Saturday, December 22, 2018

Addressing stated concerns from Ray Bates around the SR1.5

Ray Bates has provided a critique of the IPCC Special Report on 1.5C promoted as a ‘paper’  (it is no such thing in that it is not a rigorously peer-reviewed piece) by the GWPF – a highly questionable think tank.

This piece has garnered some interest in the Irish press and I am quoted as critical of it and the associated views in a piece in the Irish Times. To be crystal clear: a free press is an essential component of a healthy, vibrant democracy and it would be strange for the media to completely censure views. Furthermore, journalists only have a finite amount of real estate to work with. Also, journalism is a real-time business and a tough gig. Hence here I am taking the opportunity to more clearly lay out from a scientific viewpoint why the piece being reported upon is fundamentally flawed and the IPCC SR1.5 report stands on its merits.

The societal response and policy aspects I shall leave for another day but suffice to say I am more optimistic and view national leadership on acting on climate change, if done right, as a real national opportunity for all Irish citizens.

Wilful misinterpretation of AR5 attribution findings

Ray Bates begins by claiming that the central attribution statement of AR5 is:
It is extremely likely that more than half of the observed increase in global average surface temperature from 1951 to 2010 was caused by the anthropogenic increase in greenhouse gas concentrations and other anthropogenic forcings together.
And that the statement did not necessarily assert that more than half of the warming is due to human influences. He goes on to contend that SR1.5 departs from AR5 in contending that all the warming is due to human influence. 

Why is this a wilful misinterpretation of AR5? Well, the complete statement rather than a partial lift is:
It is extremely likely that more than half of the observed increase in global average surface temperature from 1951 to 2010 was caused by the anthropogenic increase in greenhouse gas concentrations and other anthropogenic forcings together. The best estimate of the human-induced contribution to warming is similar to the observed warming over this period. 
Spot the difference? It’s this:
The best estimate of the human-induced contribution to warming is similar to the observed warming over this period. 
I see no other way to describe this presentation other than wilful misinterpretation. Don’t believe me?  p17. Section D.3 first bullet. Want more evidence? Try to reconcile only half the change being explained by anthropogenic influences in Figure SPM 6:
Figure SPM6 from IPCC AR5 WG1 report. Changes for simulations with human influences (pink) and without (blue) for a range of global and continental and ocean / cryosphere indicators. Human influences explain effectively all the change seen since the early 20th Century.

Or to reconcile the statement with Figure 10.5 from the attribution chapter in the report:
Figure 10.5 showing attributable contributions to the observed change (black bar) from Grrenhouse gases, all anthropogenic forcings, non-greenhouse gas anthropogenic forcings, natural forcings and internal variability. The best estimate is that anthropogenic contribution to warming slightly exceeds that observed (balanced by variability / natural forcings). It is in no way only half the warming.

So, does SR1.5 start from a distinct premise from the AR5 conclusions around attributable warming contribution from us humans? That’ll be a resounding no.

Appropriate use of temperature period and series

Ray Bates goes on in his piece to imply nefarious intent behind the IPCC considering only the post-1960 series of GMST in their SPM figure. The choice is reasonable because the attribution statement in AR5 (and the prior figure) pertained to post-1950 changes. Furthermore, in the underlying chapter 1 Figure 1.2  the full series from 1850 is shown. The SR1.5 is hardly ignoring the early period as implied is it? 

Ray Bates then goes on to show a satellite dataset that shows less warming than the surface. It is true that the UAH dataset does, indeed, show less warming than the surface. What’s the issue? Well, satellite data are no gold standard. You put a delicate instrument on top of a controlled explosion to send it into space then expose it to vagaries in space weather. Unlike the surface there are not hundreds of instruments there is one. It can drift in the sampling and it can degrade in performance over time. What happens if another group also go to the trouble of creating a product from the same instruments in space? Well, they find greater warming than at the surface. I’m not going to argue one of these is right and the other wrong. I’m just pointing out that satellites aint no gold standard. Indeed, AR5 concluded that the surface was robust whereas the tropospheric changes while certain of the sign (warming) were uncertain in magnitude. As I lead that portion of the AR5 assessment I am uniquely qualified to make this statement.

Ocean red herrings

Ray Bates goes on to throw in a couple of red herrings on the Oceans for good measure. He states that an emerging signal of warming of land faster than oceans in the observed record is problematic. Well, no, its actually what we expect as an emerging signal of climate change and it will only get worse. As SR1.5 points out much of the global land domain already experiences seasonally or annually changes exceeding 1.5C. 

Don’t believe me? Here is Figure 12.11 of the SPM of AR5. Note how the land warms more than the oceans in all cases. Note also how our choices matter in determining future generations climate.
Figure 12.11 from AR5. Land robustly warms more than oceans. I sure don’t want my descendants inheriting that bottom right world. Can we please make efforts to make the top right reality?

