The Weekend Wonk: 2017 was HOT. Climate Models Affirmed. Climate May be MORE Sensitive than Thought

January 21, 2018

Damn, anyone else have that Flu from Hell?
I’ve been clobbered ever since I got back from the American Geophysical Union meeting, a month ago.
At least functional over last couple weeks, but mental cobwebs and nagging cough are annoying – starting to clear.

Anyhow, it’s 2018, I’ve got a wealth of important new interviews and footage, so let’s get going.

And Then There’s Physics:

There have been a couple of recent papers presenting analyses that claim to have narrowed the likely range for equilibrium climate sensitivity (ECS). One is Dessler et al. (currently a discussion paper under review) which suggests that the 500hPa tropical temperature better describes the planet’s energy balance and infers an ECS of 2.1K to 3.9K. The other is Cox et al. who use variability of temperature about long-term historical warming to constrain the ECS to 2.2K to 3.4K. Both suggest a narrower range than that suggested by the most recent IPCC report (1.5K to 4.5K).

James Annan has a post which suggests that these new papers are interesting but that there may be unaccounted for uncertainties. I largely agree and won’t say any more myself. I was, however, going to mention a few aspects of this that I think are relevant.

I thought how this was framed in the media was somewhat unfortunate. For example, Yes, global warming will be bad. But these scientists say it won’t reach the worst-case scenario.It does indeed seem that these studies are suggesting that the worst case scenarios might be less likely than we had previously thought. However, the public debate seems to be dominated by those who think everything will be fine (Lukewarmers) and those who are mostly in the middle of the mainstream. In fact, there is often quite a lot of pushback against any who present worst case scenarios.

The significance of these new studies to the public climate debate therefore seems to be that they largely rule out the Lukewarmer position. Yet, this is not really how they’ve been presented. One prominent Lukewarmer has even claimed that these studies are a vindication for Lukewarmers. Presenting these studies as having ruled out the worst case scenarios, rather than the best case scenarios, probably hasn’t shifted the public climate debate very much.

Another issue is that there is often an apparent confusion between climate sensitivity and how much we will warm. Yes, climate sensitivity is relevant, but so is how much we emit. These new studies have potentially narrowed the range, but they don’t really change the best estimate (about 3K). The more extreme scenarios (both low and high) may be less likely, but we can still potentially emit enough to warm substantially. In a sense, how much we will probably warm is largely unchanged.

Additionally, a common way to quantify how we could achieve some temperature target is to present a carbon budget; the amount of CO2 we can emit if we want some chance of staying below the target. Given that these new analsyses don’t really change the ECS best estimate, the carbon budget that would give us a 50:50 chance of staying below some target should be unchanged. Often, however, the carbon budget is presented as giving us a 66% chance of staying below some temperature target.


Given that the range might now be narrower, this might suggest that the carbon budget for a 66% chance might be slightly bigger. However, it’s not only the uncertainty in climate sensivity that constrains this; there are also carbon cycle uncertainties (i.e., what fraction of our emissions will be taken up by the natural sinks). Hence, I suspect that the impact on the carbon budget framework might be small (I might be wrong about this).

Also, although I’m mostly in favour of working within a carbon budget framework (it’s a pretty straightforward metric) I do sometimes think that there might be a better way to present it (to be fair, I don’t have any good suggestions as to what this should be). A carbon budget that gives us a 66% chance of staying below some temperature target doesn’t mean that we will do so 2/3 of the time, and fail 1/3 of the time. There is only one outcome; we will either stay below the temperature target, or we will not.

If we think that there is now a bigger chance of staying below some temperature target, it’s not clear to me that we should then adjust the carbon budget. Maybe it would be better to present it as there now being a bigger chance of succeeding, than suggesting that we can now emit more while still having the same chance.

Okay, I think that’s long enough, so I’ll stop there. The latter part of this post is not as clear as I would have liked, so if anyone has other suggestions as to what we should do given these new results, feel free to make them in the comments.

Andrew Dessler’s comment is worth reading. I had forgotten that their paper was more presenting what might be a better way to constrain the energy balance, rather presenting a firm estimate for an ECS range. (see comment below – Peter)


In the interview above, Andy Dessler mentions the new Brown and Caldeira study. I interviewed Patrick Brown recently, and will be expanding on that in future vids, but for now, here is Patrick’s own video explanation of the study, which implies a higher climate sensitivity.

