RealClimate

Climate science from climate scientists...

Blog – realclimate.org – All Posts

The IPCC Sixth Assessment Report

by

Climate scientists are inordinately excited by the release of a new IPCC report (truth be told, that’s a bit odd – It’s a bit like bringing your end-of-(seven)-year project home and waiting anxiously to see how well it will be received). So, in an uncharacteristically enthusiastic burst of effort, we have a whole suite of posts on the report for you to read.

If/when we add some more commentary as we digest the details and we see how the report is being discussed, we’ll link it from here. Feel free to discuss general issues with the report in the comments here, and feel free to suggest further deep dives we might pursue.

Net Zero/Not Zero

by 100 Comments

At the COP26 gathering last week much of the discussion related to “Net-Zero” goals. This concept derives from important physical science results highlighted in the Special Report on 1.5ºC and more thoroughly in the last IPCC report that future warming is tied to future emissions, and that warming will effectively cease only once anthropogenic CO2 emissions are balanced by anthropogenic CO2 removals. But some activists have (rightly) pointed out that large-scale CO2 removals are as yet untested, and so reliance on them to any significant extent to balance out emissions is akin not really committing to net zero at all. Their point is that “net-zero” is not zero and hence will serve as a smokescreen for insufficient climate action. To help sort this out some background might be helpful.

[Read more…] about Net Zero/Not Zero

Forced responses: Nov 2021

by 390 Comments

A bi-monthly open thread related to climate solutions. This month will start off with COP-26 and many targets and plans and mechanisms will be proposed and discussed. Look out for the updated impacts of the evolving NDCs such as this one from Climate Resource, suggesting that the world could be on track for just a little less than 2ºC warming (relative to the pre-industrial) (if everyone does what they pledge and we are lucky with respect to climate sensitivity). Please be respectful and constructive.

Unforced variations: Nov 2021

by 119 Comments

This month’s open thread. The first two weeks will be dominated by COP-26, and various science updates that will be announced there, including this year’s Global Carbon Project report. Curiously, there is some archival interest in the climategate affair possibly in connection to COP-26 (a BBC dramatization “The Trick“, a BBC radio series on the security aspects “The Hack that Changed the World”, and a couple of months ago, a podcast episode of “Cheat!”). Please stick to science-related topics on this thread.

A science-based move to climate change adaptation

by 76 Comments

All countries in the world urgently need to adapt to climate change but are not yet in a good position to do so. It’s urgent because we are not even adapted to the present climate. This fact is underscored by recent weather-related calamities, such as flooding in Central Europe and heatwaves over North America. It’s also urgent because the oceans act like a flywheel, making sure that cuts in emission of greenhouse gases will have a lagged effect on global warming.

Climate change adaptation was addressed in the Paris Agreement from 2015, the Climate Adaptation Summit in January 2021, and will be one of four key priorities during the upcoming COP26. Proper climate adaptation of course needs meteorological and climatological data for mapping weather-related risks to prepare us for future extreme weather. However, I would argue that the climate research community has not had a visible presence during any of these meetings. Instead the summits have been dominated by politicians and NGOs.

[Read more…] about A science-based move to climate change adaptation

Tributes to Geert Jan van Oldenborgh

by 3 Comments

As many of you will know, Geert Jan van Oldenborgh died on Oct 12, 2021, and in the last week a number of very touching tributes have appeared. Notably, a lovely obituary in the NY Times by Henry Fountain, a segment on the BBC’s Inside Science from Roland Pease, a piece on Bloomberg News by Eric Roston and, of course, an appreciation from his colleagues at World Weather Attribution (including Friederike Otto, the co-recipient of the TIME 100 award to Geert earlier this year).

Geert’s work had been featured often at RealClimate (notably the rapid attribution work for the Pacific North West heat wave earlier this year), and we have made frequent reference to Climate Explorer, the tool he built to provide easier access to many sources of climate data. He also provided us with annual updates for the comparison between a 1981 climate projection to subsequent observations.

Comparison of Hansen et al (1981) projections to GISTEMP observations (to 2019) (credit Geert Jan van Oldenborgh).

He let us know earlier this year that this was likely the last update. Moge hij rusten in vrede.