Ray Bates then goes on to contend that CERA-20C ocean heat content in the 0-300m horizon casts doubt on the significance of ocean heat content changes. The changes in the early period of that record are clearly spurious. Variations of the degree shown would lead to changes in sea level that simply are not present in the tide gauge records around the world. The abstract of the paper makes clear that the result is experimental: 
These preliminary results are considered of interest for the ongoing community efforts focusing on coupled data assimilations.
Why would IPCC be expected to rely upon an experimental data assimilation system analysis? Well, exactly.

Climate model tuning

The section on climate model tuning is very confused indeed. I was involved, when at the Met Office, in the model tuning and evaluation exercises. I saw lots of effort concerned around getting critical processes such as El Nino, mid-latitude storm tracks and monsoon variations adequately as well as the mean state. I never saw an effort to tune the model to reproduce the observed trends or to determine some state in 2100 in the projections. I also saw significant efforts to run perturbed ensembles and these have been done even more so in the citizen science www.climateprediction.net experiments. To claim that the model tuning is an issue that is unexplored or a major issue is incorrect.

Equilibrium Climate Sensitivity

The section on ECS is, to put it politely, selective. Fortunately I don’t have to waste many words here as there has been a recent major review piece written by Reto Knutti and colleagues - the figures are not behind the paywall and any number of these give a much more holistic assessment as to the state of our knowledge on the topic. The bottom line is that contrary to the assertion of Ray Bates if anything the lower bound on ECS has increased since AR5. The good news here is there is a whole chapter devoted to it and related matters in AR6.

To conclude

Ray has had a long and distinguished career. But that career has been in atmospheric dynamics and not climate. Yes, both are to do with the atmosphere, but when your toilet is backing up you call the plumber and not the electrician. In the same way when looking for guidance on climate change it is advisable to listen to the climate scientists of which there are many thousands the vast majority of whom (and I mean vast) concur with the broad findings of the IPCC and various national assessments and national academies that climate change is real, its due to us, and that our choices now are of critical import. Equally, if you want to discuss the intricacies of atmospheric dynamics please don’t come knocking at my door!

The analysis of Ray Bates is not a peer reviewed paper and finding substantive flaws in it is really not that hard. This has taken me all of an hour of an evening. Compare and contrast to the rigor of the IPCC assessment process with multi-author teams, multiple drafting meetings, substantive rounds of in-depth review, and final sign off on the summary word by word with the governments of the world. Only the most robust findings survive such a process. It is to borrow a colloquialism like contrasting night and day. I know who I’m going to pin my choices as to the future for my kids on. It’s the guys saying this:







7 comments:

David Young said...

I think this article rather misstates the concern with model tuning. The problem here is that credible ranges of parameters for cloud and convection processes give a wide range of model ECS. Thus, the ECS is not really constrained well by these models.

Peter Thorne said...

David,

In response to your comment I have carefully re-read the segment in the original. I do not disagree that sub-grid scale parameterisations are important, nor did I imply this. It is just that my experience is that the tuning effort is at a stage before historical or projection simulations are even started. The tuning is all about getting climatology and important processes. By the time congruence with historical changes or projections are apparent the version is operationally producing the requisite runs and totally locked down.

So, congruence with 20th Century change or falling into some range of ECS, in my experience has no influence in decision making contrary to what is implied by Ray Bates. I had the good fortune to attend as an observational expert an IPCC meeting on model validation last year where considerable effort was expanded on the matter and there it was clear that the many modelling groups present do not explicitly tune to the 20th Century observed changes or a given ECS.

This may not mean it is universally true. But, from direct experience and honest discussions with colleagues I simply do not see evidence that tuning is to anything other than mean state and variability. It would be perverse not to try to ensure models had a reasonable present-day climatology and captured important processes.

Pieter Zijlstra said...

Perfect! Cristal clear.

Victor Venema said...

I would like to politely disagree with your first point. With billions of people there will always be numerous fools on any topic. Especially if the dark money donors of "think" tanks like the GWPF are willing to pay for it.

When I studied physics the professors showed us hilarious letters for alternative theories that violated basic physical laws and it could be fun and sometimes even hard to work out what was wrong. These people never got into the newspapers.

The damage to society is done by journalists picking up the nonsense. I fully realise they have a hard job and that the media system is going through a tough transition, but they are the ones who make nonsense into a problem for humanity. (Just for the record. No need to reply, I understand that you would like to keep good relations with the local press.)

rembrand said...

https://www.thegwpf.org/bates-reply-to-thorne/
It's your turn again :)

Peter Thorne said...

Rembrand,

If Ray Bates is unable to admit that two sentences in a single bullet of the AR5 SPM were intended to be read and interpreted together I think that says all one needs to know to be honest. If an author places two sentences together in a single bullet point it is for good reason. If it is signed off in the final plenary it has a sound basis. The rest is simply rehashed long-debunked arguments.

Lars Wilmer Thomas said...

If you please, though a year and a half since, it would be interesting to get the referrals to the long-debunked arguments that are rehashed
:)

Post a Comment

Please note all comments are moderated. Comments containing profanities, unwarranted accusations, deemed off-topic etc. shall be deleted. There may be a substantial delay between posting a comment and its acceptance owing to moderator availability.