Washington Post:

Brown and Caldeira are far from the first to study such models in a large group, but they did so with a twist.

In the past, it has been common to combine the results of dozens of these models, and so give a range for how much the planet might warm for a given level of carbon dioxide emitted into the atmosphere. That’s the practice of the leading international climate science body, the United Nations’ Intergovernmental Panel on Climate Change.

Instead, Brown and Caldeira compared these models’ performance with recent satellite observations of the actual atmosphere and, in particular, of the balance of incoming and outgoing radiation that ultimately determines the Earth’s temperature. Then, they tried to determine which models performed better.

“We know enough about the climate system that it doesn’t necessarily make sense to throw all the models in a pool and say, we’re blind to which models might be good and which might be bad,” said Brown, a postdoc at the Carnegie Institution.

The research found the models that do the best job capturing the Earth’s actual “energy imbalance,” as the authors put it, are also the ones that simulate more warming in the planet’s future.

Additional Comment from Andrew Dessler on There’s Physics post above:

The main point of our paper is a critique of energy balance estimates of ECS based on the 20th century historical record. We think that low estimates of ECS from the 20th century historical record (i.e., Otto et al., Lewis and Curry) are not guaranteed to get the right answer, even if we knew the input variables (i.e., forcing) perfectly.

By itself, our paper does not tell us the low value are wrong — but it gives a mechanism that demonstrates they COULD be wrong. However, Kate Marvel and Kyle Armour both have analyses they talked about at AGU that shows that, indeed, the surface pattern of warming we’ve experienced over the past few decades will imply a much lower ECS than reality. My guess is that, in a year or so, no one will take those low energy-balance estimates seriously and everyone will agree ECS > 2 K.

The ECS estimate in our paper is just an example of what you can do with our framework, not something I’d every cite as a firm estimate. In fact, it’s an emergent constraint estimate, which, if you read my twitter feed, you’d know I’m not crazy about. To understand why it’s in there, you have to know something about the history of the paper. We had previously submitted a version that does not have the ECS estimate in it to another journal. The reviewers seemed confused about the utility of our revised energy balance framework and questioned what the point was. In response, we decided we had to better show what the potential uses were, so we put that short discussion of ECS in the paper. However, you’ll notice that we don’t site as numbers in the abstract or the conclusions — that’s a signal that the values are for illustration and not to be considered a firm value. We have another paper that is basically ready to submit that has a more rigorous ECS range (likely 2.4-4.5 K) based on our revised energy balance framework.


17 Responses to “The Weekend Wonk: 2017 was HOT. Climate Models Affirmed. Climate May be MORE Sensitive than Thought”

  1. Sir Charles Says:

    Meanwhile, at 1.44 times the CO2 concentration and before equilibrium we find a world which is already 1.2°C warmer than pre-industrial levels.

    • indy222 Says:

      Yes – glad you pointed that out Sir Charles. I’ve been doing the same in my talks and pdf’s, to show that the low ECS’s sure don’t make any sense for TODAY, let alone for the past. So let’s flesh out the math. Close of 2017 we were at GISS temp Lowess smoothed T=+0.95C from 1951-1980 baseline. Now add +0.254C to refer that instead to the conventional 1880-1910 avg as “pre-industrial” and that gets you to +1.204C today. But wait! Schurer, Mann et al 2017 argue that true “pre-industrial” T is 0.2C cooler than the 1880-1910 average (which, after all, is used mostly because it’s when global data quality got good), so that would give you not +1.204 but +1.404C as our T anomaly at present. But we’re only 44% of the way to 2x CO2. So let’s be conservative and assume that temp rise rate will follow CO2 rise rate so we can straight-line extrapolate and get 1.404/0.44 = +3.19C when we cross 2xCO2. And we’ve neglected the fact that dt/d(CO2) has been rising steeply during the past century. Temps didn’t really get cranking upward until after 1975 and CO2 had already gone most of the way to where it is today. Still, let’s let that sit and not panic. Now, this +3.19C is just TRANSIENT climate sensitivity, not ECS. Equilibrium would add another +0.6C or more, waiting a century or two for it to get there (not counting Hansen’s “slow feedbacks” which would continue for thousands of years. But let’s just neglect the slow feedbacks). So we arrive at a very conservative ECS based on TODAY’s climate change of +3.19+0.6 = 3.79C and that’s w/o tipping points (like Arctic Ocean ice loss albedo feedback) or any state-dependence of ECS on background climate. So ECS=+4C is quite within the conservative error bars and even Friedrich et al’ 2016’s , and the many other hotter-> higher ECS studies of the past few years value of ECS=+4.9C is not unreasonable.

