Friday, March 25, 2016

On Hansen et al

By Peter Thorne

As I am getting dribs and drabs of queries from several angles on the Hansen et al paper that appeared this week in final form I thought I'd write a post (possibly against my better judgement). I was an invited reviewer and my initial and second reviews are available online. I stand by those. The whole paper has been a highly unusual experience to be part of the review process of, and not because its in the EGU Open Access journals. I have reviewed many times for EGU journals articles that have proceeded by a more traditional route. In this blog post I shall try to reflect upon the process, the potential issues that remain, and its reception. Much more can be found in my reviews which will likely help if you have issues with falling to sleep at night.

Before getting into the nitty gritty let me be clear on several points:
  • There is value in exploring possibilities of future behaviour of the climate system. We have not observed it long enough or well enough, and we do not (yet) have powerful enough and complete enough models, to absolutely rule out 'nasty surprises'. 

  • We are undertaking an unintended geoengineering project that is pushing the climate system away from what has been a remarkably (in geological terms) stable climate that has allowed agricultural, then industrial revolutions and modern society to develop. Married to other ongoing large perturbations to aspects of the Earth System such as ecosystem disturbance on an unprecedented scale we truly are entering uncharted territory. From a purely scientific perspective, the precautionary principle of reducing our environmental footprint should be pursued strongly.

  • It is the absolute right of the journal and its editors to publish any piece using their best judgement upon completion of a proper peer review process. I would strongly support any editor and any journal in the right to publish what they view as fit so long as the paper has undergone rigorous peer-review.

Peer-review process observations

The whole process was as if we had fallen through Alice's looking glass. The paper was trumpeted at a major news conference several days before the discussion paper was even published online and available for review. Many media outlets, completely incorrectly, reported this as a new paper. Discussion papers on EGU journal sites are not peer reviewed articles and do not therefore constitute scientific papers. Coincidentally, since then EGU now no longer typeset the discussion papers so that it is now more obvious whether the paper is a submitted manuscript or the final real deal. I have no idea whether these issues are linked.

This deliberate publicity surrounding a discussion paper (which to my knowledge is unique) led to unprecedented interest in the paper. By the time the comment period was closed there were over three times as many reviews as to the next most commented discussion paper in the journal's history. This included many off-topic comments including a long thread on the existence of the greenhouse effect. Ironically, this was one of the better responded to comments by the authors (more later ...).

Amongst the greenhouse effect deniers and other off-topic comments were unsolicited reviews from a large number of very well respected scientists expert in many fields pertinent to the paper including several colleagues who were (Coordinating) Lead Authors in the Fifth Assessment Report of IPCC or who have contributed to major works such as the annual state of the climate series. These reviews highlighted very many salient issues that the official reviewers failed to spot, and hence added substantial value.

In my view the responses from the paper author team to very many of the comments they received were inappropriate. Scientific peer review has a set of norms that you respond to the issues raised in a calm and measured manner including point-by-point responses that detail whether changes were made, what these were, and why. Instead, the authors chose to respond in many cases by writing discursive policy pieces that were too often non-responsive and often verged on playing the man and not the ball.

The reviewers (invited or otherwise) all donated the precious gifts of their time and their expertise to the peer review process. Their contributions both required and deserved scientifically substantive responses. Sadly, in many cases this is not what they received. The most egrerious example is the Drijfhout et al comment and its response. The reader should consider whether the response is polite and addresses the substantive points raised by the reviewers in a measured manner. In my judgement it does not. This is but the most obvious example of a systemic issue in how the reviews were dealt with by the authors during the public review segment of the process.

It is beholden upon senior members of the community to set an exemplar of expected behaviour. They are role models and they rightly or wrongly set or modify expectations of cultural norms, be that in climate science or elsewhere. My view is that the authors treated many of the reviews as a nuisance and did not provide the response that was justified to them that allowed the reviewers to fully understand how each of their review comments was dealt with. This included the public version response to my own invited review. It was not the behviour I would expect from such senior colleagues.