A Nobel pursuit

by 106 Comments

Klaus Hasselmann and Suki Manabe

Last week, the Nobel physics prize was (half) awarded to Suki Manabe and Klaus Hasselmann for their work on climate prediction and the detection and attribution of climate change. This came as quite a surprise to the climate community – though it was welcomed warmly. We’ve discussed the early climate model predictions a lot (including some from Manabe and his colleagues), and we’ve discussed detection and attribution of climate change as well, though with less explicit discussion of Hasselmann’s contribution. Needless to say these are big topics which have had many inputs from many scientists over the years.

But RC has a more attuned audience to these topics than most, and so it might be fun to dive into the details of their early work to see what has stood the test of time and what has not, and how that differs (if it does) from their colleagues and rivals at the time.

Manabe’s Climate Modeling

Fortunately, Manabe recently wrote a retrospective on his early work in response to receiving the Crafoord prize in 2018. That paper (Manabe, 2019) gives a good overview of Manabe’s particular philosophy of climate modeling which was very much focused on getting things to work, and not worrying too much about the details. He makes a eloquent argument for a hierarchy of modeling where simpler, functional, models can contribute a lot to understanding in advance of the more complete and more detailed versions turning up. In this, he is in violent agreement with Isaac Held, his colleague at GFDL, and indeed most climate scientists.

But let’s go back to the beginning. Manabe’s early focus was on radiative-convective equilibrium, and his seminal 1967 paper (with his longtime collaborator Richard Wetherald, who passed in 2011). The Manabe and Wetherald (1967) paper has been described as the most influential climate paper ever.

The key aspects were the inclusion of water vapour feedback as temperatures increased, and the use of ‘convective adjustment’ to maintain stability of the lower atmospheric column. While not a great parameterization of the complexity of real convection, it served to keep the troposphere and surface linked in ways that match what happens in the real world. In practice, it was a big advance towards realism over the work of Plass or Möller from a few years before (despite the lack of cloud feedback). Two examples of the sensitivity of their model (which have mostly held up) are useful to look at at:

Two results from Manabe and Wetherald (1967)

What they showed are the distinct fingerprints of two kinds of forcing; increasing solar activity which warms all parts of the atmosphere, and carbon dioxide increases which warm the surface and troposphere, but cool the stratosphere and above. The source of this result is the spectral resolution of the radiative transfer model they were using, but oddly enough they don’t discuss it at all. In a subsequent short paper Manabe (1970), Manabe extends this result to predict a temperature increase by 2000 of 0.8ºC based on a 25% increase in CO2, which was pretty close. (Funnily enough, this paper appeared in volume about environmental risks that was edited by a young(er) S. Fred Singer, before his turn to the dark side).

Manabe’s subsequent work led to the development of the GFDL GCM, initially just including the atmosphere, but eventually with an ocean, and then the transient results shown in Manabe and Stouffer (1993). Famously, these early results were half the input into the Charney report‘s estimate of climate sensitivity in 1979 (the other half being the preliminary results from Jim Hansen’s model at GISS). Both these predictions have been evaluated in recent years to see how well they did. The time series were included in the Hausfather et al (2020) paper and in the latest IPCC report:

Evaluation of old projections of global temperature, including Manabe (1970) and Manabe and Stouffer (1993)

The next step in climate modeling was to couple dynamic ocean models to the atmospheric models, and again, Manabe and his colleagues were pioneers (notably Manabe and Bryan (1969), but more comprehensively in Manabe et al. (1975), and Bryan et al. (1975)). But as expectations increased that coupled models could help climate predictions, there was a growing realization that there was a problem with how they were being designed.

The basic issue stems from the different timescales of the ocean and atmosphere. Given the ocean temperatures, an atmospheric model will equilibriate in a year or so of model time (maybe a decade if you care about the water vapour distribution in the upper stratosphere). However, given information from the atmosphere, an ocean model takes centuries to millennia to equilibriate the deep ocean. The tail of the age distribution for water parcels in the deep Pacific can reach 10,000 years or so. But back in the day, running a coupled model anything like that long was prohibitively expensive. So in order to get a coupled model simulation for near-present, the ocean needed to be ‘spun-up’ for a good while on it’s own, and then, once it was in equilibrium, the coupling was turned on, and voila! a coupled simulation of the present-day climate. Except…

… it didn’t generally work. The newly coupled model would drift away from today’s climate, sometimes with a collapse of the Atlantic overturning circulation, other times just towards a much warmer or cooler climate, or with a terrible ‘double ITCZ’. This was a problem because it’s not at all clear that the sensitivity of the simulated climate (which was off in serious ways) would be the same as the sensitivity of the real world. This stymied progress for a while (maybe a decade or so) as people worked to understand why the models drifted so much, and to find ways to fix it.

