On Quora someone asked:
What is the most agreed-on figure for our future carbon budget?
My answer:
Asking “what is our future carbon budget?” is a bit like asking how many calories a day you can eat. There’s really no limit on how much you can eat if you don’t care how overweight and unhealthy you become. So, to set a carbon budget, you need to say how much global warming you will accept.
That said, here’s a picture of how we’re burning through our carbon budget:
It says that our civilization has burnt 60% of the carbon we’re allowed to while still having a 50-50 chance of keeping global warming below 2 °C.
This chart appears in the International Energy Agency report World Energy Outlook Special Report 2015, which is free and definitely worth reading.
The orange bars show CO2 emissions per year, in gigatonnes. The blue curve shows the fraction of the total carbon budget we have left to burn, based on data from the Intergovernmental Panel for Climate Change. The projection of future carbon emissions is based on the Intended Nationally Determined Contributions (INDC) that governments are currently submitting to the United Nations. So, based on what governments had offered to do by June 2015, we may burn through this carbon budget in 2040.
Our civilization’s total carbon budget for staying below 2 °C was about 1 trillion tonnes. We have now burnt almost 60% of that. You can watch the amount rise as we speak:
Quoting the International Energy Agency report:
The transition away from fossil fuels is gradual in the INDC Scenario, with the share of fossil fuels in the world’s primary energy mix declining from more than 80% today to around three-quarters in 2030 […] The projected path for energy-related emissions in the INDC Scenario means that, based on IPCC estimates, the world’s remaining carbon budget consistent with a 50% chance of keeping a temperature increase of below 2 °C would be exhausted around 2040, adding a grace period of only around eight months, compared to the date at which the budget would be exhausted in the absence of INDCs (Figure 2.3). This date is already within the lifetime of many existing energy sector assets: fossil-fuelled power plants often operate for 30-40 years or more, while existing fossil-fuel resources could, if all developed, sustain production levels far beyond 2040. If energy sector investors believed that not only new investments but also existing fossil-fuel operations would be halted at that critical point, this would have a profound effect on investment even today.
Since we seem likely to go above 2 °C warming over pre-industrial levels, it would be nice to make a similar chart for a carbon budget based on 3 ° C warming. The Trillionth Tonne website projects that with current trends we’ll burn 1.5 trillion tonnes, for a warming of 3 °C in a cautious scenario, by 2056.
But: we would never burn the 1.5 trillionth tonne if emissions dropped by 1.2% per year from now on. And we’d not even burn the trillionth tonne if they dropped by 2.6% per year.
Azimuth 
Given that the climate sensitivity has an uncertainty of at least +/-50% (using the range of climate model projections as a proxy for uncertainty), is there not a commensurate level of uncertainty in how much we can burn to stay below a specific level of warming?
Yes. If you look at TrillionthTonne you’ll see you can get different numbers by choosing different assumptions on climate sensitivity and how much warming you’re willing to endure.
The International Energy Agency graph I showed says how much we can burn to have a 50% chance of staying below 2 °C, so they are essentially choosing a probability distribution on climate sensitivities and saying that with 50% chance the climate sensitivity is low enough to stay below 2 ° C warming in their scenario.
I haven’t had time to check, but since they cite the IPCC, I’m hoping they chose this probability distribution based on the latest IPCC report.
The summary for policymakers in this report says:
Here “likely” means “greater than 66% chance”, “very unlikely” means less than 10% chance, and so on. You can see more details and graphs of the probability distribution function of the climate sensitivity in Box 12.2 on pages 1110-1111 in the report.
I think it also should be noted that it’s not just about probabilities here. Two matters.
First, the cost to civilization of encountering an error to the high side of the density is not the same as encountering an error to the low side. For, if the actual sensitivity is lower than the midpoint, some of the mitigation done, assuming we do cut emissions, will be done sooner than strictly needed, even if it is eventually needed. If the actual sensitivity is higher than the midpoint, then we may incur costs to civilization in excess of those which would be borne if the sensitivity were at the midpoint, because our mitigation by cutting emissions is not enough. It also depends what should be minimized … For instance, a case could be made to use something like the 0.8 quantile of climate sensitivity since the associated damages to civilization would then be outrageously high.
Second, climate sensitivity is not a constant. It is known, for instance, that climate sensitivity goes up as temperatures go up. So, should we, say, allow temperatures to get to the +3 degree Celsius range, it’s not correct to expect an increment of additional increase of carbon dioxide in atmosphere will have the same effect as it would at +1 degree Celsius (essentially where we are now). Just how much that sensitivity increases with temperature is an open question, some of which are addressed here and there’s a plan to study these here.
