Back in 1998, I got a mailing that included a very professional looking paper casting doubt on the link between human activities and global warming. Two things about the paper caught my eye. First was its source: it was mailed from the Oregon Institute of Science and Medicine in Cave Junction, Oregon. Now it so happens I know where Cave Junction is. It's a wide spot in the road, so called because there's a turnoff for Oregon Caves National Monument. Cave Junction is not a major population center, much less the site of a university or major research labs. (Caveman Campers, which obviously take their name from the caves, are located not in Cave Junction but in nearby Grants Pass.)
The paper was called Environmental Effects of Increased Atmospheric Carbon Dioxide by Arthur B. Robinson, Sallie E. Baliunas, Willie Soon and Zachary W. Robinson. It had a lot of interesting graphs of carbon dioxide and temperature trends, and looked for all the world like a reprint from a journal. Same type fonts, similar paper stock, the works. The problem came when I looked for a citation, because if I were going to use any of this information, I would certainly need to provide a citation. And all scientific reprints have one, either at the beginning, the end, or on the top or bottom margin of each page. This one didn't. Now if you want to disseminate information, you can simply photocopy your paper and send it out. Even some high-class journals do this. But this one was professionally printed, folded, and stapled on the spine. Someone had gone to a lot of trouble to make this look like a published scientific paper. And that bothered me. I filed the paper away.
So in October, 2007, I got a mailing that was deja vu all over again. This one was from the Petition Project. It had a cover letter by Frederic Seitz, President Emeritus of Rockefeller University and past President of the National Academy of Sciences, asking me to sign a petition urging Congress to reject the Kyoto Accords. There was a petition card, and there was a reprint, this one on glossy paper, in color, and again bearing the address Cave Junction, Oregon. This article, also titled Environmental Effects of Increased Atmospheric Carbon Dioxide, was by Arthur B. Robinson, Noah E. Robinson and Willie Soon. This one did bear a citation, from Journal of American Physicians and Surgeons (2007), vol. 12, pp. 79-90.
Just to close the circle, the first two references cited in the paper were to Environmental Effects of Increased Atmospheric Carbon Dioxide by Arthur B. Robinson, Sallie E. Baliunas, Willie Soon and Zachary W. Robinson, in Journal of American Physicians and Surgeons (1998; then known as the Medical Sentinel, vol. 3, pp. 171-178), and Environmental Effects of Increased Atmospheric Carbon Dioxide by Willie Soon, Sallie E. Baliunas, Arthur B. Robinson, and Zachary W. Robinson, in Climate Research (1999, vol. 3, p. 149-164). (I'll say this - they make bibliographic searches easy!) So it's possible the 1998 paper was a preprint or reprint of the Medical Sentinel paper.
So why publish a paper on climate change in a medical journal? And just what is the Journal of American Physicians and Surgeons? A look at the contents from the last few issues is revealing:
A scan of the journal's contents shows very little real science. It may well be that the two papers on climate change are the most scientific papers the journal has ever published. There are a lot of articles critical of medical peer review and government regulation. And there are a host of tangential articles and book reviews favorable to far right and libertarian ideology. I could see a medical journal publishing a review of an extreme book like The Bell Curve that was widely publicized and had a bearing on medicine, but what serious medical journal would ever have a reason to review books on privatization of roads, the Minutemen, the 101 most dangerous academics (I was excluded both from this and People's list of sexiest men. O the injustice!), or a hostile revisionist work on Lincoln?
Depending on which site you visit, Frederic Seitz was either a brilliant scientist who adds great weight to the skeptical arguments against global warming, or a corrupt charlatan in the pockets of any industry willing to pay for junk science. The tone of most articles on both sides of the divide is about equally hysterical.
There's no doubt his early career was impressive. From 1949 to1968 he was professor of physics at the University of Illinois, where a research laboratory is still named for him. He helped bring future Nobel laureate John Bardeen to the university. He was president of the National Academy of Sciences from 1962 to 1969. He left Illinois in 1968 to assume the presidency of Rockefeller University, where he served until 1978. By that time he was 67 and in a position to retire with an illustrious career behind him.
