Increased carbon budgets in IPCC SR1.5: Bittersweet news (part-I)

(Note: To improve its readability this post has been split in two parts)

The new IIPC SR1.5 (‘Global warming of 1.5°C’, released in October 2018) reports an important  increase in the available carbon budgets (CB) to comply with different degrees of global warming with regard to those reported in the fifth IPCC assessment report (AR5), which was the main reference for CBs since its publication in 2014.

The first impression from this IPCC-backed increased CBs is a sweet one, with certain relief, because it seems to provide a buffer and additional room to materialize a transition with chances to stabilize climate change with global warming below 1.5°C, and therefore to limit important climate change impacts.

However, there is a bitter aftertaste from these IPCC-backed increased CBs, which rises as soon as one delves into the details of its origin and the likely implications of its dissemination into our socio-economic and institutional framework. The two main points on this regard are:

• The new IPCC SR1.5 CBs do not represent an update of the former IPCC AR5 CBs, in the sense that they are not based on improvements over the modeling approach used to obtain the AR5 CBs. In this sense the AR5 CBs keep on being conceptually valid, and the new SR1.5 simply presents CB results from a different methodological approach. In fact, the SR1.5 CBs use the same slope of the dependency between global warming and cumulative emissions (transient climate response to cumulative emissions: TCRE) as evaluated in AR5. The difference from SR1.5 and AR5 CBs basically stems from displacing the curve that relates global warming and cumulative emissions as a response of a discrepancy between measured and calculated historic annual emissions, which is more significant around 1940. The SR1.5 gives priority to measurements over modeled results for annual CO₂ emissions, but since the measured CO₂ emissions have high uncertainty, and particularly around the 1940 period, one can’t really place much confidence on the updated CB values, especially given their high discrepancy with the AR5 CBs and the fact that they move in the direction of reducing our safety buffer as a society.

• Indeed, in spite of the appearance of providing a buffer (higher CBs should provide more chances to comply with climate boundary conditions), the updated IPCC SR1.5 CBs reduce our safety buffer for a successful transition towards sustainability. This is because our socio-economic and institutional current setups, frameworks and dynamics, have been and are still geared towards the silly goal of maximizing the use of fossil fuels (FF) consistent with a specified climate boundary. This approach is deeply imbedded into policy making driven by FF interests, but also into most of the scientific approaches used to evaluate the transition potential: Indeed, most of the Integrated Assessment Models (IAMs) used to draw the IPCC conclusions implement a shortsighted ‘economic’ optimization algorithm that basically provides the answer to the question of 'how slowly we can transition to comply with the specified climate goal' (i.e., how can we maximize the FF emissions within this climate boundary), let alone the fact that they keep on relying on equilibrium models that do not approach the reality of how the economy works, and on a deeply underdeveloped concept of the ‘economy’ itself, that considers it as a self-contained system that does not acknowledge its embedded nature into society and the Planet. This reality becomes still more pathetic when we hear the reaction from the big oil corporations to the calls for urgent climate action from the IPCC SR1.5 (ExxonMobil CEO Depressed AfterRealizing Earth Could End Before They Finish Extracting All The Oil ). Under this fully alienated context, it really does not sound like a good idea to provide updated IPCC-backed CBs that in spite of being full of uncertainties basically convey the main message that we have room to burn more carbon than the one we thought before while still remaining within appropriate boundary conditions, since from now onward all policies and modeling efforts will just adjust their goal towards the increased carbon allowance, instead of focusing on what really matters, which is transitioning as fast as possible.

Below we will develop further some of these issues, but let’s upfront state the main insights from this analysis:

• Do not take the higher IPCC SR1.5 CBs as a license to still emit more carbon to the atmosphere. The old IPCC AR5 carbon budgets are still as valid as they were before the IPCC SR1.5 release. Moreover, the huge uncertainties embedded into the CBs could very well mean that in fact we have already exhausted them.

• The fundamental transition goal right now should migrate from the alignment with specific global warming goals towards unlocking all the potential and capacity to transition as fast as possible. We already know that the later we manage to transition the highest will be the costs for the society and the Planet. And we also know that due to the lack of effective transition action up till now, even if we deploy a fast transition we will have to manage and try to adapt to strong climate impacts: even at 1.5°C global warming the impacts will be strong, they will basically double at 2°C global warming (IPCC SR1.5C), and they are already important at the present 1°C global warming). Therefore, in any case we will need to adapt to the impacts of climate change, and for this adaption capability and associated resilience to be unlocked we need to address the very structural aspects of the transition as soon and fast as possible. We should pay attention to avoid 'green washing' transition strategies, which ultimately will produce additional delays, and focus without any further delay on the structural aspects of a fair and just transition, as well as avoiding those transition strategies that under the flag of one specific global warming goal rely on false solutions that ultimately will further weaken the social and planetary capability to manage and navigate the climate (and potentially other global) impacts that are already in the pipeline. This somehow simplifies the subject and decision-making process: Just run as fast as possible along a pathway that addresses structural issues.