All that said, I think this is a bit of fiddling while Rome burns, this arguments about temperature sensitivity. In fact, emissions from most countries continue to increase, let alone decrease, and some of them are accelerating. Also, no one can expect any economy to “stop on a dime”. Finally, the closer one gets to the end of the budget, any budget, the more rapid the deceleration needed so as not to overshoot it. The more rapid the deceleration the more damage to the economies which attempt it, including risk of recession. This is the basis for Thomas Stocker’s paper “The closing door of climate targets“.
Atmospheric CO2 has increased by ~42% since 1870, yet global average surface temp has already increased close to 1 deg C.
Not clear what you mean here. If you mean we’ve already seen +1 deg C, then, yes. we have. If you mean we’ve ONLY seen +1 deg C after 40% atmospheric CO2, then IMO you do not understand this question at all. Commenting would seem to require a thourough understanding of what should be done and how.
Observations suggest that ECS is likely to be well beyond 1 deg C, and it’s well known that ECS is not a fixed value.
Climate change ethics shows clearly that there’s a moral obligation to act, since there’s already clear evidence of harm, especially to people of poorer nations who have least contributed to the problem and have not consented to be put at risk. Furthermore, economic arguments cannot be ethically used as justification for inaction.
The IEA usually works with a carbon budget that’s estimated to give us a 50-50 chance of keeping global warming below 2K. However it’s arguable that a sensible carbon budget aimed at keeping global warming below 2K should give better odds than 50-50. The IPCC WG3 provides a useful chart in their Summary for Policy Makers (AR5; p13) that outlines a range of carbon budgets, and their likelihood of producing warming that stays below various thresholds.
Speaking of carbon budgets, you may find this paper interesting, if you haven’t seen it already: http://www.nature.com/nature/journal/v517/n7533/full/nature14016.html
It explores the geographical distribution of fossil fuel resources left in the ground in a world where humanity sticks to a 2C carbon budget.
I’m stumbling a little on the arithmetic here. The orange bars presumably give CO2 (not carbon!) emissions in GtCO2/year. This rate maxes out at what looks like 35 GtCO2/year, in 2040. The ratio between CO2 emissions and carbon emissions happens to be about 3.5, so this gives a max carbon emission rate of 10GtC/year before 2040. But there are only 25 years between now and 2040, so in that time we would only emit ~ 250 of the 400 Gtc remaining in our carbon budget.
But, the plot legend says “energy-related emissions”, so this is presumably not counting cement production and land-use change. Recent numbers (for instance,
here
and
here
)
suggest that taking these sources into account raises emissions by ~ 15%. But that raises our estimate of max cumulative emissions between now and 2040 to ~ 290 GtC, still far from 400 GtC. Is the percent of total CO2 emissions which comes from fossil fuels supposed to change drastically in decades to come? If not, how is the preceding arithmetic consistent with the plot and the claim that we’ll emit 400 GtC in the next 25 years ?
jeevanje wrote:
Yes, my comment about that was confusing. Fixed!
I’m too tired to tackle your other issues tonight, though they’re very much worth thinking about. I hope someone solves them while I’m snoozing!
So the links you’ve provided just talk about CO2 emissions, which are responsible for about for roughly three quarters of anthropogenic GHG emissions. I think the problem you’re having is resolved when you account for non-CO2 GHGs.
According to the IPCC AR5 WG3 SPM, total anthropogenic GHG emissions from all sources were ~49 Gt CO2eq in 2010, with CO2 from fossil fuels accounting for ~32 Gt of that. This seems roughly consistent with the IEA chart, which seems to give a value of a little less than 32 Gt for energy-related CO2 emissions in 2010. 49/32 = ~1.5, so taking into account non-CO2 GHGs raises emissions by ~50%, as compared to the increase of ~15% you calculated based on including just non-fossil fuel CO2. This raises the estimate for cumulative emissions from now to 2040 to ~375 GtC, which makes more sense.
The IEA chart appears slightly misleading in this context, IMO, as it sort of makes it look like we’re drawing down our carbon budget just though CO2 emissions, when in fact there are a whole range of non-CO2 forcings that need to be accounted for, including non-GHG forcings such as aerosols and albedo changes due to land conversion.
Great, thanks. The apparently quite significant contribution from non-CO2 GHGs was not part of my zeroth-order picture of climate change (and plots such as the EIA one in this post do not help in that regard, as you point out), but it is now. Much appreciated!
Thanks, Alex!
I agree the IEA chart is potentially misleading, though technically honest: they say the orange bars are ‘energy-related emissions’, but one might get the idea that this is the only thing causing the blue curve to drop to zero. Their focus on energy-related emissions probably comes from their overall focus: they’re the International Energy Agency, after all.
Reblogged this on Simple Climate and commented:
This is a very clear analysis – upshot: humanity can burn a trillion tonnes of carbon before being committed to the 2°C warming limit governments have agreed. We’re 60% of the way there, and on course for 1.5 trillion, which would commit us to 3°C.