Instead, he began working as a paid permanent consultant for the R.J. Reynolds Tobacco Company, in which capacity he served until 1989. According to a purported company document posted on the anti-tobacco site tobaccodocuments.org, by 1989 some in the company viewed him as "quite elderly and not sufficiently rational to offer advice." By the mid 1990's he was emerging as a climate change skeptic. His role in promoting the 1998 paper led to the National Academy of Sciences dissociating itself from him. A New York Times story dated April 22, 1998 stated:
The National Academy of Sciences has taken the extraordinary step of disassociating itself from a statement and petition circulated by one of its former presidents that attack the scientific conclusions underlying international efforts to control emissions of industrial waste gases believed to cause global warming.
Many atmospheric scientists and ecologists who believe global warming to be a serious threat had expressed anger and alarm over the article because it was printed in a format and type face similar to that of the academy's own journal, Proceedings of the National Academy of Sciences. [emphasis mine] In his letter, Dr. Seitz, a longtime skeptic on the question of global warming, also identified himself as a past academy president.
Dr. Arthur B. Robinson, the leading author of the article circulated with Dr. Seitz's letter, said it had been submitted for review by other scientists and for publication in a scientific journal but had not yet been published. [Medical Sentinel?]
Dr. Robinson said he ''never intended to imply'' that the academy endorsed it. He also said that he ''just wanted to put it into familiar format,'' and that he liked the academy journal's format. He noted that nowhere did the article say it was endorsed by the academy or published by its journal.
Dr. Seitz, a physicist who was president of the academy in the 1960's, said, ''It's true; the academy was not involved.'' He also said he had urged the authors of the Robinson paper to ''withdraw it'' and submit it to a journal for publication.
By 1998, the Internet was well established and there were all manner of pleasing formats available for documents. So there was no justification for mimicking the NAS format, much less using a special paper stock and spine stapling to make it look like a real reprint. And if Seitz felt the article should be withdrawn, why didn't he insist they remove his name from the mailing, or at least issue a public disclaimer once it went out?
The tobacco industry reached all time lows of ludicrous crank science in their denials of the health dangers of smoking. For my money, Seitz' association with tobacco, and even worse, his failure to acknowledge any wrongdoing, hopelessly taints his credibility.
With all the hype abut "Climategate" and the supposed discrediting of global warming, it's worth noting that the movement was spearheaded for some time by someone who had already been deemed too senile to front for the tobacco industry.
All the figures and data presented in the 2007 paper appear legitimate, and I've heard no allegations of any fudging or manipulation. It's the interpretation rather than the actual data that are flawed. Since there are 27 figures in the paper, the bulk of the most interesting information resides there, and I will simply reproduce the figures in order and comment on them, with occasional references to the text. Apart from cleaning up file conversion blemishes and changes in the color palette to reduce file size, no substantive changes have been made.
Figure 1. The above figure is based on Keigwin, L. D. (1996; Science v. 274, 1504-1508), and, apart from connecting lines across gaps in the older data, is just as originally published. It shows estimates of sea surface temperature in the Sargasso Sea, a measure of climate variation. A warm period called the Medieval Climatic Optimum (now more commonly called the Medieval Warm Period) extended from about 800 to 1200, followed by the Little Ice Age, from 1200 to perhaps the present.
This graph clears up a major mystery about the Little Ice Age. Various authors use the term in wildly different ways, beginning anywhere from 1200 to 1500, ending anywhere from 1400 to now. From the figure, we can see there were two Little Ice Ages, with a break in the 15th and 16th centuries.
The question whether the observed warming in recent decades is merely the recovery from the Little Ice Age or something else is one of the more legitimate scientific debates about global warming.
The table above simply shows that the graph in Figure 1 represents the overall consensus of the scientific literature. Although I am a bit perplexed by the probabilities. Rewording the last question a bit for direct comparison, we can add a fourth table row:
|20th Century not the warmest century?||64||7||14||99.9999|
The idea that there was a cold period from 1300-1900 has a 105-2-2 consensus and a 99.99 per cent probability. The idea that the 20th century is not the warmest century in the record has a much weaker consensus 64-7-14, but nevertheless has a 99.9999% probability?
Figure 2. The above graph shows glacier shortening plotted against rising fossil fuel consumption. And here's where the science starts to get wobbly. The text and figure caption assert that glacier shortening began before the increase in the use of hydrocarbons and follows a mostly linear trend despite a sharp rise in the use of hydrocarbons.
