Climate science reporting: detecting bias

The Roger Pielke Sr. Research Group News (ClimateSci.org; one of Nature magazine’s Top 50 Science Blogs) has reported what it considers to be a case of bias against its research by EOS, the online publication of AGU, a “a worldwide scientific community that advances, through unselfish cooperation in research, the understanding of Earth and space for the benefit of humanity”.

The allegation is the result of the refusal to publish their article, based on a survey of climate scientists, that suggests “there is not a universal agreement among climate scientists about climate science as represented in the IPCC’s WG1“.

The results of the ClimateSci.org survey, of 140 climate scientists, reported these results:

• No scientists were willing to admit to the statement that global warming is a fabrication and that human activity is not having any significant effect on climate [0%].
• In total, 18% responded that the IPCC AR4 WG1 Report probably overstates the role of CO2, or exaggerates the risks implied by focusing on CO2-dominated Anthropogenic Global Warming (AGW), to a greater or lesser degree.
• A further 17% expressed the opinion that the Report probably underestimates or seriously underestimates the consequences of anthropogenic CO2 -induced AGW and that the associated risks are more severe than is implied in the report.
• The remaining 65% expressed some degree of concurrence with the report’s science basis

ClimateSci.org then tried to get this published in EOS, and Nature Precedings. They were turned down by both.

How should an environmental journalist respond to this?
Firstly, an assessment of whether or not ClimateSci.org is worth your time. And that’s by assessing how influential it is. So, back to their claim that they have been identified by Nature magazine as one of the Top 50 science blogs.

Top 50? Looking at this ranking shows that ClimateSci.org is ranked joint 47th out of 50; and joint with the blogs of a Canadian phd-student studying “the evolution and diversification of early jawed vertebrates (gnathostomes)” and Afarensis, the blog of a part-time blogger and anthropologist. (On a personal note, I am not knocking either of these sites, as they seem both to be genuine, honestly written, intelligent (as opposed to intelligent design) and entertaining). The top rated by Nature magazine (and in reality it is simply a Technorati ranking) was another scienceblog.com blog, Pharyngula, getting upwards to 700 comments on his articles. In publishing the Technorati ranking, the Nature.com badge of ‘Top 50′ must be seen as a commons vote, rather than a scientific evaluation of the content.

The IPCC Report. The ClimateSci.org article claims that:

there is not a universal agreement among climate scientists about climate science as represented in the IPCC’s WG1. The claim that the human input of CO₂ is not an important climate forcing is found to be false in our survey. However, there remains substantial disagreement about the magnitude of its impacts.

The issue is a dificult one for journalists without science training and without having read the IPCC WG1 report, to which ClimateSci.org are referring and against which their survey is measured. So a starting point would be to read the IPCC WG1 report (the summary for policy-makers, technical summary and FAQ would be a good start). Another good document, for student journalists affiliated to a University with an Athens account, was the editorial of the journal Global Environmental Change 18.1 (2008), that calls the WG1 Report:

very solid scientifically and is full of useful information, in general clearly presented with the degree of certainty or uncertainty in statements or projections explicity provided.

The editorial also states that:

On some of the most fundamental and important aspects of climate change, such as whether the world is warming and what is the primary cause, the debate is over in the scientific community. However, for many other important aspects of climate change, such as sensitivity of climate to greenhouse gas forcing, the role of aerosols in the radiative balance and in the hydrological cycle, and carbon cycle feedbacks, there remain significant and somewhat stubborn uncertainties.

Compare this to the framing of the debate from the ClimateSci.org article:

The issue of whether scientists agree about the causes of climate change has persisted in discussions of climate science in general and in the development of policy to respond to the threats implied by climate variability and change.

But ‘whether’ (or not) scientists agree about the ’causes’ of climate change is not a persistent question any longer within the scientific community and, importantly, it is a very different claim than the one that the data itself is making. The data is saying, as quoted above, that scientists disagree on the ‘magnitude’ of the ‘impacts’ of ‘human input of CO₂‘. But the written article is framing the argument in terms of disagreement over ’causes’, not ‘magnitudes’. This for me is the crux of the reason why I would as an editor have serious concerns about publishing such a report. If the langauge of the article reframes and distorts the argument so that it is saying something other than the data, on which it bases its argument, then there is something questionable going on.