When I started in climate modeling (in the early 1990s), this was still a relevant issue, though two approaches had been adopted. One, advocated by Manabe’s group (and Hasselmann’s!), was the imposition of ‘flux corrections’ or ‘flux adjustments’ (Manabe and Stouffer, 1988; Sausen et al, 1988) which added artificial fluxes at the ocean-atmosphere boundary that gave the ocean and atmosphere what they both needed to stay stable, correcting for what would have been calculated, and then keeping that fixed in all future sensitivity experiments. This (by design) produced a good climatology, but effectively buried the models’ poor physics. The other approach was to work with models that had offsets from the real world (which you would keep trying to reduce) but would have sensitivities that were more physically coherent.

The implications of the two approaches are difficult to assess without a perfect model simulation to compare against, and if we had that, there’d be no need to worry about drifts. Thus during the early 90s there was a fair bit of unresolved religious-like discussions about what should be done. Manabe was vocal that you needed a reasonable model to play with and make progress, while others were of the opinion that the sensitivity of a flux-corrected model wasn’t informative of the real world, and that using flux corrections as a crutch, was actually holding back work on the physics that would (eventually) remove the need for the corrections in the first place. (Minor aside, I was a co-author on a paper that assessed this concept for a slightly simpler class of model, and found that the ‘flux-corrected’ version was not predictive of the ‘true’ sensitivity Bjornsson et al., 1997).

Over time the issue more or less resolved itself as models got incrementally better and computational resources increased so that longer coupled model simulations could be done more routinely. Occasionally, the issue still comes up (i.e. Gnanadesikan et al., 2018), but I think it’s fair to say that few modelers think it’s a useful tool anymore. For context, 10 out of 17 models in CMIP2 (~1995) used flux adjustments, and 6 out of 24 in the CMIP3 ensemble (~2001), but none in CMIP5 or CMIP6, while each generation has greater skill than the previous one. In his 2018 retrospective, Manabe doesn’t discuss the issue at all.

The proof of the pudding in climate model terms though are the quality and skill of the predictions. A recent paper Stouffer and Manabe (2017), assessed how good the Manabe and Stouffer (1989) predictions were. These came from an idealized 1% increasing CO2 experiment after 70 years, when CO2 has approximately doubled, and so is warmer than we would expect for 2020, but the pattern is quite robust:

Comparison of the pattern of warming from Manabe and Stouffer (1989) compared to the subsequent observations.

Not too shabby!

Hasselmann’s Statistical Insights

[I have to admit to not knowing Hasselmann’s oeuvre as well as Manabe’s, and to my recollection I don’t think we’ve met, so this might need some amendment…]

I think the key paper to look at is Hasselmann (1979), which really set the stage for formal methods of detection and attribution of climate change. Later papers, notably Hasselmann (1997)(pdf) extended this to multi-variate attribution problems (written in tensor notation no less, so that probably helped 😉). The basic idea is that although there are a vast number of degrees of freedom in the atmosphere/climate system, you can make a lot of progress by reducing the degrees of freedom and looking just at the dominant modes of variability changes and comparing them with expected patterns from simulations. A key insight is that depending on how the noise and the forced patterns line up, the ‘optimal’ pattern to detect might not be what you first thought. But note that this was written when “continuous model time series of comparable length to analyzed global or hemispheric data [were] not available”, so the paper is mainly conceptual. It really is only in the late 1980s, and more clearly and with more models, the mid-1990s, that the data became available to really apply these methods.

The only figure in Hasselmann’s 1979 paper.

The challenge with all detection & attribution (D&A) work is that it must rely on counter-factuals – i.e. estimates of how the climate would behave in special cases – for instance, if the only forcing was greenhouse gases, or if there was only natural forcings or only internal variability. Since the real world has all of these things going on at the same time, it’s hard to extract them from the observations, particularly since good direct observations don’t stretch back more than a century or so, and proxy climate observations have their own, increased, uncertainties. But even with perfect observations, getting a full characterization of internal variability would be hard, and perhaps impossible. So in practice, the ‘noise ellipsoid’ in the above figure is almost always taken from control runs of coupled climate models which, as Hasselmann acknowledged, were not available in 1979.