Why hydrocarbons? Those only include oil and natural gas. If we're going to assess the impact of carbon dioxide, we need to look at all the sources of carbon, including coal. In fact, before coal became widely mined, the Industrial Revolution was fueled by wood. Considering the small coal use before 1900, the wood contribution was probably very small.
Considering the natural fluctuations in glacier length, which make it difficult to decide exactly where to define the start of glacier recession, the onset of glacier recession and the output of significant industrial carbon dioxide coincide closely. Putting a divide at the start of hydrocarbon production is a bit of sleight of hand, a subterfuge to distract from an actual close correlation.
Why don't the slopes match more closely? There could be many causes, but the list would certainly include the 19th century clearing of farmland and deforestation of North America. Those both would contribute significantly to carbon dioxide output. Although, if we remove the trend line, the glacier curve is concave upward until 1900, just like the curve of coal usage, levels for a bit, and then becomes concave upward again as petroleum usage takes off. The curve actually tracks fossil fuel use much better than the trend line suggests.
Figure 3. This is a plot of solar activity, Arctic temperature, and fossil fuel use. Now sunspots and solar flares affect solar brightness, so a link between solar activity and climate is certainly possible. But notice again the emphasis on hydrocarbons and the neglect of coal. Actually, it looks very much like the upward slope of temperature and the use of fossil fuels match quite closely.
Figure 4. Okay, here's the graph of temperature showing the increase with time, but note that we don't see the graph of fossil fuel use. This graph shows that superimposed on the long term increase are pretty substantial ups and downs of a few decades in duration. Very true. Also quite irrelevant to the fact that there is a long term increase.
Figure 5. Comparing this graph with Figure 3, we see that a tenfold increase in fossil fuel use is supposed to have had no effect, but a change in solar radiation of 0.2% is supposed to have affected climate. The paper states:
Between 1900 and 2000, on absolute scales of solar irradiance and degrees Kelvin, solar activity increased 0.19%, while a 0.5 C temperature change is 0.21%. This is in good agreement with estimates that Earth’s temperature would be reduced by 0.6 C through particulate blocking of the sun by 0.2%
This whole section is just wonderful in its physical awfulness. How did Willie Soon, who has a background in astrophysics, let his name get attached to something like this?
Figure 6. Now this graph is just so silly it has no place in a scientific journal at all. To see why, imagine this is a study on obesity, and it finds that over a certain period of time the average weight of Americans has increased by one pound. That means that the weight Americans have gained more than offsets others have lost. In a population of 300 million, that's a very significant change. It takes a lot to change the average of a large data set by even a small amount. The bottom two bars would be like trivializing the weight gain by comparing the range in weights of individuals, say toddlers against fullbacks and sumo wrestlers.
The second bar, showing 50-year variations in the Sargasso Sea, is meaningless, because a comparison with Figure 4 shows short-term cycles superimposed on a long-term trend, and the short term variations are a lot larger than the long term change. So what? There is still a long term change.
Changing the average of a very large data set means changing the state of a very large number of data points. That's why it is so hard to change averages of large data sets, and why even a small change is so significant. Considering the enormous variability in daily and seasonal temperatures, the fact that the overall average of this data has increased is extremely significant.
Figures 7-9. Graphs of rainfall, severe tornadoes, and hurricanes making landfall.
Figure 10. I don't know about these captions. Having five hurricanes with wind speeds over 80 meters per second (180 miles an hour) since 1969 (the infamous Camille) and none before then sure looks like an increase to me. The number of violent hurricanes looks more like a decrease followed by an increase.
Figure 11. They really shouldn't have put those red trend lines in. Supposedly the trend lines show that the rise in sea level is erratic and complex, with steep and flat stretches. What we actually see is a steep rise from 1860 to 1905, corresponding to the increase in coal usage, a flattening from 1905 to 1935 corresponding to a flattening of the usage curve, a steep pitch 1935-1970, corresponding to the steep increase in usage of hydrocarbons, a flattening 1970-1985 (connected to OPEC and ceilings on output?) and then another rise corresponding to rises mostly in natural gas and coal use. Looks like a nearly perfect correlation to me. If I set out to show a detailed correlation between fossil fuel use and environmental effects, I couldn't have done any better.
