08 June 2014

Carbon budgets, climate sensitivity and the myth of "burnable carbon"

by David Spratt

Breakthrough National  Climate Restoration
Forum 21-22 June,  Melbourne
In my previous post explaining why there is no carbon budget left for burning fossil fuels for a 2-degree Celsius (°C) target, I explained that these carbon budget calculations are expressed in probabilities of not exceeding the target. This reflects a number of uncertainties in understanding, including climate sensitivity, ocean heat uptake inertia, the influences of non-carbon dioxide forcing agents, and because results vary somewhat among model ensembles.

Of these, climate sensitivity is the biggest issue, because of the possibilities that climate change might proceed more rapidly than currently estimated because of reinforcing feedbacks, thresholds or tipping points in the climate system, or less rapidly because of dampening feedbacks. 
Listen to David's carbon budget interview on Radio EcoShock

The real budgetary emergency and the myth of "burnable carbon"

No carbon left to burn
(audio + slides, 17 minutes)
Another significant issue is whether the modelling used for the most recent IPCC report is too conservative in projecting the loss of Arctic sea ice, and the consequences for Arctic-driven warming.

This post looks at these two issues

Arctic modelling underestimates sea-ice loss, albedo change and warming

The IPCC's 2013 carbon budget work uses Coupled Model Intercomparison Project Phase 5 (CMIP5) computer modelling results. Results are given for the RCP2.6 (~2°C warming) scenario, which show a 43% reduction in September Arctic sea-ice extent by 2100 (compared to a 1985–2005 reference period), so that ~ 4 million square kilometres on Arctic sea ice still remains in summer by 2100.

This is so at odds with the reality on the ground as to be not credible.  With  less than 1°C of global warming, the Arctic sea-ice extent has dropped by half, and the sea-ice volume by three-quarters. At the 2012 summer minimum, sea-ice extent was 3.4 million square kilometres, less than the CMIP5's projection for 2100!  

Many Arctic experts think that the Arctic is likely to reach an sea-ice-free state (defined as less than 1 mil. sq. km.) in the northern summer within the next decade or so, and perhaps sooner, with the number of ice-free days growing from then on. Prof. Will Steffen told "The Age" in September 2012 that: “I’m pretty certain that we have now passed the tipping point for Arctic sea ice”.  This reflects work by researchers include Livina, Lenton, Wadhams and Maslowski (1).

A reasonable scenario would be to look at a sea-ice-free Arctic in five-to-ten years, with the number of ice-free days expanding from then on to several weeks, perhaps even months, before +2°C is achieved. This is important because a more rapid loss of sea ice changes the Arctic's albedo (reflectivity), as dark seas absorb more heat than white ice, increasing warming. This feedback effectively squeezes down the carbon budget, and is underestimated in IPCC's 2013 report.

A 2011 study, for example, found that if the Arctic were ice-free for one month a year plus associated ice-extent decreases in other months then, without taking cloud changes into account, the global impact would be about 0.2ºC of warming. If there were no ice at all during the months of sunlight, the impact would close to 0.5ºC of global warming (2).

It is a very credible scenario that the Arctic could indeed be sea-ice-free for a month in summer before warming reaches 2°C, but this has not been considered in any carbon budget considerations as far as I can ascertain. Warming of 0.2°C from a month of sea-ice-free conditions is roughly equivalent to ten years of current human emissions, which would have to be subtracted from the IPCC's 2013 carbon budget, reducing it by around 40%. 

Climate sensitivity

Short-term, or Equilibrium Climate Sensitivity (ECS), is the temperature increase resulting from doubling of atmospheric carbon dioxide (CO2) levels, including such factors as rapid changes in snow and (sea) ice melt, and the behaviour of "fast" feedbacks including clouds and water vapour. Thus, doubling of atmospheric CO2 from the pre-industrial level of 280 parts per million (ppm), to 560 ppm, would resulted in a 3°C global temperature increase using ECS of 3°C.  A related function is the Transient Climate Response to Cumulative CO2 Emissions (TCRE), which is used in the IPCC's carbon budget.

Arctic sea-ice loss, and the associated albedo change, is a fast feedback that is included in the CMIP5 models for IPCC AR5 carbon budgets, but as discussed above, that process appears to have been significantly underestimated.

The mid-range ECS estimate is generally around 3°C (range 2–4.5°C), and it plays out in the first hundred years or so after an injection of carbon dioxide into the atmosphere. The 2013 IPCC report finds that "Equilibrium climate sensitivity is likely in the range 1.5°C to 4.5°C" but "No best estimate for equilibrium climate sensitivity can now be given because of a lack of agreement on values across assessed lines of evidence and studies."

