"Our greatest responsibility is to be good ancestors."

-Jonas Salk

Tuesday, November 2, 2010

Curry's Hallowe'en Piece

Well, I don't think she'll be able to save the Italian Flag stuff. But she does keep a man guessing.

I have read Judith Curry's Hallowe'en post Decision Making Under Uncertainty Part I . The criticism I can offer here, that the Obersteiner and Weitzmann critiques are rather less inclined to catastrophic outcomes and probably biased to the relatively near future, is of a very different sort than my critiques of the Italian flag, or others' critiques of other seemingly baldly elementary errors. And the statement
The net result of such a feedback loop is an overconfident assessment of the importance of greenhouse gases in future climate change.
while slyly implying an overstatement of said importance, is at least relatively unobjectionable in its literal content. We could, after all, be overconfidently underestimating greenhouse gas sensitivity as well. Now I'm reasonably convinced by James and Jules that we have the thing well bracketed, while much of the community thinks the low side is better constrained than the high side. Presumably Curry's audience does not want to hear this sort of thing.

Once she gets off trespassing directly on James' turf without mentioning James (surely a faux pas among climate bloggers) and starts talking about the near term future of modeling, she seems to be channeling me.
If we assume that CO2 sensitivity dominates any conceivable combination of natural (forced and unforced) variability, what do the simulations actually say about 21st century climate? Well, the sensitivity range for the IPCC calculations are essentially in the same range (1.5-4.5C) that was estimated in the 1979 Charney report. And the calculations show that the warming proceeds until about 2060 in a manner that is independent of the emissions scenario.

So exactly what have we learnt about possible 21st century climate from the AR4 relative to the TAR (and even relative to the 1979 Charney report) that refines our ability to set an optimal emissions target? I suspect that we are probably at the point of diminishing returns from learning much more in the next few years (e.g. AR5) from additional simulations by the large climate models of the current structural form.
The big new push in the climate modeling enterprise is for Earth Systems Models. These models are beginning to include biogeochemistry (including a carbon cycle) and ecosystem interactions. Some are proposing to incorporate human dimensions, including economics models and energy consumption models. Such models would could in principle generate their own scenarios of CO2, and so reduce the scenario uncertainty that is believed to become significant towards the end of the 21st century.

There is also a push for higher resolution global models to provide regional information, particularly on water resources. There is currently no evidence that global models can provide useful simulations on regional scales, particularly of precipitation.

Another push is for credible predictions on a time scale of decades (out 20 years in advance). This necessitates getting the natural variability correct: both the external forcing, and the decadal scale ocean oscillations. I don’t expect the models to do much of use in this regard in the short term, but a focus on the natural variability component is certainly needed.

So it seems like we are gearing up for much more model development in terms of higher resolution and adding additional complexity. Yes, we will learn more about the climate models and possibly something new about how the climate system works. But it is not clear that any of this will provide useful information for decision makers on a time scale of less than 10 years to support decision making on stabilization targets, beyond the information presented in the AR4.

In fact I'd say that 10 years is extremely optimistic given current methodologies.

Further climate modeling has relatively little foreseeable policy relevance. We have done as much as we can in terms of policy advice. It's far better than some claim, and yet limited in what it can reliably achieve in future. We know the scope and scale of the damage we are doing, but we don't know the details, and probably won't until they come to pass. It's one thing to predict a falling vase will shatter, another to predict where the pieces will end up. Policy makers want regional predictions, and the models are not likely converge regionally. After the fact, somebody can always find a study saying "See! Just like we said!" and ignore the contradictory study.

At Curry's conclusion,
The strategy (primarily model based) has provided some increased understanding and a scenario with about 3C sensitivity that is unlikely to budge much with the current modeling framework. A great deal of uncertainty exists, and emissions target policies based on such uncertain model simulations are not robust policies.
she refers to a formal definition of robustness given earlier in the piece as [a property of] "a strategy that formally considers uncertainty, whereby decision makers seek to reduce the range of possible scenarios over which the strategy performs poor".

I don't know as it's fair to suggest that strict emissions targets are not robust in that sense, though. Essentially, one can always squeeze more toothpaste out of the tube; one cannot easily put it back. That being the case, one could argue that the most stringent emissions scenarios that are consistent with economic well-being would be the most robust in this sense.

So, on careful reading, I find several points of emphasis to disagree with, which seem to be preparing to lay the groundwork for a somewhat skewed risk assessment. We'll see. To my surprise, I also find an extensive section with which I am entirely in agreement.

The main point, though, is that it is hard to imagine that this was written by the same person who has made so many elementary errors in the recent past. What we see here is a tolerable amount of counterspin, surely no worse than some of the spin out there by other mainstream scientists.