Science and content

Peer-review is a necessary but grossly insufficient condition for eventual acceptance of a new hypothesis as a scientific theory. Its a very weak and imperfect filter. So, publication is the first step on a long road to eventual either acceptance or rejection of the hypothesis. I'll briefly highlight here some of the open issues as I see it that lead me personally to put a very low prior on the work being correct. There are many more open issues detailed in mine and others' reviews at http://www.atmos-chem-phys.net/16/3761/2016/acp-16-3761-2016-discussion.html

The sea-level rise is prescribed and not predicted

The authors prescribe an extreme freshwater hosing (pumping of fresh water) into the sub-polar oceans. Hence the extreme short-term sea-level rise is not a prediction arising from the model at all, and assertions to the contrary are patently false. The authors basically give up on ice-sheet processes in climate models or even ice sheet models and decide to play what-if with some, in my view, poorly physically justified assumptions.

It is important to stress that there is therefore no rigourosly physically justified mecanism or basis underlying the posited multi-metre sea-level rise. They basically assume you start off adding one olympic swimming pool today, two tomorrow, four the next day etc. then when they determine they have added enough swimming pools of water they switch it off.


This is meant to approximate large-scale ice-sheet disintegration of Antarctic and Greenland ice sheets that are posited to add a huge volume of water on a short multi-decadal period. Such pulses have occured in the past, but primarily to my knowledge when transitioning from glacial to interglacial and presumably associated with the disintegration of the much more unstable Laurentide and Eurasian ice sheets. Antarctica and Greenland have remained ice caps whilst these other giant ice sheets have come and gone. There is therefore evidence the present ice sheets are much more stable. You'd have to look to colleagues who provided comments for further clarification though.

That said, West Antarctic is unstable, and we almost certainly have passed a tipping point that will see eventual c.8m rise but the scientific literature generally suggests this shall be a multi-century process. East Antarcic and Greenland have a few retrograde bedding outlet glaciers that with water intrusion could surge in the short term. But they are a bit like funnels draining huge areas and therefore the potential rate of addition will quickly become constricted as a result. A short term acceleration would not likely be a precursor to continued acceleration posited by the authors based upon current understanding of ice sheet dynamics. Reviewers more expert than I pointed these issues (and the lack of perceived realism) out but to my judgement were never adequately addressed.

I therefore see the claimed multi-metre sea-level rise as not a prediction and their characterisation as such is at best unfortunate. The fresh water injection is simply a prescribed forcing of the model system. The authors have no robust basis and ample comments on the record associated with the review calling into considerable doubt the verity of the underlying assumptions. So, the central headline of multi-metre sea-level is probably best considered an assumption instead of a prediction.

The hosing experiments also have an issue that the water is injected into the Oceans at -15C, which you can do in a model. I could inject water at absolute zero if I wished (although the model might, admittedly, crash). But in the real-world the water will be added at or close to -1.8C or higher. It is an open question whether this step is valid to approximate ice calving effect, and how it impacts the model predictions (more, again, on that later).

The Boulder deposit is at least as plausibly tsunami-mediated

One of the most eye-catching aspects of the paper was the apparently storm tossed relic boulders atop a current day cliff that was used to support an assertion of increased storminess in the Eemian. There is little doubting the boulder deposit was wave mediated and whilst there is some uncertainty inherent in dating it was at or around the late Eemian maximum. What is highly uncertain is that the waves were storm driven. These boulders were likely well in excess of 1,000 tonnes and i) dislodged from the sea-floor then ii) raised 20 metres. The amount of energy required to do this is tremendous. There are likely only a few places in the world where meteorologically driven waves get large enough to do this and in the modern climate the location is definitively not one of these. Wave power is a function of windspeed, fetch and duration. It is only really in mid-latitude locations, such as the west coast of Ireland, where such waves can plausibly occur. In the tropical locations although wind speed can be considerably higher fetch and duration are both limited. A local point tsunami which may have also led to local uplift (hence reducing the work required) as mentioned by expert reviewers is at least as likely.