The CB concept it is an important one for climate change discussions, analysis and policies, and therefore an effort should be made to reduce noise around it. Indeed, the power of the CB concept is that it provides a very direct, clear and effective means to communicate and calibrate any transition policy or initiative. This clarity and communicative capability is of paramount importance to articulate an effective, engaging and participative transition towards sustainability. And among the enormous amount of often blurry seas of climate change information, pledges and policies, CBs stand as one of the very few elements promoting and supporting this clarity and communicative capacity.

The accounting concept of the CB is conceptually very close to citizens and politicians: The carbon budget available to maintain global warming below one specific target (1.5°C, 2°C,…) is the amount of CO₂ that still can be emitted at any point at time. Of course, as times goes by and we keep on emitting, the CB reduces, just like the monetary or resources budget decreases as we keep on spending.

But first of all: Let’s have a look at how big the current CBs are, according to both AR5 and SR1.5 IPCC reports.

The IPCC AR5 provided CBs for 2011, and the IPCC SR1.5 provides CBs for 2018. To compare both budgets we need to discount historic CO₂ emissions until a common year. Since we are already on the point of turning into 2019 (and still without having done anything effective to tackle climate change…), here we will present the 2019 carbon budgets: The amount of CO₂ that we can still emit as of 1/1/2019 to comply with different global warming goals.

Figure-1 presents a direct comparison of the 2019 carbon budgets for 2°C@66% (this means the budget for global warming of 2°C with a likelihood of 66%), 1.5°C@50% and 1.5°C@66% for the IPCC AR5 and the IPCC SR1.5. The SR1.5 CBs presented in this figure are the main values provided in the SR1.5 report. The figure also presents the associated remaining years before budget exhaustion and the year of budget exhaustion (assuming annual emissions equal to those in 2017).

First think to clarify is that the quantification for the likelihood of staying under the specified global warming has different interpretations in AR5 and SR1.5. In AR5 the likelihood is evaluated through the dispersion of the results obtained from different climate models. In SR1.5 the likelihood is associated to the uncertainty in the slope of the dependence between global warming and cumulative emissions (transient climate response to cumulative emissions: TCRE) as obtained by the AR5 model runs.

The second issue that is worth pointing regarding the likelihood, are the meager values that we are considering: 66% or 50% likelihood of staying within the specified global warming if the emissions do not exceed the associated carbon budget. Hence, for 1.5°C@50% we have a 50% chance that even if emissions are within the associated carbon budget we still exceed the 1.5°C global warming…

The third thing that needs to be commented regarding the likelihood is the fundamental uncertainty embodied into its very value. Indeed, these likelihoods are not real likelihoods: they are estimates that come from the dispersion of results from climate models. But there is an implicit bias in this way of estimating likelihoods, because all the climate models used for evaluating global warming in AR5 basically have the same fundamental limitations in their modeling approach, which stem from the current understanding and modeling capability of the climate and Earth systems. Therefore, any uncertainty associated to real aspects not captured by these climate models (like several climate feedbacks) is simply not captured into this likelihood estimate. Because of the nature of the current ‘unknowns’ from the response of the climate system we could therefore expect that the real likelihood of limiting global warming to a specified value if the corresponding CB is not exceeded is lower (and perhaps significantly lower) to the one indicated. Hence, for instance, for the 1.5°C@50% carbon budget we could expect the real likelihood of staying within a 1.5°C global warming if this carbon budget is not exceeded to be lower than 50%, which really does not leave us in a very good position EVEN in the VERY unlikely event that we would comply with the 1.5°C@50% carbon budget.

Coming back to Figure-1, we can appreciate the very significant increase of CBs reported by IPCC SR1.5 with regard to those from IPCC AR5: The SR1.5 2°C@66% CB is 90% higher than the AR5 value, the 1.5°C@50% CB is 227% higher in the SR1.5, and the 1.5°C@66% CB is 623% higher in the SR1.5.

