This is something I should have blogged about long ago. Nothing I am doing here is new, just a rehash and recombining of other’s work.
(In fact, I forgot to actually post this until I got a comment today – March 15, 2012 – from Tony at The Air Vent who pointed me to a post by Jeff Condon there dated March 6, 2010 entitled Historic variations in CO2 measurements.)
Please have patience. The main point here is that the standard 19th century (read: pre-CO2, supposedly) CO2 level of 290 ppm was derived in 1938 from TOTALLY cherry-picked data – by someone whose paper was intended to show that CO2 was going to send us into a runaway greenhouse effect. According to a later scientist who looked at his methods, the average should have been 335 ppm.
My attention was drawn to this by Zbigniew Jaworowski on CO2, on the late John Daly’s site (which is still up and running, amazingly).
Have a look at the real data and the ones used in 1938:
Do YOU think the circled data points are representative? Or are they cherry-picked?
Jaworowski is extremely critical of G.S. Callendar’s 1957 paper, “On the amount of carbon dioxide in the atmosphere.” That paper was a rebuttal of a 1955 G. Slocum paper, after Slocum had written about Callendar’s 1938 paper, “The artificial production of Carbon Dioxide” in very unkind but civilized ways. Jaworowski (as did Slocum) accuses Callendar of cherry picking his data. Slocum’s paper,“Has the amount of carbon dioxide in the atmosphere changed significantly since the beginning of the twentieth century?” Month. Weather Rev., 1955(October): p. 225-231 shows how Callendar arbitrarily left out hundreds of CO2 data points and kept in only CO2 data that showed low CO2 levels.
Callendar’s Figure 1:
Note the 290 ppm line. Slocum showed that that level should have been 335 ppm.
I want you, the reader here, to think about that difference before going on.
To give a perspective on it, look at this NOAA chart of CO2 levels since before 1960:
Note the 1960 level. It is below 320 ppm. Also note that at Mauna Kea – currently the definitive CO2 data point – the level did not reach 335 until around 1979. Keep that all in mind as you go forward in this post. According to Slocum, the pre-1900 CO2 average was HIGHER than it was when he did his paper in 1955.
One aspect of this is that if Slocuim is correct, the CO2 level has only gone up from 335 to 385. That is an increase of only 15% – in over 100 years. Callendar’s 290 pre-1900 level increased to 385 is an increase of 33%. Which one do you think the alarmists want to use?
Later I will detail Callendar’s data selection. He does, in fact, explain why he uses some and not others, so we will then present both sides of this and let the reader make up his own mind (though my two bits are going into this, too). But for now I want to go present Jaworowski’s criticisms.
JAWOROWSKI’S SENATE STATEMENT
In his March 2004 Statement to the U.S Senate, Jaworowski has this to say about Callendar:
The notion of low pre-industrial CO2 atmospheric level, based on such poor knowledge, became a widely accepted Holy Grail of climate warming models. The modelers ignored the evidence from direct measurements of CO2 in atmospheric air indicating that in 19th century its average concentration was 335 ppmv (Figure 2) . In Figure 2 encircled values show a biased selection of data used to demonstrate that in 19th century atmosphere the CO2 level was 292 ppmv. A study of stomatal frequency in fossil leaves from Holocene lake deposits in Denmark, showing that 9400 years ago CO2 atmospheric level was 333 ppmv, and 9600 years ago 348 ppmv, falsify the concept of stabilized and low CO2 air concentration until the advent of industrial revolution .
Footnote 11 is Slocum’s paper on Callendar, and footnote 12 is Callendar’s paper.
Not one to take Jaworowski’s word for it and was fortunate enough to find Slocum’s paper critiquing Callendar. Jaworowski clearly states that the 19th century CO2 level was 335 ppm, not 290 ppm. This I thought was something I could sink my teeth into, and I was lucky that all the papers were available online and not behind paywalls.
Jaworowski got the 335 ppm number from Slocum’s paper. here is what Slocum had to say in his Abstract:
In this paper, the physical knowledge of atmospheric GO2 is examined and the available nineteenth and twentieth century observations of the atmospheric CO2 concentration are summarized to ascertain the extent to which they corroborate claims that the amount of atmospheric CO2 has increased since the nineteenth century. In the light of the uncertainty of both physical knowledge and of statistical analysis, it is concluded that the question of a trend in atmospheric CO2 concentration remains an open subject.
Slocum in the 1950s lived in a very proper and sedate time, apparent in his in saying that “the question of a trend in atmospheric CO2 concentration remains an open subject.” He, in fact, does everything in the paper but call Callendar a fraud – except when writing his conclusions about Callendar. In those Slocum is very gentlemanly.
WHAT DID SLOCUM SAY ABOUT CALLENDAR’S DATA?
First, Slocum talked about statistics:
“Since Callendar, by basing his hypotheses on statistical data, has tacitly invoked the laws of statistical evidence, it is fitting to examine the validity of his procedure, that
of using only the data he believed to be of the best quality available, rejecting the rest. The mathematics of statistics, and the experience of statisticians both indicate, as a general principle, that arbitrary rejection of data, without specific knowledge of their unreliability or unapplicability, is questionable.”
Fair enough, yes?
He goes on…
“At best, the omission of part of the data is not as necessary or as helpful as might appear at first thought, since it can be shown that when the means of two sets of data are compared, the presence of a given average bias in each set will not affect the difference nor the standard error of this difference, except as an added contribution to the variance of the sample. If, however, some of the data be selected to the exclusion of the rest, for the purpose, perhaps, of reducing the magnitude of the residual variance, due to crudity in some of the measurements, then, in addition to any unintentional bias that might be introduced in the comparison of the means, there might also result an underestimate of the standard error of the difference, due to the mistaken rejection of those of the extreme values which actually belong to the distribution. The result may be an entirely spurious accuracy in the means, which leads to unjustified conclusions. In the light of these considerations, a reexamination of the entire body of available measurements of the relative proportion of CO2 in the atmosphere may have some value.”
That is the purpose of the paper, then, to reexamine Callendar’s data and his omission of data.
ALL of Callendar’s data looked like this:
Take a moment and persuse the values in those two “MEAN” columns. ONLY LOOK AT THE 1800-1899 values, though. We are not so interested in the post-1899 values just now. (Slocum goes into them and rakes Callendar over the coals over his SECOND cherry picking in his paper, and I will let him speak for himself – later.)
How many values at or below 290 ppm do you see? I see three below 290 (tw0 just barely). A basic rule of thumb (to see if you have made a gross error) on simple mean calculations is that, for a sizable data set, approximately 40-50% of the data points will be below the mean and about the same number above the mean. If that doesn’t hold true, it is a good reason to go back and see if you did something wrong. There are, as I count them, 132 data points at or below 290, Callendar’s mean. I count 4,157 above 290. You tell me if 290 ppm can be the mean.
Slocum raked Callendar over the coals, but did not, in the end accuse him of out-and-out cherry picking the data. But he did everything short of it.
Callendar’s 290 ppm value – upon which so many CO2 claims are made – simply is false. His paper should be retracted. And any studies that reference it need to be either redone or retracted also. CO2 increases have not been from 290 ppm to whatever. They should be based on Slocum’s values of 335 ppm.