However a recent paper by Sherwood, Bony et al. looking at clouds and atmospheric convective mixing finds that on “the basis of the available data… the new understanding presented here pushes the likely long-term global warming towards the upper end of model ranges.” Taking “the available observations at face value,” they write, “implies a most likely climate sensitivity of about 4°C, with a lower limit of about 3°C” (3).

Writing in the Guardian, Skeptical Science's Dana Nuccitelli explains that these "results are consistent with Fasullo & Trenberth (2012), who found that only the higher sensitivity climate models correctly simulated drying in key cloud-forming regions of the atmosphere. Likewise, preliminary results by scientists at the California Institute of Technology Jet Propulsion Laboratory presented at the 2013 AGU meeting showed that higher sensitivity models do the best job simulating observed cloud changes. These results are also consistent with Lauer et al. (2010) and Clement et al. (2009), which looked at cloud changes in the Pacific, finding the observations consistent with a positive cloud feedback" (4).

If indeed ECS is more likely at the higher end of the range, this would diminish the remaining carbon budget.  Quantifying a carbon budget for a ~4°C mid-point ECS has not been done as far as I can ascertain. 

Long-term earth system sensitivity

Paleoclimatology (study of past climates) suggests that if longer-term feedbacks of "slow" factors are taken into account, such as the decay of large ice sheets, changes in the carbon cycle (changed efficiency of carbon sinks such as permafrost and methane clathrate stores, as well as biosphere stores such as peatlands and forests), and changes in vegetation coverage and reflectivity (albedo), then the Earth's sensitivity to a doubling of CO2 could itself be double that of the "fast" climate sensitivity predicted by most climate models, or around 6°C (5). These "slow" feedbacks amplify the initial warming burst. A measure of these effects for a doubling of CO2 is known as Earth System Sensitivity (ESS).

Longer-term ESS is generally considered to come into play over periods from centuries to several millennia, depending on how fast is the rate of change in greenhouse gas levels and temperature.

The problem is that rate of climate change now being driven by human actions may be as fast as any extended warming period over the past 65 million years, and it is projected to accelerate in the coming decades. This means that longer-term "slow" events associated with ESS – such as loss of large ice sheets, and changes in Arctic and biosphere carbon stores – are starting to occur now, are happening much more quickly than expected, and likely will proceed at a significant scale in the current hundred years. We face an event unprecedented in the last 65 million years of "fast" short-term and "slow" long-term climate sensitivity events occurring alongside one another in parallel, rather than one after the other in series as is usually the case. Thus, even as some of the "fast" warming is still to be realised due to thermal inertia, some of the "slow" feedbacks are already coming into play:
Evidence from Earth’s history suggests that slower surface albedo feedbacks due to vegetation change and melting of Greenland and Antarctica can come into play on the timescales of interest to humans, which could increase the sensitivity to significantly higher values, as much as 6°C … the slow feedback climate sensitivity has relevance in the Anthropocene era, since ice sheet/vegetation feedback may become significant on decadal-to-centennial timescales of interest to humans (6).
Unfortunately, slow feedbacks are amplifying on time scales that humans care about: decades, centuries, even millennia. As the planet warms, for example, ice sheets melt, exposing a darker surface that increases warming. Also warming causes a net release of long-lived greenhouse gases from the ocean and soil. Vegetation changes that occur as climate warms from today's situation will also have a significant amplifying effect, as forests move into tundra regions in North America and Eurasia (7).
The problem is that the IPCC carbon budget analysis assumes that none of these longer-term feedbacks will be materially relevant before 2°C of warming, and so exclude the possibility of large-scale permafrost, methane clathrate or less efficient biological stores (Amazon, tundra etc) making contributions to atmospheric greenhouse gas levels and impacting on the carbon budget.

Thus the IPCC 2013 report notes that "Accounting for ... the release of greenhouse gases from permafrost will also lower..." the target, and that the CMIP5 modelling used for the IPCC's carbon budgets does not include "explicit representation of permafrost soil carbon decomposition in response to future warming". It also notes that "the climate sensitivity of a model may... not reflect the the sensitivity of the full Earth system because those feedback processes ["slow feedbacks associated associated with vegetation changes and ice sheets"] are not considered".

Several lines of evidence suggest theses assumptions are not robust. Recent research shows that the Amazon may often be releasing huge quantities of CO2 to the atmosphere, acting not as a carbon sink but as a source (8); and that the seafloor off the coast of Northern Siberia is releasing more than twice the amount of methane as previously estimated and is now on par with the methane being released from the Arctic tundra (9).

In February 2013, scientists using radiometric dating techniques on Russian cave formations to measure historic melting rates warned that a +1.5ºC global rise in temperature compared to pre-industrial was enough to start a general permafrost melt. They found that “global climates only slightly warmer than today are sufficient to thaw extensive regions of permafrost.” Lead researcher Anton Vaks says that: “1.5ºC appears to be something of a tipping point” (10).