Hasselmann’s work before this paper was heavily related to measurements and understanding of ocean waves and the role of ‘random’ weather forcing on long term ocean variability, and that has been widely cited, and afterwards, he played a key role in developing the MPI climate model (i.e. Cubasch et al. (1992). But much of later well-cited work built off the 1979 paper and involved further refinements on the theme of D&A, often working with Gabi Hegerl (i.e. Hegerl et al., 1996; Hegerl et al., 1997).

These were very much the ideas that set the discussions in climate science in the 1990s. As you will recall, Hansen had declared in 1988 that “the greenhouse effect is here!”, based on a 3-sigma signal detected with the original GISS model. But the ocean model used there was a simple ‘Q-flux’, no-dynamics, module, and so had no ENSO, or other coupled modes of variability. The implicit estimate of the internal variability here was, to be clear, not widely accepted. There are a couple of articles and responses at the time that give a flavor, for instance “Hansen vs. the World” by Richard Kerr reporting from a workshop where Manabe, and Hasselmann’s coauthors (notably Cubasch and Barnett) were present, and the responses from Wally Broecker and James Risbey.

Hasselmann himself commented on this in Science in 1997, after the 1995 Second Assessment Report from IPCC declared that “the balance of evidence” suggested that the greenhouse gas signal had indeed been detected. The figure he showed there:

The D&A of forced climate change as it stood in 1997 (10-year smoothed data).

… supported the IPCC conclusion, and his last line is worth repeating:

It would be unfortunate if the current debate over this ultimately transitory issue should distract from the far more serious problem of the long-term evolution of global warming once the signal has been unequivocally detected above the background noise.

Klaus Hasselmann (1997)

Today, 24 years later, the detection and attribution of anthropogenic climate change is “unequivocal”, but we are still being distracted by ultimately transitory issues…

What if the prize had been given a decade ago?

The two restrictions on the award of disciplinary Nobel prizes are that the awardees must still be alive, and that there is a limit of three laureates per prize. For advances made in the 1960s and 1970s, the first is extremely relevant, and makes the second condition somewhat less so. But without wishing to take anything away from the two awardees this year, ten years ago the decision would have been much tougher. Norman Philips published what is recognised as the first GCM in 1955 – he died in 2019. Akio Arakawa was the conceptual leader of climate modeling directly influencing both Manabe and Hansen – he died earlier this year. Of the published papers predicting global warming in the 1970s (as catalogued in the Hausfather et al paper), the authors Rasool, Schneider, Benton, Sawyer, Broecker and Mitchell have all passed. Only Nordhaus and Manabe are still alive – though now both have won Nobel prizes.

But the building of climate models and their application is broader than can be recognized like this. There are no prizes for the people that actually wrote the code for the models – people like Gary Russell or Ernst Maier-Reimer (nicely eulogized by Hasselmann), the specialists who designed the parameterizations, or the teams that developed the inputs and processed the outputs or the technicians that kept the old supercomputers running. In recent papers documenting model development, it’s not unusual to have dozens of authors – not the level of the CERN collaborations, but significantly beyond the Nobel limit. The huge advances in understanding we’ve seen since the 1970s have been the work of thousands of smart and dedicated people all around the world, only a few of which will ever be recognized as widely as this. We should always remember this while we celebrate the winners.

Finally, while it is many scientists’ dream to win a Nobel Prize, Hasselmann’s statement that he would rather have “no global warming and no Nobel Prize” captures the ambiguity that many of us feel in successfully predicting events and trends that we don’t want to come true.