The largest impacts go almost unmentioned


From an impacts perspective the biggest impact is not the sea-level rise at all. The model runs show very large perturbations to, in particular, northern hemisphere climate in the mid-21st Century. Basically many areas in Europe in particular enter a period at least as cold if not colder than the little ice age. The associated rainfall changes are equally as impressive. This would lead to large-scale challenges around provision of food, services etc. for global society and large-scale disruption of ecosystems. It is also entirely opposite to the direction of climate change that policy makers are currently planning for on this timescale. If (and in case you haven't twigged this yet that is an extremely big if in my expert judgement) the paper hypothesis were to eventuate, we'd have made massively incorrect adaptation decisions and investments and it would have large-scale implications for society.

Change of title

The original title was:
"Ice melt, sea level rise and superstorms: evidence from paleoclimate data, climate modeling, and modern observations that 2 ◦C global warming is highly dangerous"

and became:
"Ice melt, sea level rise and superstorms: evidence from paleoclimate data, climate modeling, and modern observations that 2 ◦C global warming could be dangerous"

This change has been variously discussed e.g. here amongst many others. I see no way that the original title could remain given:
  • The speculative nature of the freshwater injection (amount, rate and method)
  • Open questions over interpretation of the supporting observational and in particular palaeo-evidence and its interpretation
  • The non-deterministic nature of the climate system
  • The fact that the findings contravene accepted science based upon many decades of research and publications by many respected and established groups and individuals
The title was (and has to be) a reflection upon the paper contents and not on whether 2C does or does not reflect dangerous interference with the climate system. Those who willingly or unwittingly conflate these two issues are falling into a very obvious and dangerous (no pun intended) logic trap. The title should reflect the paper and not be a position statement upon whether some nominal departure from pre-industrial is safe or not. The two are completely seperate issues. To answer the second requires a holistic assessment of the totality of evidence for impacts of various thresholds on subcomponents of the Earth System. For that you want to wade through IPCC WG2 (book a few days out to do so) and not a single paper! The editor was entirely correct in that context to follow reviewer advice and insist on the change. To not do so would have been to discount the points raised above and have been extremely foolhardy.

Reception 

Here I shall just concentrate, briefly, upon the reception to the revised paper published this week. There is an interesting twitter thread here (see also here) on how two journalists came to polar opposite views. Obviously, I agree with Seth Borenstein's position here. Indeed, possibly the worst piece of journalism on the paper that dispensed with even any pretence of scientific balance was the piece at Slate. If you have got this far firstly congratulations and secondly you should be well sensitized to the fact that peer review is an imperfect filter. Being published does not make it part of a scientific 'canon' and nor should science be conflated with religion in such a manner (actually, at all).

Elsewhere I've seen pieces in the Washington Post and several other places that provide a degree of balance. Climatehome ran a nice piece after a phone interview. The Guardian has also decided to dispense pretty much entirely with balance. Twitter has been somewhat 'fun'.

Some of the articles, discussions and discourse have shown up a small number of the people supporting the paper to be no better in terms of behaviour, conspiracy theorising or reasoned logic than those on the opposite side of the aisle. Science works by a slow, deliberative methodical process. Single papers do not overturn received wisdom. Nor do blog articles (this one included) or comments below the line in various fora. A lot more light and a lot less heat is needed.

And people wonder why many of my colleagues are in general reticent to wade into the wider public debate ...

12 comments:

Rocky Rex said...

Thanks for this measured coverage of this very unusual sequence of events.

karl schwartz said...

Thank you for your thoughtful piece and reminder of how science ought to be done, published, and how hypothesis should be discussed and verified with data.

