Tuesday, December 15, 2009

Don't Overstate the Climategate Case

UPDATE: below

The right-o-sphere is buzzing about Willis Eschenbach's guest post on Watts Up With That? looking at Australian climate records. The piece purports to compare temperature data as reported by the IPCC with the unadjusted records first in northern Australia (defined by the IPCC (SM.9-9) as land measurements between 110E to 155E, 30S to 11S) then, second, at a particular station in that region, Darwin airport.

Ed Morrissey and others see it as Climategate's "smoking gun", establishing that warming stems from shadowy "manipulation" of the underlying temperature data. This is a point I've made before -- that the climate models are all wrong, if not phony -- particularly about measurements from the U.S. and New Zealand. But on initial inspection, I don't think Eschenbach makes an air-tight (ha!) case.

The first part of the analysis, covering the northern Australia region, is the most troublesome. Eschenbach accurately reproduces the IPCC's "global mean temperature changes (°C) from 1906 to 2005" in Northern Australia, from its 2007 Fourth Assessment Report:

source: IPCC Fourth Report, WG1 at 695

Next, Eschenbach plots the raw data from the Global Historical Climate Network (GHCN), i.e., the average temperature anomaly actually recorded at all 222 stations in the same geographic region:

source: Eschenbach Figure 4

Eschenbach indignantly concludes that the latter chart "looks nothing like the UN IPCC data, which came from the CRU, which was based on the GHCN data." But in truth, the broad "Northern Australia" anomaly reported by IPCC and Eschenbach's raw plot in figure 4 show about the same temperature anomaly -- a half of a degree C. Initially, this isn't much (unless it's per decade), which could be an argument in itself; but Eschenbach doesn't make that point. More importantly, much of the difference between the charts flows from the IPCC's decision to start its series in 1906, which is the lowest year of the raw data. Eschenbach starts his in 1880. Eyeballing Eschenbach's plot from 1906 on makes the two charts appear far more similar.

So Eschenbach's analysis mostly demonstrates the significance of where one starts plotting trend lines, a deception often practiced by warming alarmists. That itself is a bias -- cherry picking -- but Eschenbach doesn't address it. And it falls short of showing that the subsequent adjustments to the data are a smoking gun.

Eschenbach's analysis of the adjustments made to temperatures recorded at Darwin airport is, in some ways, more probative. His two plots compare the averaged raw and adjusted temperature anomaly numbers at that site:

source: Eschenbach Figure 7

source: Eschenbach Figure 8

I haven't double-checked Eschenbach's calculations, which I presume are correct. If so, the warming appears to derive solely from post-recordatation manipulation of the data. And, to the extent that typical "adjustments" are applied to historical records to lower them -- as apparently done in New Zealand -- or delete them -- as apparently done in Orland, California -- that's particularly suspicious (and potentially illegal under various freedom of information laws). Still, Darwin airport is just one station. Meaning, a logical response from warming zealots is that this, too, is cherry picking. Ken Hall suggests as much in comments on WUWT? (4:50am on December 8th).

Don't mistake my argument. I think it far from certain that warming is caused by man-generated greenhouse gas emissions, that the urban heat island effect accounts for much of the asserted increase, and, in any event, that warming has plateaued lately -- as even the IPCC chairman admits. And the media's behavior -- the New York Times, Washington Post (Ebell letter to editor) and even Google -- has been shameless.

Given the manipulation of the data and peer review by supposed scientists, the IPCC and other warming alarmist data must be fact checked before being the basis for treaties or EPA regulation. If they rush ahead regardless, that tends to confirm that enviros' real agenda is socialism, not science.

Yet, that doesn't prove a widespread conspiracy faking the data. As Tiger Hawk says:
The more partisan skeptics who have argued the conspiracy theory, or at least alluded to it, are setting themselves up for the obvious response, which is that true conspiracies involving thousands of people are virtually impossible to organize, sustain, and cover up. Megan McArdle describes the much more probable case, which is that the community of climate scientists are practicing a subtle sort of collegiality bias, in which nobody wants to find large errors in the reasoning of their colleagues.
As Richard Fernandez concludes, the likely culprit is a form of confirmation bias, where supposed scientists -- like this one -- only seek answers conforming to their pre-conceived warming theory.

That's bad enough; indeed, it makes alarmism dubious. But, without more evidence, conspiracy assertions make warming skeptics vulnerable to accusations of the opposite confirmation bias. So turn down the hype and concentrate on getting a consensus to let the numbers speak for themselves. Put differently, insist everyone share the data--and eat it raw.


They're not taking my advice; indeed, the CRU may now be hiding data formerly available on the web. Prompting a "litigation hold" notice (prohibiting document destruction) to United States Department of Energy employees regarding CRU data.

(via reader OBloodyHell, Wolf Howling, twice, Berman Post)


OBloodyHell said...

As I've noted in private --

I'd concur with what you say, but the solution for the AGW crowd to directly reject it is to simply pull up a half-dozen more records and show that it's not so -- then challenge the opposition to match that six-to-one, with the promise of counter-matching THAT six-to-one if they can do so... in short, to prove it's wrong by showing the data.

But the real fact is, the AGW crowd DON'T want the actual record examined closely, and I say that's because too many records will show exactly what the Darwin data shows -- specious methodology and practices -- which is why they consistently refuse to provide data, to the point of even destroying it when confronted by a legal demand to do so.

And here's YET another "cherry picked" example of cherry picking:
Frigid Folly: UHI, siting issues, and adjustments in Antarctic GHCN data -- "Antarctic is warming" claim based on data from one single station. A station in an atypical location surrounded by one of the denser human habitats IN the Antarctic. Can you say "Urban Heat Island Effect"?

Yeah, all by itself, it's just a cherry...
At what point do you notice that the cherry's been et by a trout and he's in the milk?

"Hey, look at all those OTHER milk buckets!! No trout there!!"


AND THAT is the functional response to that "cherry picking" counter claim:

"Fine. Share your data so we can discuss it ALL like real scientists, rather than having to independently reconstruct it from scratch for each location. WE aren't getting paid for it."

They don't get to hold us to rules they won't follow, when those rules are in clear violation of the scientific process. We have to stop ceding ground to them on that.

So the reply to the claim of cherry-picking in response is simple -- either the above "quotation" or this one:
"We'll cease cherry-picking, when YOU cease cherry picking, and share your data so we can ALL discuss what it actually says"

I concur with you in principle, but in practice it doesn't work -- the people doing this reconstruction shouldn't have to be doing it blind and attempting to guess what the heck it is that the Goebbels Warming crowd has pulled to get their numbers. Prior to publication, yeah, you can keep things "secret" -- once you publish, all methods and techniques are revealed to provide transparency and reproducibility.

That's how Science WORKS. Not by manufactured (aka "manipulated"!) consensus. Science is not Democracy. It's not even a Republic, even though we're well on the way to not keeping that.

Oh, and, while you're at it, here's a bushelful of cherries to chew on:

815 new snowfall records, 304 low temperature, and 403 lowest max temperature records were set this week. And, for integrity, as you can see if you look at the map, some "high" records (36/51) set down in the Deep South. More here.

Hey, "we cherry picked the week"... right?

Trout. Milk. Cherry. 'Nuff said.


Anonymous said...

You really don't need ANY science to uncover this scam. One needs only recognize the fallacious arguments, inconsistency of purpose, rejection of obvious remedies (nuclear), and other tells.

I could tell years ago that this was too fishy to be true even when I was predisposed to endorse it.