References

  1. S. Manabe, "Role of greenhouse gas in climate change**", Tellus A: Dynamic Meteorology and Oceanography, vol. 71, pp. 1620078, 2019. http://dx.doi.org/10.1080/16000870.2019.1620078
  2. S. Manabe, and R.T. Wetherald, "Thermal Equilibrium of the Atmosphere with a Given Distribution of Relative Humidity", Journal of the Atmospheric Sciences, vol. 24, pp. 241-259, 1967. http://dx.doi.org/10.1175/1520-0469(1967)024<0241:TEOTAW>2.0.CO;2
  3. S. Manabe, "The Dependence of Atmospheric Temperature on the Concentration of Carbon Dioxide", Global Effects of Environmental Pollution, pp. 25-29, 1970. http://dx.doi.org/10.1007/978-94-010-3290-2_4
  4. S. Manabe, and R.J. Stouffer, "Century-scale effects of increased atmospheric C02 on the ocean–atmosphere system", Nature, vol. 364, pp. 215-218, 1993. http://dx.doi.org/10.1038/364215a0
  5. Z. Hausfather, H.F. Drake, T. Abbott, and G.A. Schmidt, "Evaluating the Performance of Past Climate Model Projections", Geophysical Research Letters, vol. 47, 2020. http://dx.doi.org/10.1029/2019GL085378
  6. S. Manabe, and K. Bryan, "Climate Calculations with a Combined Ocean-Atmosphere Model", Journal of the Atmospheric Sciences, vol. 26, pp. 786-789, 1969. http://dx.doi.org/10.1175/1520-0469(1969)026<0786:CCWACO>2.0.CO;2
  7. S. Manabe, K. Bryan, and M.J. Spelman, "A Global Ocean-Atmosphere Climate Model. Part I. The Atmospheric Circulation", Journal of Physical Oceanography, vol. 5, pp. 3-29, 1975. http://dx.doi.org/10.1175/1520-0485(1975)005<0003:AGOACM>2.0.CO;2
  8. K. Bryan, S. Manabe, and R.C. Pacanowski, "A Global Ocean-Atmosphere Climate Model. Part II. The Oceanic Circulation", Journal of Physical Oceanography, vol. 5, pp. 30-46, 1975. http://dx.doi.org/10.1175/1520-0485(1975)005<0030:AGOACM>2.0.CO;2
  9. S. Manabe, and R.J. Stouffer, "Two Stable Equilibria of a Coupled Ocean-Atmosphere Model", Journal of Climate, vol. 1, pp. 841-866, 1988. http://dx.doi.org/10.1175/1520-0442(1988)001<0841:TSEOAC>2.0.CO;2
  10. R. Sausen, K. Barthel, and K. Hasselmann, "Coupled ocean-atmosphere models with flux correction", Climate Dynamics, vol. 2, pp. 145-163, 1988. http://dx.doi.org/10.1007/BF01053472
  11. H. Bjornsson, L.A. Mysak, and G.A. Schmidt, "Mixed Boundary Conditions versus Coupling with an Energy–Moisture Balance Model for a Zonally Averaged Ocean Climate Model", Journal of Climate, vol. 10, pp. 2412-2430, 1997. http://dx.doi.org/10.1175/1520-0442(1997)010<2412:MBCVCW>2.0.CO;2
  12. A. Gnanadesikan, R. Kelson, and M. Sten, "Flux Correction and Overturning Stability: Insights from a Dynamical Box Model", Journal of Climate, vol. 31, pp. 9335-9350, 2018. http://dx.doi.org/10.1175/JCLI-D-18-0388.1
  13. R.J. Stouffer, and S. Manabe, "Assessing temperature pattern projections made in 1989", Nature Climate Change, vol. 7, pp. 163-165, 2017. http://dx.doi.org/10.1038/nclimate3224
  14. R.J. Stouffer, S. Manabe, and K. Bryan, "Interhemispheric asymmetry in climate response to a gradual increase of atmospheric CO2", Nature, vol. 342, pp. 660-662, 1989. http://dx.doi.org/10.1038/342660a0
  15. K. Hasselmann, "Multi-pattern fingerprint method for detection and attribution of climate change", Climate Dynamics, vol. 13, pp. 601-611, 1997. http://dx.doi.org/10.1007/s003820050185
  16. U. Cubasch, K. Hasselmann, H. Höck, E. Maier-Reimer, U. Mikolajewicz, B.D. Santer, and R. Sausen, "Time-dependent greenhouse warming computations with a coupled ocean-atmosphere model", Climate Dynamics, vol. 8, pp. 55-69, 1992. http://dx.doi.org/10.1007/BF00209163
  17. G.C. Hegerl, H. von Storch, K. Hasselmann, B.D. Santer, U. Cubasch, and P.D. Jones, "Detecting Greenhouse-Gas-Induced Climate Change with an Optimal Fingerprint Method", Journal of Climate, vol. 9, pp. 2281-2306, 1996. http://dx.doi.org/10.1175/1520-0442(1996)009<2281:DGGICC>2.0.CO;2
  18. G.C. Hegerl, K. Hasselmann, U. Cubasch, J.F.B. Mitchell, E. Roeckner, R. Voss, and J. Waszkewitz, "Multi-fingerprint detection and attribution analysis of greenhouse gas, greenhouse gas-plus-aerosol and solar forced climate change", Climate Dynamics, vol. 13, pp. 613-634, 1997. http://dx.doi.org/10.1007/s003820050186
  19. K. Hasselmann, "Are We Seeing Global Warming?", Science, vol. 276, pp. 914-915, 1997. http://dx.doi.org/10.1126/science.276.5314.914