In 2011, Schaefer, Zhang et al. warned: "The thaw and release of carbon currently frozen in permafrost will increase atmospheric CO2 concentrations and amplify surface warming to initiate a positive permafrost carbon feedback (PCF) on climate…. [Our] estimate may be low because it does not account for amplified surface warming due to the PCF itself….We predict that the PCF will change the Arctic from a carbon sink to a source after the mid-2020s and is strong enough to cancel 42-88% of the total global land sink. The thaw and decay of permafrost carbon is irreversible and accounting for the PCF will require larger reductions in fossil fuel emissions to reach a target atmospheric CO2 concentration" (11).

This very strong and disturbing finding – that permafrost decay is "irreversible" and requires a lower carbon budget – is not reflected in the PCC's figuring.


Climate change with its non-linear events, tipping points and irreversible events –  such as mass extinctions, destruction of ecosystems, the loss of large ice sheets and the triggering of large-scale releases of greenhouse gases from carbon stores such as permafrost and methane clathrates – contains many possibilities for catastrophic failure. 

If climate sensitivity is, in reality, at the high end of the range used for the IPCC's carbon budgets, then as a consequence that means that we must adopt a very low-risk of exceeding the target.   As the previous post explained, If a risk-averse (pro-safety) approach is applied – say, of less than 10% probability of exceeding the 2°C target – to carbon budgeting, there is simply no budget available, because it has already been used up. The notion that there is still "burnable carbon" is a myth. 


(1) Livina, V.N. and T.M. Lenton (2013) “A recent tipping point in the Arctic sea-ice cover: abrupt and persistent increase in the seasonal cycle since 2007”, The Cryosphere  7: 275-286; UWA (2012) “Arctic scientist warns of dangerous climate change”, http://www.news.uwa.edu.au/201201304303/climate-science/arctic-scientists-warn-dangerous-climate-change, accessed 30 July 2013; Wadhams, P. (2012) “Arctic ice cover, ice thickness and tipping points”, AMBIO 41: 23–33 ; Maslowski, W., C.J. Kinney et al. (2012) "The Future of Arctic Sea Ice", The Annual Review  of Earth and Planetary Sciences, 40: 625-654
(2) Hudson S. (2011) “Estimating the global radiative impact of the sea ice–albedo feedback in the Arctic”, JGRA, 16 August 2011; For a more detailed discussion, see: http://www.climatecodered.org/2012/10/after-arctic-big-melt-1-hotter-planet.html
(3) Sherwood, S.C., S. Bony et al. (2014) "Spread in model climate sensitivity traced to atmospheric convective mixing, Nature 505: 37-42
(4) Nuccittelli, D. (2014) "Global warming is being caused by humans, not the sun, and is highly sensitive to carbon, new research shows", The Guardian, 10 January 2014.
(5) The Geological Society (2013) "An addendum to the Statement on Climate Change: Evidence from the Geological Record", London, December 2013,  www.geolsoc.org.uk/climatechange; Hansen, J. (2013) "Climate Sensitivity, Sea Level and Atmospheric Carbon Dioxide", Philosophical Transactions of the Royal Society A, 371, 20120294, doi:10.1098/rsta.2012.0294.
(6) Previdi, M., B.G. Liepert et al (2011) "Climate sensitivity in the Anthropocene", Earth Syst. Dynam. Discuss., 2, 531–550
(7) James Hansen, An Old Story, but Useful Lessons, 26 September 2013, http://www.columbia.edu/~jeh1/mailings/2013/20130926_PTRSpaperDiscussion.pdf
(8) Kirby, A. (2014) "Drought ‘makes Amazonia emit carbon’", Climate News Newtwork, 5 March 2014, http://www.climatenewsnetwork.net/2014/03/drought-makes-amazon-emit-carbon, accessed 7 April 2014; Brando, P.M., J.K.Balch et al. (2014) "Abrupt increases in Amazonian tree mortality due to drought–fire interactions", PNAS, doi: 10.1073/pnas.1305499111
(9) Science Daily (2013), "Arctic seafloor methane releases double previous estimates", 25 November 2013, http://www.sciencedaily.com/releases/2013/11/131125172113.htm, accessed 7 April 2014.
(10) Vaks, A., O.S. Gutareva et al. (2013) “Speleothems Reveal 500,000-Year History of Siberian Permafrost”, Science 340: 183-186
(11) Khvorostyanov, D.V., P. Ciais et al. (2008) “Vulnerability of east Siberia's frozen carbon stores to future warming”, Geophysical Research Letters, 2008; 35:L10703; Schaefer, K., T. Zhang et al. (2011) "Amount and timing of permafrost carbon release in response to climate warming", Tellus 63:165-180.