Friday, May 23, 2014

A Statistical Analysis of Tol's Demon

The bunnies have been asking for some hints about the dance Dickie Tol is attempting.  Ok, as Eli wrote over at Retraction Watch a couple of days ago (May 21)

What Tol is saying is that if one includes his outlier even with corrections the curve is rather banana like with a net positive benefit at modest warming. As Frank Ackerman has shown this is a result of the curious way in which Tol’s FUND model calculates agricultural effects, the positive benefit is large and completely different from every other study.
If you do not include the Tol and Anthoff result, then you pretty much get a simple aX2 fit which is negative everywhere, rather than what Tol shows (see And Then There is Physics for a long discussion of this). If you force the fit using Tol’s data, then to encompass the other results the curve has to descend more rapidly past 3 C of warming taking into account the other models including the questionable Tol mode. This means that Tol’s fit predicts more damage at higher warming than the aX2 fit however, as the IPCC AR5 points out, Integrated Assessment models (IAMs) past 3C warming are economic fiction, because the damage would be so great and from so many directions not included in the IAMs that the models lose validity.
Given that, Tol's statement that
The assessment of the impacts of profound climate change has been revised: We are now less pessimistic than we used to be.
is just parlor games.  Neither old nor new are valid much past 3 C .  But what does Eli know, he's only a bunny. Perhaps the good Professor Tol will accept the analysis that Andrew Gelman posted today


12 comments:

Bernard J. said...

It's not just about a few borked signs, although that's bad enough. It's also about missing factors, to which I think Ol' Dick might refer as "externalities", in traditional dismal fashion...

As I, Eli, and others have noted, once the water's under the bridge the curve is going to look rather different by around 3°. What shape exactly may be an idle post hoc curiosity for the bowl of petunias follwing the whale, but I would suggest that the absolute value of second derivative of the curve will demonstrate a rapid increase in magnitude.

You can take that to the bank...

William Connolley said...

Tol makes Retraction Watch (again) http://retractionwatch.com/2014/05/24/weekend-reads-scientific-bullying-again-science-journalism-looks-inward/, who link to http://www.washingtonpost.com/blogs/monkey-cage/wp/2015/05/23/the-gremlins-did-it-iffy-curve-fit-drives-strong-policy-conclusions/

Lots of nice quotes to mine there, and some good thoughts too.

Thomas Lee Elifritz said...

Quicksand indeed. It's weird how he just can't seem to own up to this. Or maybe not. I guess we'll just have to consult with a psychologist.

Anonymous said...

Of course, what we can observe is that corresponding with the warming to date, humans have never enjoyed a lower death rate, greater income, nutrition, education, or general well being.

Have a nice day.

Steve Bloom said...

Much like heroin, the fossil fuel high is great while it lasts, anon. The earlier stages of an OD feel best of all, I'm told.

BBD said...

Anon.

I can't match Steve Bloom's eloquence, but a mundane paraphrase would be: current conditions are uninformative about future conditions under a continued increase in forcing.

False equivalence is a logical fallacy.

Lowballing TCR and the projected economic impacts of warming is a nifty tactic, but it is excessively reliant on outliers at the bottom end of sensitivity and economic analyses.

Shaky ground all round.

Anonymous said...

Trying to "extrapolate" the curve based on just a few points from a few economic models with all sorts of built in assumptions (that are not even constrained by physical laws) is a fool's errand.

This really is just so much mathturbation.

Tol, Dick indeed.

Thomas Lee Elifritz said...

Right, but the problem here is that the media took it and ran with it and that contributed to the delayed action by something like five years now, and get this, he will not acknowledge that even with the sparse data, the curve is below the axis, the result is different and he will not fully retract it. Is there a psychologist in the house, this gentleman needs immediate help! This is what post peer review by blog dissection is all about.

Anonymous said...

Tol's reaction is pretty typical among academic types.

They don't like to admit they are wrong (particularly in a public forum) and will go to great hem-and haw-lengths to make it seem they are not wrong, or at least not as wrong as they really are.

Or, on their own blogs, when you point out their mistakes, if they think they can get away with it, they will simply ignore the criticism (crickets) or simply wait til one of their rabid apologists chimes in to defend their stupidity so they don't have to.

dbostrom said...

Anon: "Have a nice day."

There's something very funny about a rhetorical flourish supposedly punctuating a blistering riposte, when the whole affair is limp from end to end.

climatehawk1 said...

Not much eloquence here, either, but another analogy:

Today is a really, really nice day here, just lovely. I've been told that there is a massive storm front approaching, and perhaps some tornadoes, but I can't remember when I've seen such a nice day, so I'm pretty sure it's no problem.

Bernard J. said...

Climatehawk1.

It's even worse than your analogy indicates - permit me to tweek it:

"Today is a really, really nice day here, just lovely, because I've spent all of my kids' trust fund on loose women and fast cars. I've been told that there is a massive storm front approaching, and perhaps some tornadoes, but I can't remember when I've seen such a nice day, so I'm pretty sure it's no problem because having loose women and fast cars doesn't change the climate... does it?"

I'm not surprised that the Anonymous didn't use a nym - it's embarrassing indeed to make such a facile leap of fallacious logic using such a clumsy85838885 475 false equivalency. So embarrassing indeed that even putting a pseudonym to it would contaminate his (it's almost alwasy a 'he') karma.