For the record, I seriously doubt that global environmental changes track fluctuations of fossil fuel use that closely. The global feedback loops and interactions between systems are far too complex, and they will exhibit time lags. But it's a far better interpretation of the data than the claim that there's no correlation at all.
Figures 12. So here we go again with the graphs and fossil fuel use. This time it's the relation between glacier shortening and sea level rise. Again we see the same artificial divide when oil and gas use steepen, and the same neglect of the contribution due to coal. And when we take coal into account, we find a very strong correlation. Note that in Figure 12 the slopes of the sea level and glacier curves on the one hand and the fossil fuel use actually match quite well.
Figure 13. Graph supposedly showing the link between temperature and solar activity. The glacial and sea level curves have been shifted left by 20 years to factor in the time lag between temperature change and response.
First, note that the temperature curves and solar activity are only loosely correlated at best. It would be interesting to calculate a correlation coefficient rather than rely on visual comparison. Note that the Arctic temperature peak happens about 1940 but the solar activity peak happens about 1950 - ten years later. The lowest dip in solar activity about 1795 predates the dip in temperature by about 20 years.
Glaciers aren't the only thing with a time lag. An increase in CO2 ought to result in a delayed warming of the climate because it takes time to store heat in the oceans as well as saturate all the carbon dioxide sinks in the environment. If we shift the temperature curves left along with the glacier and sea level curves, then the first warming trend (1910-1940 gray band) could be interpreted as a delayed response to increased coal usage in the late 19th century, the cooling trend (white band 1940-1975) as a delayed response to the level fossil fuel use from 1910-1940, and the present climb (gray band 1975-present) as a delayed response to the steep increase in fossil fuel use after 1940.
Again, I don't really think changes in major global systems track changes in fossil fuel use that simply. But as with Figure 11, it's a better interpretation of the data than the one presented by the authors.
Figure 14. Somebody help me wrap my head around this. We have a graph of temperature change from 1980 to 2005. There's an overall increase globally, in the Northern Hemisphere, in the Southern Hemisphere, and in the tropics. In a paper opposed to the idea of global warming, we see that temperature curves everywhere - both hemispheres, global, and tropics, show a steady increase. So for the many people who still deny there's a warming trend at all, what's their answer to this?
Figure 15. This graph shows that weather stations in more populous counties show a steeper temperature increase due to the well known urban heat island effect. Very true, and a known complication in interpreting long-term temperature trends. The figure caption states:
The points marked “X” are the six unadjusted station records selected by NASA GISS for use in their estimate of global surface temperatures. Such selections make NASA GISS temperatures too high.
GISS is Goddard Institute of Space Studies. The total number of stations used for establishing temperature trends globally is in the hundreds. But while looking this up, I encountered an interesting sidelight. A correction to the global temperature trends was widely trumpeted as showing that recent years were not unusually warm, that in fact 1934 was the warmest year on record. A page on the GISS site explains:
Contrary to some statements flying around the internet, there is no effect on the rankings of global temperature. Also our prior analysis had 1934 as the warmest year in the U.S. (see the 2001 paper above), and it continues to be the warmest year, both before and after the correction to post 2000 temperatures. However, as we note in that paper, the 1934 and 1998 temperature are practically the same, the difference being much smaller than the uncertainty.
So if there's no change in ranking, and GISS had already said in 2001 that 1934 was the warmest year, and that it was, for all practical purposes, tied with 1998, who's making all the misleading claims?
Figure 16. So we have carbon dioxide released from the oceans and we have interglacials at the same time. This shows that anthropogenic carbon dioxide does not contribute to global warming how? The paper says:
The hypothesis that the CO2 rise during the interglacials caused the temperature to rise requires an increase of about 6C per 30% rise in CO2 as seen in the ice core record. If this hypothesis were correct, Earth temperatures would have risen about 6 C between 1900 and 2006, rather than the rise of between 0.1 C and 0.5 C, which actually occurred. This difference is illustrated in Figure 16. The 650,000-year ice-core record does not, therefore, agree with the hypothesis of “human-caused global warming,” and, in fact, provides empirical evidence that invalidates this hypothesis.>
Carbon dioxide increased by 30% at the start of interglacials, and resulted in a 6 degree global temperature rise. Hmm. If only six degrees marks the difference between an ice age and an interglacial, how exactly does this support the idea that changes of only a few degrees are no big deal?