Unforced Variations: Oct 2021

by

Fall is here (in the northern hemisphere at least), along with articles about the impact of climate change on autumnal colors. LandSat9 successfully launched to continue an almost 50 year long series of remote sensing (since 1972!), and the World Economic Forum has proposed and Earth Operations Center to monitor greenhouse gases and climate change. Please stick to climate science topics, and remember that (most) other commenters are real people.

The definitive CO2/CH4 comparison post

by

There is a new push to reduce CH4 emissions as a possible quick ‘win-win’ for climate and air quality. To be clear this is an eminently sensible idea – as it has been for decades (remember the ‘Methane-to-markets’ initiative from the early 2000s?), but it inevitably brings forth a mish-mash of half-remembered, inappropriate or out-of-date comparisons between the impacts of carbon dioxide and methane. So this is an attempt to put all of that in context and provide a hopefully comprehensive guide to how, when, and why to properly compare the two greenhouse gases.

Historical comparisons

First of all, let’s be clear about the relative magnitude of the gas concentrations. In 2020, CO2 was at ~410 parts per million, while CH4 was around 1870 parts per billion (or 1.87 ppm, a factor of more than 200 smaller). However the relative rise since the pre-industrial is three times larger for CH4, around 150%, compared to the 50% increase in CO2.

The radiative forcing from these changes in concentrations can be easily calculated using standard formulas (from Etminan et al, 2016 which supersede the slightly simpler ones from IPCC TAR), as about 2 W/m2 for the CO2 change and 0.65 W/m2 for CH4.

But methane’s role in atmospheric chemistry and as a source of stratospheric water vapour means that it has a bigger effect on climate than just the direct effect of its concentration. Methane emissions have a feedback on its own lifetime, serve as an ozone precursor, and reduce the production of sulphate and nitrate aerosols (and consequently indirect cloud-aerosol effects), all of which amplify its net warming effect to about 1.2 W/m2 (to about 60% of the CO2 effect since 1750). There is also a very small impact of the CH4 oxidation to CO2 itself for any fossil-fuel derived methane.

(a) Historical radiative forcing by emissions (W/m2) (IPCC fig TS.15a).(b) Historical concentrations of CO2, CH4 and N2O (IPCC AR6 fig TS.9b)

This implies that if you convert the impacts of each set of emissions into temperatures, as was done in the IPCC AR6 report, you get about 0.75ºC from the changes in CO2 and 0.5ºC for CH4 (from the late 19th C, see figure below) or 1ºC and 0.6ºC, respectively, from 1750. Thus despite the smaller concentrations and changes in methane compared to carbon dioxide, the impacts are comparable.

Historical contributions to global warming by emissions (IPCC AR6 SPM)

Stocks and flows

Before we go any further though, we need to understand that the effective perturbation time for CO2 and CH4 in the atmosphere are very different. CO2 emissions embed themselves in the atmosphere/biosphere/upper-ocean carbon cycle and have very long-term impacts (under natural conditions, some 15% of the CO2 perturbation will still be in the atmosphere thousands of years from now). In contrast, methane has a perturbation time-scale of about 12 years. This implies that the impact of CO2 on temperature is cumulative (a function of the total emitted CO2 or stock), while the impact of CH4 is a function of current (~decadal) emissions (the flows). Stabilizing temperature effects from CO2 means getting down to net-zero anthropogenic emissions, while stabilizing temperature effects from CH4 means simply stabilizing emissions.