      These paint a grim picture of our future.

    • grindupbaker Says:

      I seem to recall a climate scientist noting that there was some natural warming 1750-1850 AD so I think it’s best to stick with a 1880-1900 baseline for the AGW temperature anomalies. I think that’s +1.06 degrees (check it out).

      • Sir Charles Says:

        As if it matters what you think, grindupbaker. NOAA is using the 1880-1900 baseline because this is the pre-industrial period where they have the best reliable data. The IPCC 5th Assessment Report used 1850-1900 as a historical baseline, but this period includes some large volcanic eruptions and is after greenhouse gas concentrations had already started to rise. Some scientists suggest that the earlier period of 1720-1800 is a better choice for this baseline. Have a look here if you wanna dig a bit deeper =>

  2. MorinMoss Says:

    So we’re basically buggered 3.7x ways from 1850.

  3. MorinMoss Says:

    The latest idea making its way around the Denier-verse is that the time-of-day effects of clouds will balance or mitigate any & all supposed warming.
    And I don’t see them giving up on low ECS

  4. MorinMoss Says:

    “Damn, anyone else have that Flu from Hell?
    I’ve been clobbered ever since I got back from the American Geophysical Union meeting, a month ago”

    Hmm, everyone I know who’s had it are frequent flyers, mostly overseas.
    Is it being spread by cabin air?

    • Sir Charles Says:

      Interesting. I was flying in Europe before Xmas. Three days later I ended up in hospital with a severe viral chest infection on top of my COPD. But there was a little epidemic in Ireland too. Very nasty indeed. Suffocating not nice.

  5. schwadevivre Says:

    I’m wondering if there’s a case for throwing these models and figures at Alpha Go and seeing what it comes out with.

  6. Sir Charles Says:

    Rasmus Benestad just came out with an article on a recent study on ECR by Cox et al => The claim of reduced uncertainty for equilibrium climate sensitivity is premature

  7. brucehparker Says:

    we’re only 44% of the way to 2x CO2, but CO2 is not the only “factor” that contributes to warming. We’re at about 75% (490PPM) with all the GHGs, and the IPCC showed that RF for CO2 was about 78% of the total RF (1.8/2.3) in 2011. And then there are aerosols, albedo changes in the Arctic that are beyond what was expected for climate sensitivity, etc. So is TCS based on just CO2 really meaningful? Would a TCS based on total RF (if we really knew the correct value!) be a better metric?

    • greenman3610 Says:

      I think ECS takes into account feedbacks like water vapor, sea ice albedo, etc, so called “fast feedbacks” — Hansen talks about “slow feedbacks” like ice sheets and forests, deep ocean layers, that may take centuries to millennia to adjust.

    • grindupbaker Says:

      It’s 53% of +CO2 effect at present ln(406/280).

  8. brucehparker Says:

    Since ECS is a outcome of a model run and is supposed to include all “fast” natural feedbacks (which includes the Arctic ice and snow and should probably include permafrost), what seems to be missing is way for the public to “calibrate” the results of a the model runs. For example, RCP 2.6 shows a median temperature increase of about 1.8°C for CO2 at 420 PPM. Since the temperature is not changing much around 2100, that implies an ECS of about 3.5 for CO2 alone (1.8/(420-278)/278) and perhaps an ECS of 2.75 for all GHGs (where CO2 RF =2.2 and total GHG RF = 2.8 in 2100). But it is known that the models did not include emissions from a thawing permafrost and that they underestimated the Arctic Ocean ice decline. The ECS for RCP2.6 would likely have been much higher the model had been “calibrated” to more accurately reflect the models expected anthropogenic greenhouse gas emissions and what is now known about natural feedbacks.

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: