Saturday, February 19, 2011


So Eli has been looking for a couple of weeks at a paper by Georg Feulner, from the Potsdam Institute for Climate Impact Research, "The Smithsonian solar constant data revisited: no evidence for cosmic-ray induced aerosol formation in terrestrial insolation data" that is under review at Atmospheric Chemistry and Physics. Eli was holding off until at least some of the reviews were in to avoid the food fight that blogs can bring to the table, but he has been gazumphed by Rasmus at Real Climate.

The bottom line is that an earlier paper by Werner Weber claimed to have found a "Strong signature of the active Sun in 100 years of terrestrial insolation data" (the Acrobat file can be downloaded), but Feulner, who understands the data set that Weber used for the period 1905 to 1954, the Smithsonian Astrophysical Obvervatory solar irradiance series, thinks it's hooey and explains why.

A careful re-analysis of the data on which these claims are based shows that these trends are due to the effects of volcanic eruptions (and other sources of aerosols) and due to seasonal variations. None of the three quantities shows any significant trend with sunspot 10 number once these effects are taken into account (see the summary in Table 1). This illustrates once more that extreme care must be taken to understand any systematic bias of a dataset when investigating possible trends.
Well, the first referee's comment is in and if Feulner were O'Donnell the gloves would be off (actually, if you read the review, the gloves are off)
This paper is effectively a technical comment on the paper by Weber in Ann Phys
(Berlin). For a long time scientific practice has generally been to publish comments
in the same journal as the original paper, to allow comments and corrections (and
the courtesy of a reply from the author) to be associated with the first piece of work.
The hurumphing goes on a bit, but the review is open and Eli suggests that interested bunnies go read it. However the next part is precious
Throughout there is an assumption of a linear relationship between sunspot number and cosmic ray aerosol production. What is the basis for this ? The variable experimental work probably does not support this, for example there is a square-root relation between ion production and ion concentration.
Well maybe because Weber analyzes the relationship between sunspot number R, the top of the atmosphere solar insolation S, pyrheliometer measurement of solar intensity I, pyranometer values of the solar aureole and the water vapor content W using linear regression?
I have analyzed the S, I,A andW data with respect to their linear dependence on the sunspot numbers R of the respective days [19]. For this analysis various data prearrangement strategies have been employed; all of them produce very similar findings. Here, I report results based on the simplest scheme which evaluates a data group by standard linear regression
and Feulner does exactly the same analysis. There is no assumption about cosmic ray aerosol production being linear with sunspot number or not. Feulner does not need to get that far to show that Weber's analysis was, wait for it, Eli's favorite description of this stuff, naive. Weber, of course is a condensed matter theoretical physicist who, judging from his paper, has a full blown case of physicist's arrogance.

Eli is not going to engage in a Steve and Jeff act, wondering who wrote this review but it is a great illustration of the rough and tumble of peer review.


David B. Benson said...

So this time we are not going to throw the snake oil around?

Pinko Punko said...

The reviewer works at GPMI.

Goal Post Moving Inc.

A joke I have amongst my biologist colleagues, about when we present data appearing to rule out some model or other and get comments from the gallery relating to unsupported "what if?" scenarios, I think to myself "we have also not ruled out space rays in this phenomenon."

I chuckle at the irony.

EliRabett said...

That, dear David, is the bunnies job. Eli just sits there and tutt tutts. OTOH, the Rabett eagerly awaits the replies and further reviews. Oh yes, GPMI, know them well(unfortunately, and have seen their fine work.

EWI said...


While Richard Tol may have fled from the questions on Mr. Rabett's blog and the Internet in general (apart from updating his family blogs), he does appear to have found time to pester the BBC with emails, and send the results to Delingpole (his newest bestest friend):

J Bowers said...


Great!! Here's a more complete report:

Absolutely brilliant! Beddington tells it straight.

Anonymous said...

"Beddington said he intends to take this agenda forward with his fellow chief scientists and also with the research councils."

The sleeping bear is stirring, and ain't waking up in a good mood. Actual defamation suits, rapid response teams, now the redoubtable Beddington (a chap with bottom, as we Brits say). There'll be squeals before bedtime, mark my words.


VDP said...

I just found a page indicating a "Guido Fuchs" (managing editor) and a "Werner Weber" has been in the same school only 2 years apart. I have no idea if they really are the same people but it begs further investigation...

A google search on their names.

It is a stroke of genius to publish in an Interactive Open Access Journal. It allows us all to see GPMI in action + it ensures a fair editorial process.

VDP said...

Werner Weber from that school-page is not the correct person. The age cannot be correct.

I still think it was the correct choice to publish somewhere else (and this open discussion type journal is great). Eckern (editor) and Weber does seem to frequent the same circles. I believe that Feulner would have received a fair process. However, there is no need to put Eckern in that awkward position.