Agree that the super-storm boulder explanation is highly speculative. However the severe typhoons of 2014 and 15 (and Sandy) already demonstrate the danger we face today. Clearly the warming oceans are driving extreme weather events. My lay perspective is that the Hansen hypothesis deserves full attention and rational discussion ... due to the catastrophic impacts that would occur if his hypothesis is correct ... and assuming that it is a science-based hypothesis -- one that is plausible. I agree with you 100% about the title change for the paper and that the uncertainty about his hypothesis should be included in any reporting of it.

EliRabett said...

Eli commented, or to be more accurate provided the several reviewers comments on the title. It was . . . interesting

As to the intemperate responses, there has never been a good explanation and Hansen in his response to the second stage review elided over it, but the Bunny would suggest that he thought the time for responding to the reviews had passed and the paper was dead, that, and of course, a couple of fine whiskeys

http://rabett.blogspot.com/2016/03/the-title-is-neither-wine-nor-sausage.html

Victor Venema said...

Interesting. The title of the Slate piece according to the html is "James Hansen sea level rise climate warning passes peer-review" and also the URL has that title "james_hansen_sea_level_rise_climate_warning_passes_peer_review.html".

Apparently only later an editor changed "passes peer review" to the erroneous title: "Is Now Part of the Scientific Canon". Weird that the author later defended the Canon-title on twitter.

Thomas Peterson said...

Authors frequently try to guess who anonymous reviewers are. Reading this blog post reminded me of an old technique of mine: When I didn't want the authors to know who I was, I wrote whilst once or twice in the review. I hope I haven't made authors mad at Peter. The other key feature is how science in general owes a large debt of gratitude to selfless peer-reviewers, anonymous or not. Peter obviously put a lot of time, energy, care and consideration into his review - time that could have been spent doing his own science. For that we thank you.

Åsmund Ervik said...

I'm curious about the -18 C liquid water detail: if my numbers are correct, the heat that must be provided to melt 1 kg of ice is 10x the heat that must first be provided to heat 1 kg of ice from -18 C to -2 C. Has there been any explanation of why Hansen et al. claim it is reasonable to neglect the large effect and include the small one? Alternatively, is it known what happens to the model result if this water is injected at -2 C instead?

PeterThorne said...

Asmund,

I'm afraid this is one of many aspects that were still, to me at least, unclear throughout the process. I and other reviewers suggested several such experiments or rerunning with a completely different model to build confidence in results. That didn't happen which is fair enough as its unreasonable to expect all suggestions to be enacted. This is one of many questions and avenues for further investigation which I feel still remain.

Victor Venema said...

You are right that in the end the decision is up to the editor and the role of the reviewer is to advice the editor. Still I find it very demotivating when significant problems pointed out be the reviewers are not fixed in the final article. Does not make me more eager to volunteer to do reviews.

Unknown said...

Alice 'stepped' Through The Looking Glass.
Whereas she 'fell' down the rabbit hole in Alice in Wonderland.

I mention this in the interests of scholarly accuracy. :-)

Ps - if A in W had been written 100 years later it may have been given the title A in Wabbitholeland.

Brandon R. Gates said...

Dr. Thorne,

I came to this post by way of ATTP's blog. Just wanted to leave you a note of appreciation for writing it. You have helped me better understand why I might not want to rely on this paper to substantiate my own pro-mitigation position and arguments.

Best regards.

Michael Sweet said...

Dear Dr Thorne,

Thank you for the time you put into your comments and the review. For this paper, it seems to me that more thoughtful discussions on the net contribute to the undersatnding that interested laymen (like myself) get. I came here from a link at Skeptical Science.

Anonymous said...

I understand the concern and issues raised. Nonetheless, I think it is important to continue the public debate, even if it leads to disagreements like this. In my opinion, publicly airing these kinds of disagreements is good, because the public gets a better sense of the enormity of the uncertainty involved. Knowing what we don't know is also valuable knowledge - for all of us.

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.