Monday, April 01, 2013

Is Marcott et al a "Gross Misrepresentation?"

Roger Pielke Jr has a provactive post this morning that comes within a hair's breath of accusing Marcott et al of misconduct. He does accuse them, and the NSF of "gross misrepresentation," which I just don't see at all.

I like Roger, I like a lot of his writing, I loved his book, and he always gives me good interviews. But I have to disagree here.

Roger writes,
There is a big problem with the media reporting of the new paper. It contains a fundamental error which (apparently) originates in the NSF press release and which was furthered by public comments by scientists.

In a belatedly-posted FAQ to the paper, which appeared on Real Climate earlier today, Marcott et al. make this startling admission:
Q: What do paleotemperature reconstructions show about the temperature of the last 100 years?

A: Our global paleotemperature reconstruction includes a so-called “uptick” in temperatures during the 20th-century. However, in the paper we make the point that this particular feature is of shorter duration than the inherent smoothing in our statistical averaging procedure, and that it is based on only a few available paleo-reconstructions of the type we used. Thus, the 20th century portion of our paleotemperature stack is not statistically robust, cannot be considered representative of global temperature changes, and therefore is not the basis of any of our conclusions.
Got that?
How is this "startling?" They say in their paper that it's not robust:
Without filling data gaps, our Standard5×5 reconstruction (Fig. 1A) exhibits 0.6°C greater warming over the past ~60 yr B.P. (1890 to 1950 CE) than our equivalent infilled 5° × 5° area-weighted mean stack (Fig. 1, C and D). However, considering the temporal resolution of our data set and the small number of records that cover this interval (Fig. 1G), this difference is probably not robust.
(Emphasis mine.)

  • on the first page, right-hand column, they explicitedly note their reconstruction stops at "100 yr B.P," and their "present" is defined as 1950 CE.
  • The graphs on page 2 have the blade in different colors, and the captions make it clear these aren't their results.
  • On the last page: they write, "Our results indicate...." which is not the same as writing "Our results show...." Anyway this is the discussion portion of their paper, and what's wrong about discussing how their results look in light of the recent instrumental record?
That's just the paper. After its publication both Shakun (in the NY Times) and Marcott (in an email to McIntyre) indicate their blade isn't robust. That's a pretty clear statement (though I wish scientists would stop using the word "robust" when talking to the public, because it sounds overly wonky). Should they have completely avoided any mention at all of how their results compare to what's happening today? Every reader in the world is going to want to know that, and who better to answer it than these scientists?

For the same reason, the press release, which is setting the context, is going to address what every journalist is immediately going to want to know: what does this say about today's warming? Not, what does this categorically prove -- there is no such thing -- but what does this say, what does it "indicate?" Because here we have this 11,000 year reconstruction, and on the other hand we have the instrumental record over the last 130 years or so, so what do they say together? It would be irresponsible (and a waste of taxpayer-funded research) not to say something about the broader context in an issue of this importance.

So it seems to me that
(a) the scientists made it clear in their paper their results were not robust for 1850 CE onward
(b) they addressed, in their discussion and in interviews, the immediate question on everyone's mind: so how do your results compare to today's warming?

And today's warming looks pretty darn troubling, and my cynicism is wondering if that's what almost all of the paper's post-publication criticism is trying to suppress by any means necessary (although my guess is that's not Roger's motivation, which (I'm guessing) he sees along this line).

Update: I made much this same comment on Roger's blog, and he responded here.



William M. Connolley said...

Yeah, I think RP's gone a bit off the rails on this one.

Sou said...

Have to agree with William Connolley that Pielke Jr has jumped off the edge. Don't know what he was imbibing at the time. It looks as if he's still intoxicated with being on the same bandwagon as the rabble at WUWT.

His response to you, David, doesn't make any more sense than his original rant. It's pretty obvious that the earth is now getting as hot as ever it did during the Holocene - thanks to the work of Marcott et al and others (compared with current global surface temperature data). I see nothing in the NSF press release, interviews or the paper itself that says otherwise or any conflicts with the Marcott et al paper.

Roger will turn into a Curry or worse if he doesn't watch himself, and forever ruin whatever professional reputation he might currently enjoy. He's not got my respect. From this and gross errors in other blog articles of his I've learnt not to take what he says seriously - at least not without checking for logic and veracity.

I wouldn't take any notice of his off the cuff bloviations except to chide him appropriately (if you're a mate who has his best interest at heart). (His academic work can stand or fall on its own merits.)

Anonymous said...

Thanks Dave for this...It should be noted that the errors have yet to be corrected even though all the offenders have been notified. Each gives the impression the authors were specifically trying to hide information from people, when in fact, it is the offenders who are hiding information to give their readers the ability to evaluate the situation.

At this point, I cannot come to any conclusion other than the offenders, are in fact, purposefully hiding this information. I don't know about Lomborg, but McIntyre doesn't surprise me. Pielke, you'd think would care about this, but does not seem to. I'll also point out that one of the links that Pielke included to Tom Yulsman's post, has a correction specifically emailed from Marcott, before the FAQ, that told Tom "it would be incorrect to say the rate is greater than anything else seen in the past 11,000 years. When looking at our reconstruction of temperature it appears this way. [But] it is, in fact, just a resolution issue related to this dataset."

This is yahoo cached page from March 13th. It would seem dubious and deeply opportunistic to use this paper to show Marcott was attempting to misrepresent his research in the manner portrayed, upticks, resolution, and all.

Susan Anderson said...

I've been fascinated by this one. Simply put, but the language is far from easy to parse:

It appears they want to use the past, in the paper they wish to discredit, as evidence to discredit the present.

Go figure. I appointed myself to imitate Saint Sebastian (all those arrows) at DotEarth, but do not recommend anyone go see the attack machine at work over their, ably assisted by Prof. Revkin's choice to support McIntyre and Pielke and take McI's worse that Tamino plagiarized, as well as his "unbiased" noninterference with comments.