If Dr. Curry proceeds at this level of cogency, maybe she can start a meaningful conversation after all. I'd advise her to let the dreadfully posed flag business drop. I won't pursue it if she doesn't. And I'd advise her to take comparable care in future. This isn't some veiled threat, it's simply an assertion that people are on the lookout for elementary errors now, and that Curry will be called out on them.

This piece is, mysteriously, far less disappointing than several recent ones. While there are points with which I disagree it doesn't seem remotely as amateurish as recent ones have. I genuinely hope it represents the sort of thing we can expect in future.


David B. Benson said...

So I shouldn't take the recent NCAR future drought study (and also USGS study), based on the outcomes of many different climate models, seriously?

Drought under Global Warming

Michael Tobis said...

Yes, that is a pretty strong prediction. But broadly speaking we probably only get a few of those.

David B. Benson said...

Michael Tobis --- I want to be quite sure what you are stating.

I can be what% sure that
(1) US Southwest dries up,
(2) US Great Plains dries up,
(3) Mediterranian countries dry up,
(4) Central Chile dries up,
(5) Northern Argentina becomes wetter,
(6) India and China become wetter?

Its decision time for me.

Michael Tobis said...

You can't be sure and I certainly won't venture a number.

On the grand scale it looks like this:

Roughly speaking zones will migrate poleward. Equatorial edges of deserts may shrink a bit. Polar edges of deserts will expand. Probably.

Places that do get precipitation will get much more per episode because the saturation vapor pressure increases dramatically with temperature.

My personal opinion is that these trends will have a susbtantial and increasing wobble on them. But I don't know how to prove that.

Broadly I think your predictions 1-5 are quite plausible, while 6 covers too varied of an area to generalize.

Percentages? Nope.

adelady said...

OK. So we take our climate warming baby to the climate warming pediatrician. She says, given the measurements of length and weight and head size, and the growth rate, we predict baby will grow to be a six feet tall adult.

And you turn to me and say "But she can't tell us what size shoes to buy when he's three years old! Why should we believe her?"

I'm not at all convinced by all this uncertainty discussion. Even if the material presented is reasonable on the face of it, it's still singing along with a very familiar tune. The chorus certainly seems able to come in on cue.

Michael Tobis said...

Yes, Adelady, of course.

But she's within bounds this time, which is an improvement. As long as she doesn't pump falsehoods and confusion out there, we can argue like adults, and maybe even make progress.

Rattus Norvegicus said...

What she said in the head post is reasonable. I know that's weird, but it is true. Should I go and read the comments?

We need these sort of changes in modeling to get a better idea of what exactly the effects might be, but they don't add a whole lot to the bigger picture. 30 years of modeling have not changed the overall picture of climate sensitivity. I would be surprised if the AR5 simulations moved things much in any direction. How is her point "not the IPCC"?

Rattus Norvegicus said...

I went back and read her post. Ugh. Her claim that because the UNFCC cut down on research into natural variability is silly. I just couldn't get much farther, how did you? The silliness was just too great.

EliRabett said...

The problem, of course, is context. The uncertainty argument has been a weapon of denial since before tobacco became an issue, but it is also true that many people lack the context, so at what point do you conclude that someone is dissembling or whether she has independently concluded that uncertainty is large, rather than using it as a tactic. Eli submits that you look at the surrounding context, which in this case is not very encouraging.

Michael Tobis said...

Thanks, Eli.

I am basically doing Willard a favor here, and writing off all prior context. She has done me the favor of writing a not-beyond-the-pale piece. We'll see which way she goes with this (she's already dropped some not encouraging hints.) But I'm willing to adopt a tabula rasa to convince any Willardites that I am not being unfair or personal.

If she picks up the Italian flag again I'm afraid I must be merciless.

Tom said...

You owe Curry a public apology.

Michael Tobis said...

Tom, the best I can offer at present is a truce.

Let her say nothing patently substantively wrong that no professor should ever say publicly going forward. In return, I will not dwell on the ridiculous things she has said in the past.

If that isn't sufficient, if there's an apology needed, I have to be convinced that what I said was wrong in substance. She or her defenders will have to do some hard work. Impossibly hard in my present estimation, but the whole idea is that I'm wrong, isn't it?

If she (or they) would like an apology, she (or they) must open the can of worms and defend the various things she has said that others find to be undergraduate level errors in science.

I doubt she can manage it, but if she does, she gets an apology from me, and indeed a most effusive one. Indeed, I would probably change camp if she could do it.

I have an open mind, but unlike some, the emphasis is on "mind", not on "open".

There are rules of engagement. The truce is only about asking Curry to abide by them in future. She has gotten a free pass in the past. That's over.