Well, goes the argument, we've seen an equally big change in carbon dioxide in the last century and no concomitant abrupt rise, therefore we're not having any effect. Except the temperature graph above doesn't have vertical slopes: the actual slopes during all the warming periods are steep but easily span a few thousand years. So why would we expect to see instantaneous change? It takes time to melt ice caps. A more correct conclusion is that if a 30% increase over a few millennia has dramatic effects, a much faster anthropogenic change will have the same effects, only faster.
Figure 17. Global carbon dioxide levels.
Figure 18. The caption reads: "Qualitative illustration of greenhouse warming. “Present GHe” is the current greenhouse effect from all atmospheric phenomena. “Radiative effect of CO2” is the added greenhouse radiative effect from doubling CO2 without consideration of other atmospheric components. “Hypothesis 1 IPCC” is the hypothetical amplification effect assumed by IPCC. “Hypothesis 2” is the hypothetical moderation effect.
"Qualitative" means no numbers. Water vapor accounts for 90% of the earth's natural greenhouse effect. The effect of carbon dioxide alone is shown in the second bar. The combined effect of carbon dioxide producing warming, that then results in higher humidity and more of other greenhouse gases, is shown in the third bar. The fourth bar appears to display possible moderating effects like increased cloud cover.
Except with no numbers, this is a meaningless graph. What does a roughly 20% increase in the greenhouse effect mean in terms of temperature change? Well, that's why we do those much maligned computer models.
Figure 19. From the "true but so what file." The computer model uncertainties are indeed a lot bigger than the effects of increased carbon dioxide, a problem that has been noted by everybody since day one. But if it were all this hopeless, we'd expect to see as many models predicting cooling as warming. And we don't. Because carbon dioxide traps heat, and the more there is, the more heat is trapped.
Figure 20. Why do we have graphs spanning centuries showing that temperatures go up and down, but a graph spanning only 25 years is confidently labeled "Atmospheric methane is leveling?" And since when does an upward slope become "level?" The rate of increase has slowed. We can say it has leveled when it becomes level. And why is it wrong to assume present temperatures will continue to increase but all right to assume that the rate of increase on this graph will continue to decline?
About at this point the climate science in the paper ends and it becomes an opinion piece on policy, arguing that carbon dioxide increase is beneficial and that reductions in carbon dioxide emissions will be economically disastrous.
Figure 21. Graph showing mean tree ring width in bristlecone pines, showing a steep rise in the last century, which the authors attribute to increasing carbon dioxide. Given the harsh high altitude habitat of bristlecones, the increase could just as likely be attributed to warming. But considering the sheer mass of research on the effects of elevated carbon dioxide on plant growth, if this is the best evidence the authors can muster, they haven't researched the literature very much.
Figure 22. Graph showing increase in U.S. Forests 1950-2000. Completely true (the regrowth of forests, especially in eastern North America, has been called one of the most unnoticed environmental improvements ever.) But what does this have to do with carbon dioxide or climate change? The change is mostly due to abandonment of marginal farmland and reforestation or recovery of formerly cut areas, plus improved forestry practice and fire control. To assert, as the article does, that "Much of this increase is due to the increase in atmospheric CO2 that has already occurred" is just plain absurd.
Figure 23. It just keeps getting better. Not only will we get more biomass with more CO2, but the effect is greatest for resource-stressed plants. We'll be able to farm the deserts, old gravel pits, whatever. Only note the graph is in percents relative to the control groups. What we don't know is whether the increased biomass for the stressed plants offsets the decreased biomass from the stresses. In other words, is a stressed plant with 120 per cent growth enhancement from 600 ppm CO2 enrichment bigger and healthier than a non-stressed plant with no enrichment?
Figures 24-25. Before we celebrate the imminent end of world hunger, it might be appropriate to consider some of the other research on carbon dioxide and plants. Yes, there is more biomass, if the plant has all the other things it needs to grow. Two of those other things are water (that's where all the hydrogen in organic chemicals comes from) and nitrogen. Nitrogen makes proteins, among other things. Some studies indicate there will be more biomass but poorer nutrition for herbivores - more empty calories. If carbon dioxide goes into increased cellulose, it will be good for paper manufacturers and termites and firewood and absolutely useless otherwise. Other studies suggest that to achieve these enhanced yields, we'll have to irrigate more (further stressing water supplies) and fertilize more (further increasing polluted runoff.)