The impacts of emissions of CH4 compared to CO2 then will have a time-varying component. Over a short time, the enhanced effectiveness of methane will be important but on very long time scales the effects of CO2 will be dominant. This is the source of the difference between the “Global Warming Potential” (GWP) numbers calculated at 20 years or 100 years which have been used for decades. You might recall that GWP is defined as the ratio on per-kg basis of the temperature impact of other greenhouse gases compared to CO2 over a specific time period. But as is clearly stated in AR6, the suitability of comparative emission metrics depends on your end goal or values.

For instance, if you use GWP-100 to trade off emissions on the way to a temperature stabilization scenario, it simply doesn’t work (since you can’t balance any net CO2 emissions with a particular level of CH4 emissions – you would need to have constantly decreasing CH4). Hence, newer concepts like GWP* have been developed that take that into account. Nonetheless, the UNFCCC (and the EPA) use the GWPs from IPCC AR4 for calculating CO2eq emissions and have not updated them as the science has progressed.

Forward-facing comparisons

People tend to be most interested in comparisons related to future choices, and it’s worth bearing in mind that while there are many ways to do this, most don’t relate to real choices that people have, nor do they clearly match up with a consistent set of values. I’ll return to that issue below. So let’s go:

  • Molecule-to-molecule concentrations: On a per-ppm basis, methane is 25 times more effective as a direct greenhouse gas. Including the indirect effects, increases that to 45 times as effective.
  • kg-to-kg: On a mass basis, methane is 70 times more effective as a greenhouse gas. This takes into account of the different molecular weights of the molecules. That would mean 126 times as effective including indirect effects.
  • kgC-to-kgC: an equal amount of kgC as CH4 or CO2 gives rise to the same ppm change, so kgC-to-kgC, methane is again 45 times more effective as a greenhouse gas.
  • kg to kg emitted: This is where it starts to get hairy because of the different timescales. Current (AR6) estimates for fossil-sourced methane are ~83 for GWP-20 and ~30 for GWP-100 (AR6 Table 7.15). (It’s slightly smaller than this for biogenic (non-fossil) methane since the oxidation product of CO2 in that case is carbon neutral). The assessed uncertainties in these values (largely related to direct and indirect aerosol effects) are ±25 and ±11. The AR4 value for methane GWP-100 was 25.
  • kgC emitted to kgC emitted: For some applications, for instance judging the impact of flaring natural gas vs. releasing it directly into the atmosphere, the kg-to-kg comparisons are not relevant, since the same amount of carbon is being emitted, rather than the same total mass. For that, the GWP-like value over 100 years, choosing to release methane directly would be 30*16/44 = 11 times worse than flaring [Corrected 9/20/21].
  • Emissions for temperature stabilization: Each additional GtC of carbon dioxide contributes to about 0.00165ºC of eventual warming (the TCRE), while a sustained TgCH4/yr of methane emissions (0.00075 GtC/yr), leads to ~3 ppb increase of methane concentrations (AR6 Table 5.2), about 0.0024 W/m2 in total radiative forcing, and, assuming a median climate sensitivity of 3ºC for 2xCO2, roughly 0.002ºC of equilibrium global warming. That implies you need a sustained reduction of 0.8 TgCH4/yr (0.0006 GtC/yr of methane) to compensate for a one-off GtC pulse of CO2.

Whatever way you slice this it implies that CH4 reductions have an outsize effect on climate, as well as an undeniably positive impact on air pollution, crop yields and public health (mainly through ozone reductions). It is therefore not a complicated decision to pursue methane reductions, taking care not to assume that doing so gets you off the hook for reducing CO2, whatever the EPA says.

I’d like this page to be useful and current, so if you think I should add an additional comparison, or use case, or if you think I’ve got something wrong, please let me know in the comments.

References

  1. M. Etminan, G. Myhre, E.J. Highwood, and K.P. Shine, "Radiative forcing of carbon dioxide, methane, and nitrous oxide: A significant revision of the methane radiative forcing", Geophysical Research Letters, vol. 43, 2016. http://dx.doi.org/10.1002/2016GL071930

Unforced Variations: Sep 2021

by

This month’s open thread for climate science topics. Not sure about you, but we are still reading the details of the IPCC report.  We are watching the unfolding hurricane season with trepidation, with particular concern related to the impacts of compound events (and not just those associated with climate), and anticipating another low, if not record, Arctic sea ice minimum.  

PS. At some point this month we will be switching Internet service providers, so don’t be surprised if there are some oddities as we switch everything over.