Unfortunately, in replying to you I must violate the truce by explaining its terms, but such a clarification is in order.

I genuinely regret that someone needed to call Curry's competence into question. But it's hard to avoid seeing a pattern of blatant disregard for the fundamentals of the science in which she is supposedly an accomplished teacher and textbook author.

Tom said...

Then you're a scumbag.

Paul Daniel Ash said...


def tom(post)
if post.ispositive(tobis)
print "You are a scumbag."
print "You are a scumbag."

Michael Tobis said...

Doesn't parse. You missed a colon. Also the conditional is redundant.

Quick and dirty whitespace workaround in html is to use ellipses.

class Tom(foolery):
..def reply(post):
....if post.from(tobis):
......return("You are a scumbag")
......if "tobis" in post.text:
........return("Tobis is a scumbag")
........return("You're a scumbag" +
..............."Just like Tobis.")



Aaron said...

The IPCC, JC and MT ignore the fact that large events resulting from climate change may have what risk managers call "fat tails", such that social and economic losses from those events would be larger than what would be extrapolated from smaller events. See for example:http://www.rff.org/documents/RFF-DP-09-03.pdf

Michael Tobis said...

Aaron, it is my opinion that I was the first to make the fat tail argument, long before Weitzman.

I don't see how that has bearing on the present article.

Paul Daniel Ash said...


Somehow I'm not seeing a Tobis apology in the cards here.

adelady said...


Just wait 'til mt reads the next Italian flag piece. She's doing one soon.

(I _knew_ I should've saved her comment when I read it.)

Paul Daniel Ash said...

I'm watching the thread with mixed fascination and horror: kind of like watching a car crash in slow motion... with comments

adelady said...

Well we didn't have to wait for the Italian flag.

This last piece on the 'Feedback Loop' is just plain awful.

Paul Daniel Ash said...

Yeah, adelady, that's what I was referring to. Miz Curry, she crazy.

Anonymous said...

Judith Curry is a lot smarter than me, but even I can now see that she is a bit dim. There's a grain of truth in everything, but that latest piece on the IPCC...

I'll still be following her, but only for the psychological aspect of it. Fascinating stuff.

word verif: unctiess

Anonymous said...

What is the crux? Where are the IPCC summaries wrong?

I assume many people who actually know their stuff have beef with IPCC in some ways. Not everybody can be happy.

But I've assumed most of WG1 stuff is mostly ok.

I'm reminded of a guy at my father's workplace. He went to insult the company owner, tried all kinds of things to stir up trouble so he would get fired (he would benefit from that for legal reasons) but nobody would take it seriously and just told him to work. In the end he quit himself.

Steve Bloom said...

This last piece answers a question I've had about Curry for a while, which is whether she herself is a right-winger. Yes.

The misrepresentation of the IPCC right up front (no mention of the fact that even 20 years ago the expectation was for continued warming with attendant harms, even though at the time the initial effects were at the limit of detectability) pretty much tells us all we need to know. She's not stupid enough for that to have been an error.

Her views on the science would have more value if she had had a first-rank research career. She didn't, which leads me to suspect that the bad politics may be compounded by professional resentment.

William T said...

This might help in keeping the tone of the comments up... - http://xkcd.com/810/

Tom said...

My previous comment was far too gentle:

Michael Tobis to Richard Tol: "If you tarnish my reputation in this way in today's fraught political climate, it's not impossible that it will come back to do me considerable damage."

Michael Tobis: "We have reached a point where it is impossible to judge that Curry is in touch with the science that she is supposed to be a prominent participant in. So has she lost touch, or has she never had much scientific insight to begin with?"

Michael Tobis of Judith Curry: "I mean, could this be the stuff of some subtle neurological decay, where a formerly competent scientist starts making no more sense than the peanut gallery?"

Michael Tobis of Judith Curry: "Or alternatively, is the peer review system so shabby that a person of modest intellectual accomplishments, one who, despite years of connection to the scientific community, numerous publications and promotion to a position of responsibility, is capable of such vapid, illogical, pointlessly contentious writing."

Michael Tobis to Richard Tol:"I can't believe that you are dragging my name through the mud in the name of this ridiculous hair-splitting argument. If that's all you have, I strongly request you come out and say so."

Michael Tobis said...

Fuller's comment refers to conversation about me at Eli's, at which several people including Eli have said very kind and generous things about me, for which I'm deeply grateful.

In addition, on that thread, Fuller, along with Richard Tol, have said very harsh things about me. Fuller juxtaposes my objections to Tol's characterizing me as totalitarian with my criticisms of the intellectual quality of Curry's writing.

He posted this same item there, where it made more contextual sense. For your convenience, I'll also copy my answer.

Tom Fuller's clever sandwich is off point; a context switching trick.