Figures 25-27. Graphs showing U.S. energy imports and exports and cost of various energy sources. Figure 26 shows delivered energy costs for Great Britain. All true, but the diagrams raise a host of questions:
Then there's this remarkable insight from the text:
Reactor accidents are also much publicized, but there has never been even one human death associated with an American nuclear reactor incident. By contrast, American dependence on automobiles results in more than 40,000 human deaths per year.
What's the logic here? Most likely, that Americans (people in general) overestimate risks when as risk is perceived as having a negative outcome but underestimate risks when the risk is perceives as positive. What's the relevance to global climate change?
It's sadly ironic that the Oregon Institute of Science and Medicine chose to send out a piece of counterfeit scholarship, because the 1999 paper in Climate Research is far more respectable. It was reviewed by a global warming skeptic, but one with serious academic credentials. Had they sent out a reprint of that article, along with an op-ed piece, there would be no cause for complaint.
Co-authors Baliunas and Soon both have legitimate academic credentials and numerous publications in astronomy. Baliunas has been criticized as not being an expert on climate science, but her credentials as an astrophysicist place her in a far better position to evaluate the physics of global warming than most other people on either side of the debate.
Since this article is rough on the climate skeptics, it's only fair to dust up the other side a bit. The attacks on Soon and Baliunas remind me a bit of the attacks on the late Carl Sagan and his support of the nuclear winter hypothesis (which receded from prominence after the Cold War ended but was not discredited.) I'm not suggesting that Soon and Baliunas are comparable to Sagan either in professional stature or the legitimacy of their positions, but in both cases the criticism was "what does an astronomer know about climate?" Quite a lot, it turns out. Climate is all about the physics of heat, and astronomers are better positioned to understand that kind of physics than 80 per cent of most other scientists and 100 per cent of non-scientists.
Just to show that bogus reasoning is not limited to climate skeptics, the site ecosyn.us has a page on Frederic Seitz that claims "Sallie Baliunas and Willie Soon are the next generation in mass murder deceivers." Baliunas and Soon may be guilty of bad science, but that's a bit removed from mass murder. And where people get funding doesn't necessarily make their findings invalid. The people who supply the world with energy might just be doing a few things right.
The only valid reason for rejecting research funded by the energy industry is because the research is flawed. In that case, the issue is the incorrect science, not the funding. So refute the science and don't bring in the funding at all. You're certainly entitled to scrutinize suspect research as hard as you like, but if it's sound, it's sound, regardless of who funds it. But the critique of Baliunas and Soon is mild compared to the hysterical rhetoric on some of the other pages at ecosyn.us.
The site has a bunch of complex graphics showing ties between individuals and organizations, of dubious relevance. They appear to have been inspired by the scene in Alias where the FBI agent shows Jennifer Garner the evil organization chart. Let's see. I'm at the University of Wisconsin-Green Bay, and I know for a fact that some faculty are members of the ACLU (it's on their resumes) and the ACLU has members affiliated with Islamic organizations, some of whom have ties to Al Qaeda. Well there ya go. Homeland Security, come'n'get me. It gets worse. I have colleagues who signed the Oregon Petition. That means I have ties to the organization that put out these phony papers. But it gets even worse than that, if you can believe it. During my military deployments I encountered a fair number of contractors from Brown and Root, which was later bought up by (cue the ominous music) Helliburton. Finally, the company was founded as Kellogg, Brown and Root - yes, that Kellogg - so if you ever eat Rice Krispies you're linked to mass murderers. Guess that makes us all cereal killers. Also, I'm no more than six degrees removed from Kevin Bacon.
Figures from the 2007 paper are, with predictable reliability, turning up in anti-climate change articles.
Say what you will about Al Gore and An Inconvenient Truth. You may see an excess of hype and overdramatization on the global warming side. You do not see believers in global warming mass mailing specious documents designed to look like legitimate scientific journal articles. That ought to tell you something. It tells me something.
Created 06 March 2007; Last Update 25 March, 2011
Not an official UW Green Bay site