Figure-1 also translates these CBs in terms of the remaining years before exhaustion and year of exhaustion if future annual emissions would equal to those from 2017. As we can see, in any case the current situation is VERY tight, with only 2 years left (2020) for exhausting the 1.5°C@66% carbon budget from AR5 and 13 years (2031) according to the carbon budget from SR1.5.

Figure-1: 2019 carbon budgets for 2°C@66%, 1.5°C@50% and 1.5°C@66% as per the IPCC AR5 and IPCC SR1.5. The remaining years before budget exhaustion and the year of budget exhaustion, assuming future annual emissions equal to those in 2017, are also presented.

But the Figure-1 carbon budgets from both AR5 and SR1.5 do not take into account several climate feedbacks with the potential to significantly accelerate climate change. These are those complex non-linear physical processes that still escape to the modeling capability of the current climate models. However, our current physical understanding of the climate system provides evidence that even with 1.5°C or 2°C global warming several tipping points unlocking some of these climate feedbacks could be surpassed. Therefore, we should better count on the effect of climate feedbacks when evaluating our available carbon budget. The IPCC SR1.5 provides a very high-level quantification of the effect of climate feedbacks (citing two of them: CO₂ released by permafrost thawing or methane released by wetlands), without much backing and giving the impression of being a rather incomplete estimate. The IPCC SR1.5 quantifies the impact of climate feedbacks into a 100 GtCO₂ reduction of the carbon budget up to 2100 (acknowledging that this figure should be higher if we consider impacts beyond 2100). This figure certainly seems far too low, and even one of the studies cited in the IPCC SR1.5 quantifies almost twice this figure associated to one single climate feedback (methane emissions from permafrost). Hence, we should consider the estimate of the impact from climate feedbacks on the CBs provided by the IPCC SR1.5 as extremely optimistic and with a very high uncertainty. Still, if we include this estimate into the IPCC SR1.5 presented carbon budgets, we get the results from Figure-2.

Figure-2: 2019 carbon budgets for 2°C@66%, 1.5°C@50% and 1.5°C@66% as per the IPCC AR5 and IPCC SR1.5. For the SR1.5 CBs the estimate of climate feedbacks provided in the IPCC SR1.5 has been included. The remaining years before budget exhaustion and the year of budget exhaustion, assuming future annual emissions equal to those in 2017, are also presented.

Uncertainty is a big issue here, mostly when we pretend to take carbon budgets (or equivalent climate model-predicted global warming) as goals to guide ‘economic’ optimization routines that are meant to provide us transition pathways consistent with the climate goals. The IPCC SR1.5 is really not very conclusive regarding the uncertainty associated to the provided CBs. The only think that says is a very vague statement that the uncertainty is expected to be 50%, which therefore we have to interpret as a highly uncertain figure itself. However, if we take this uncertainty estimate and look for the lower potential value (the one we should aim at for not missing the goal) of the SR1.5 CBs, we get the results presented in Figure-3. As we can see, now both the AR5 and SR1.5 become very close to each other. For the 2°C@66% CB the SR1.5 value becomes lower than the AR5 value. For 1.5°C@50% and 1.5°C@66% CBs the SR1.5 value keeps on being higher than the AR5 value but now both are rather aligned.

Figure-3: 2019 carbon budgets for 2°C@66%, 1.5°C@50% and 1.5°C@66% as per the IPCC AR5 and IPCC SR1.5. For the SR1.5 CBs the estimate of climate feedbacks provided in the IPCC SR1.5 has been included. The SR1.5 CBs herewith presented are the lower end of the 50% uncertainty range indicated in the IPCC SR1.5. The remaining years before budget exhaustion and the year of budget exhaustion, assuming future annual emissions equal to those in 2017, are also presented.

However, the IPCC SR1.5 provides more information about the potential uncertainty of the presented CBs. Specifically, it reports the expected impact of the following uncertainty sources:

•  the uncertainty associated to variations of non-CO₂ GHG emissions evolution pathways. This is what I denominated Carbon Budget Wedges (CBW) in the 1.5°C Climate Vision analysis that I developed in late 2016 for Greenpeace International (see here, here and here)
• the uncertainty associated to non-CO₂ forcing and response
• the historical temperature uncertainty
• the uncertainty associated to recent (since 2011) emissions uncertainty

Clearly, one could think of further sources of uncertainty, like those associated to historic emissions uncertainties, among which are the emissions around 1940 that underpin the fundamental difference between the reported AR5 and SR1.5 CBs, or the huge uncertainties associated to climate feedbacks and the modeling limitations of current climate models. But let’s just concentrate on the uncertainty sources quantified in the SR1.5 report.

