climate sensitivities. Note that it is not the authors’ intention to argue 2.5 °C is the correct value.
Relationship between CO2 eq. concentrations, emissions reductions and temperature changes
CO2 eq Concentrations in 2100 [ppm CO2 eq] category label (concentration range)
Cumulative CO2 emissions [GtCO2]
Change in CO2 eq emissions compared to 2010
Temperature change (relative to 1850–1900)
2100 Temperature change [°C]
Likelihood of staying below temperature level over the twenty first century
−72 to −41
More unlikely than likely
No overshoot of 530 ppm CO2 eq
−57 to −42
More likely than not
Overshoot of 530 ppm CO2 eq
−55 to −25
About as likely as not
No overshoot of 580 ppm CO2 eq
−47 to −19
More unlikely than likely
Overshoot of 580 ppm CO2 eq
−16 to +7
The outcome of our model shows global total emissions under major countries’ INDCs (red line) in 2030 will not be on track to attain the 2 °C target if climate sensitivity is 3 °C (orange line). On the other hand, the red line emissions are in line with the green line that is consistent with the 2 degree target if climate sensitivity is 2.5 °C, and if we allow a temporal overshoot of 580 ppmCO2-eq. This implies, with ECS equal to 2.5 °C, that the 2 °C target is still within reach.
It is clear from the above explanations that the impact of a mere 0.5 °C difference in climate sensitivity is of critical significance for policy objectives, which is especially significant given the large uncertainties over climate sensitivity.
It is scientific community’s vital role to narrow the uncertainty range of ECS. At the same time it is critically important for policymakers in Paris to know that they are in a position to make decisions under large uncertainty of ECS.
The authors acknowledge Dr. Judith A. Curry, Georgia Institute of Technology, for her valuable comments.