The perils of cherry picking

One trouble with cherry picking is that you have to be very careful not to change anything or the whole thing falls apart.

Dennis Avery picks up Bob Carter's "Global Warming stopped in 1998" cherry but fluffs it

The official thermometers at the U.S. National Climate Data Center show a slight global cooling trend over the last seven years, from 1998 to 2005.

Nope, they show warming.

i-587303f616324a653fbf3f567494545f-global-blended-temp-pg.png

What went wrong? Well for the Carter cherry pick, you have to thread the needle by picking 1998 as the starting point and you have to use the CRU data. (So I'm threading a needle with cherries here.) In the NCDC data, 2005 is the warmest year ever recorded. Never mind. Wait a year, and if 2006 doesn't set a new record, you can claim global warming ended in 2005.

More like this

Well for the Carter cherry pick, you have to thread the needle by picking 1998 as the starting point and the CRU data.

I'm trying to parse that sentence and failing. Are you saying that 1998 is the CRU data? Is there a word or two missing in there, like "using" or "switch to"?

[I reworded it a bit. Is that clearer? Tim]

http://www.sirhumphreys.com/node/5919

Care to add some expert comment Tim?

This guy has been sounding off on this for daays now. Not being an expert myself I'd like to know if he's talking rubbish or not......

"Using," I think. The links discuss all of this, but just to try to put it in a nutshell: CRU, NCDC (and GISS) all showed 1998 as the prior high. CRU showed 2005 as very slightly cooler than 1998, whereas NCDC and GISS showed it as very slightly warmer. The difference was well within the error bars for all three, so a fair statement would be that 2005 was about the same as 1998, with the caveat that 1998 was an outlier due to being boosted by a strong El Nino (which temporarily spreads the Pacific warm pool out over a much larger surface area and results in an artificially high global average surface temp), and so the definite warming trend clearly continues.

At some point in the next few years there likely will be a non-El Nino year that will be unambiguously warmer than 1998 or 2005. We can then look forward to Carter and his crowd proclaiming that global warming started in 2005, right? No doubt they will accompany that with a call for strong action. :)

By Steve Bloom (not verified) on 30 May 2006 #permalink

The "fluffs it" link is broken; I believe this is the same piece.
Interestingly, Avery quotes none other than Fred Singer as, in effect, admitting that starting with 1998 is cherry-picking:

Fred Singer, a well-known skeptic on man-made warming, points out that the latest cooling trend is dictated primarily by a very warm El Nino year in 1998. "When you start your graph with 1998," he says, "you will necessarily get a cooling trend."

So -- we have a non-El Niño year that (as Steve points out above) is just about as warm as an earlier year Fred Singer admits is a warm outlier, and Dennis Avery thinks this helps his case exactly how?

As with all fruit harvesting, timing is ALSO critical.

If you don't pick plums at the right time (wait too long) they become prunes.

Grapes become raisins.

Cherries become....

What DO cherries become? (other than rotten?)

By laurence jewett (not verified) on 30 May 2006 #permalink

Avery also missed the boat on the "imminent ice age" business; see http://www.cfa.harvard.edu/~wsoon/DaveLegates03-d/Bergeretal03.pdf (IMHO very much worth the time to read for those not already familiar with this area of the science). And hey, look whose web page! Notwithstanding where I found this public copy, I think Berger et al is more or less accepted science.

Oddly, I heard a talk by Tim Flannery on the radio last might in which he seemed to accept the imminent ice age stuff. He even quipped that maybe governments should buy up all the coal now so that it can be burned when we really need it!

By Steve Bloom (not verified) on 30 May 2006 #permalink

The "seven year cooling trend" meme is hot, hot, hot right now. I saw it in the comments section over at this Marginal Revolution article.

...

I wonder about economists. They seem to think that their profession is the only one that really matters, and a gut feeling from a colleague is worth more than the expert consensus of another field. I know, many fields are guilty of that, but economists just rise to a unique level of out-of-expertise hubris.

Harald, please don't judge all economists by Dennis Avery.

By Ian Gould (not verified) on 31 May 2006 #permalink

Dennis Avery is a chemical company shill, not a working economist (saying you studied ag econ doesn't make you an economist).

Best,

D

I would qualify Harald's statement:

"I wonder about economists at conservative think tanks."

I once had an emial exchange with an economist at Hudson Institute (discussing global warming, actually) and I mentioned some research Amory Lovins of Rocky Mountain Institute had done on energy efficiency - composite specifically high fuel economy vehicles that are also safe ( engineering analysis essentially, and on a subject on which Lovins is a REAL expert)

The "economist" dismissed my argument with a simple wave of the hand: "Amory Lovins is not an economist".

Perhaps not , but he IS a physicist....

and I'd rather have a physicist doing my economics than an economist doing my physics (or engineering) any day of the week.

By laurence jewett (not verified) on 31 May 2006 #permalink

Even if you use the CRU data and start in 1998 the linear trend is up. His supporters/promoters probably don't care much, about this minor technical "inconvenience". It shouldn't come as a suprise that Bob is rather vague about how he calculates a cooling "trend".

Ian Gould: I wasn't. I was thinking of the fairly reasonable economist Tyler Cowen, who nonetheless accepts the not nearly so reasonable arguments of fellow economist Arnold Kling: that climate models look a bit like macroeconomic models, and those were wrong in the past.

To me that shows it isn't just economists like McKitrick that suffer from specialisation hubris.