The IPCC SR1.5 report is quick in providing a disclaimer that the different reported uncertainties cannot be added up. However, the SR1.5 report does not elaborate on how they should be combined nor substantiate the statement that they should not be added up, to which we should add the uncertainty that seems to be attached to these uncertainty estimates themselves.

But with all these caveats, we still think that adding these provided uncertainty estimates may provide a powerful qualitative insight that should not be overlooked in the discussion about the CBs. Hence, in Figure-4 we compare again the AR5 and SR1.5 CBs, but in this case, for the SR1.5 we use the lower value of the CBs associated to the different quantified uncertainties.

The picture provided by Figure-4 changes completely. Now all the SR1.5 CBs are lower than the AR5 estimates, with the CBs for the 1.5°C@50% and 1.5°C@66% already having been exhausted (in 2011 and 2006 respectively), and the 2°C@66% being exhausted within just 6 years (in 2024).

Well, this changes everything (as Naomi Klein would probably say), because the SR1.5 report, far from pretending to communicate the ‘convenient’ message that we still have room to emit a bit more of carbon than the one we thought, should completely refocus the strategy towards deploying a sustainable, just and fair transition as fast as possible, with the priority on addressing the structural changes that can provide resilience and adaption capability to navigate the future. We could already have gone beyond the 1.5°C@50% CB, and we know by sure that the higher the final global warming the worst. But we also need to be very aware that not any transition will do the job, and that there are pathways that can significantly debilitate our capability as a society to navigate the future, aggravating those structural aspects that underpin the very climate and social crisis we are currently facing.

Figure-4: 2019 carbon budgets for 2°C@66%, 1.5°C@50% and 1.5°C@66% as per the IPCC AR5 and IPCC SR1.5. For the SR1.5 CBs the estimate of climate feedbacks provided in the IPCC SR1.5 has been included. The SR1.5 CBs herewith presented are the lower end of the uncertainty range indicated in the IPCC SR1.5 through the different quantified uncertainty sources. The remaining years before budget exhaustion and the year of budget exhaustion, assuming future annual emissions equal to those in 2017, are also presented.

I will finish with one personal note associated to CB's uncertainties: 

End 2016 I was developing the 1.5°C Climate Vision analysis (CV1.5) for Greenpeace International (see here, here  and here). The CV1.5 is a transition analysis addressing the possibilities, requirements and implications of a transition aligned with the Paris Agreement 1.5°C goal. 

The IPCC AR5 CBs were used as reference for the CV1.5 analysis. 

By this time, colleagues from the Oil Change International were preparing to launch the ‘Sky’slimit’ report, and when I was asked to review the report I noticed that they were interpreting the IPCC AR5 CBs differently than I was for the CV1.5. The IPCC AR5 provides the 2011 CB and following typical accounting practice I had assumed that the 2011 CB meant the budget available at the beginning of 2011. However, the Oil Change International colleagues were interpreting it as the CB available at the end of 2011 (what would be the 2012 CB for me). We engaged on a discussion about this, and we ended up involving into the discussion two of the principal authors behind the IPCC AR5 CB estimates, and they told us that the Oil Change International interpretation was the right one (in fact they got it from those authors on the first place), and therefore that the reported values in the IPCC AR5 report were in fact the 2012 CBs associated to conventional accounting practice. So I changed my interpretation for the CV1.5 analysis. 

However, now in the IPCC SR1.5, with one of these authors being among the main authors for chapter-2 of the current SR1.5 where CBs are discussed, the IPCC SR1.5 states that the IPCC AR5 CBs are really 2011 budgets! Hence, in the results I presented here I adopted the 2011 interpretation for the AR5 CBs, and that is the reason why the CV1.5 analysis is based on higher CBs (the difference are the emissions in year 2011).

But the issue is not trivial. This very elemental accounting error has huge relative implications in the already very meager CBs. For instance, this accounting error means a 18% error in the 2019 1.5°C@50% CB and a 54% error in the 2019 1.5°C@66% CB.

The corollary for me is that if this happens with something as simple as an accounting practice, what other uncertainties could be embedded into the reported CBs? Moreover, the needed conservative approach by IPCC could bias uncertainties. Hence, I think we should change our approach and give up the strategy of building evolution pathways and policies that seek to match specified CBs (or specific warming goals) by adjusting the amount of FF that we still can burn, and rather focus on deploying an appropriate and structural transition as fast as possible. CBs will keep on being a very useful communication instrument, but shouldn’t become an excuse for allowing us to burn more carbon.  Otherwise we are just playing with fire.