I don't deny I was harsh toward Judith Curry. These are the three middle slices of his sandwich. I provided some evidence to support these strong statements, about a persistent pattern of error if the sort one would one not expect from a scientist in Curry's claims.

Richard Tol's attack on me, the bread of Tom's sandwich, was not about substance, it was about character. He accuses me of totalitarian tendencies. To make matters worse he provides next to nothing in support.

The apparent hypocrisy on my part is therefore constructed, not real.

It is not fair to compare a baseless questioning of my character and intentions to an evidence-supported questioning of Curry's competence to represent science.

Fuller has constructed a piece of propaganda. It's a smarter attack than his usual, but it's also more malign than I expect from him. I think that before he got in with the wrong crowd on climate he might have been a nice fellow. It's a shame.

Tom said...

I imagine at one point you were a decent human, too. Maybe when you were four.

dhogaza said...
This comment has been removed by a blog administrator.
Marco said...


I personally find it hilarious that Tom demands(!) you apologise to Judith Curry, but has absolutely no problem with Richard Tol's personal attacks on you. I guess that tells us what we in reality already knew: Tom's a hypocritical tone troll.

Michael Tobis said...

Although Judith Curry's Feedbacks doesn't speak well of her abilities as a bridge-builder, it doesn't really provide any ammunition for a case that her scientific understanding is weak, which is where I stuck my neck out, and where it more or less remains stuck out.

It's wrong, in my opinion, but in a let's agree-to-disagree way, not in a QED way. It pretty much rides on the idea that IPCC is an institution with an agenda, patterns of loyalty, and career paths, like an ordinary scientific institution.

It's not that, though. It is more like a process, and very few people identify with it. It's sort of like a conference. A few people work hard to make any conference work, but the majority of attendees who go to the conference have no great loyalty to the conference as institution, and their benefit for attending has little to do with the interests of those putting on the conference.

While it's not a piece I agree with or approve of it's not relevant to the question I raised. I am not asking whether her opinions are sound or not. That's a matter of meta-opinion.

I am asking whether she is actually a capable scientist. That's a matter, at least in the extreme, of fact. The Feedbacks article sheds no light on it either way.

Tom said...

Oh, now you're asking if she's a competent scientist. After stating she is not.

Tell the world which of her papers you have read.

Michael Tobis said...

I read the flag things. They were pretty shockingly bad.

I glanced at some earlier stuff. Pretty much measurement papers. Do you have a particularly excellent paper in mind? She is sole author on very few papers.

I have already backed down as far as I'm going to unless somebody makes a sensible argument out of the flags business.

But I'm not going to say anything more unless/until more undergrad level errors come up.

A lot of people hate Curry's Feedbacks piece but it doesn't cross the same line the flag pieces did.

So, Fuller, so far so good, from your point of view, I'd think. I am NOT criticizing this very harshly. Not in the same way others are. Even Bart is beside himself on this piece, or as far beside himself as he gets. So why is your beef with me on this one?

Steve Bloom said...

It's what's for dinner.

Tom said...

My beef is a political slime job that you performed. It had at the end of the day nothing to do with the Italian flag, which has nothing to do with her science. You were looking for an excuse to trash her because you don't like her views and you don't like the attention she's getting. So you hung it on the hook of a three-frame box.


Michael Tobis said...

I want people who represent themselves as scientists to be able to do quantitative reasoning as well as an average sophomore on a science track at a good university.

I think that's reasonable.

I would hold anybody to the same standard. I don't know of anything comparable anywhere.

Tom said...

Liar. You had your slime job prepped and in the can. You just used the Italian flag as an excuse because you have nothing relevant to say about her science.


Michael Tobis said...

Tom, take a pill. You're raving.

You're wrong.

Even if you were right, how would you know it? Do you have access to my machine?

King of the Road said...

I have to admit that I'm starting to look forward to Fuller's comments for sheer amusement value. It makes me feel kind of dirty though, like catching a glimpse Maury Povich.

Paul said...

I just wish he would stop signing his posts "scumbag". He's really not that bad.

Deech56 said...

FWIW, James Annan has backed up Michael (Link 1, Link 2)

As has WC.

Is that really Tom posting and not some impostor trying to make him look bad? If Michael is wrong about the science, then let's see it.

Anonymous said...

Michael, where's your image of the distribution on how you'd estimate the probabilities of climate change?

Sensitivity on x axis in one and no effect-catastrophe in another.

I remember you doing something like that some time ago, but couldn't find it.

Should help with the Curry impressions (?) that people don't talk about uncertainty.

Maybe someone could make a quick web application where people could draw their distribution figures and sign them. It would be nice to look at.

We need better communication methods than words, and I love graphs.