There are causes (plural). This is accepted. That humans are having a significant impact on the changing climate due to the dangerous levels of emissions of greenhouse gases is also accepted, and by the ClimateSci.org survey itself:

No scientists were willing to admit to the statement that global warming is a fabrication and that human activity is not having any significant effect on climate [0%].

So, perhaps the survey and article was rejected because the data itself (0% of respondents believe human activity is not having a significant effect on climate) does not support the angle of the article as written: not about magnitude, but about causes.

Who are ClimateSci.org? (or, Who is Roger Pielke?) Taking a look at the Pielke Research Group homepage that is behind Climate Science, it is clear that Roger Pielke does not agree that the scientific and public communities should maintain a focus on the reduction of anthropocentric greenhouse gas (C02 and equivalents) emissions into the atmosphere. Papers with titles such as “Climate Scientist Roger Pielke Sr. Rejects the Notion that Elevated CO2 Levels are the Sole Culprits of Climate Change” and “Pielke Calls for New Direction in Climate Science” illustrate immediately that Roger Pielke and the research programme are antagonistic to the IPCC and its reports. However, Pielke believes his views are often misrepresented, and I haven’t spent enough time assessing them and learning about the things he’s talking about. Yet. His statement is long but interesting.

Ok, but who else has taken the time (and knows their stuff) to do an assessment? There is a raft of publication and debate around Pielke’s theories, and not all of them dismissive. In fact, one of Pielke’s recent articles on the ‘dangerous assumptions‘, along with two other climate scientists, was published in Nature. This was then picked up by Bill Miller at DeSmogBlog, a site generally known for identifying and uncovering greenwash in corporate America.

But then the first comment on this page, from David Ahlport, is scathing of Peike’s position and points a number of articles published at Grist, the environmental reporting site, that takes apart Pielke’s argument, claiming that Pielke is a ‘delayer’, one of the rhetorical categories of ‘climate denailist’, arguing that Pielke’s position is:

the predictable next step along the denialist stonewall approach.

 

1. It’s not happening!
-So lets not deploy current technology, shape policy, or build markets.

2. It is happening, but it’s not us doing it!
-So lets not deploy current technology, shape policy, or build markets.

3. It is happening, we are doing it, but maybe warming is a good thing!
-So lets not deploy current technology, shape policy, or build markets.

4. It is happening, we are doing it, warming very Very VERY bad thing, so bad infact that:
-So lets not deploy current technology, shape policy, or build markets. *Lets instead focus on adaptation.*

 

The rhetorical argument may change, but the bottomline position always stays the same.

And over at Grist, Joseph Romm continues the attack on Pielke, specifically around his claims on the IPCC Report, arguing that:

No, you aren’t a delayer because you’ve “asked what data would be inconsistent with IPCC predictions.” You are because you wrote a long post giving credence to the notion — which is clearly at odds with the data — that the climate is in a cooling trend. In fact, you begin with a graph that implies we’ve been in a major cooling trend since 2001 and you yourself write of “the recent cooling in the primary datasets of global temperature.”

So, is it a solid piece of research? Some questions that came to mind straight away: Why were there only 140 responses to the survey, of a total of 1807 contacted? This is a 7.7% response rate, which by most methodological standards is low. Who was asked, and who responded? ClimateSci.org asked:

a poll of scientists’ opinions in which authors of climate papers in the journals: Geophysical Research Letters (2007), Atmospheric Chemistry and Physics (2007), Climate of the Past (2007)… [etc]

But as ClimateSci.org also make clear:

as there is no currently available information on the complete community of ‘climate scientists’, the poll cannot be tested for statistical significance. In the absence of such information, the methodology adopted was the only practicable means of attempting such a poll. It should be recognised, however, that the methodology is not strictly statistically formal and the results should be viewed accordingly.

If the methodology is not ‘formal’ and and cannot be tested for ‘statistical significance’, I too as a publisher would be in doubt as to whether or not to publish the article, which seems to lay a strong claim to its data to back up a polemical and provocative position: of dissent among climate science as to the ’causes’ of climate change.

Summary. There are a number of tools you can use as a journalist reporting on climate change and the environment. A close reading of the texts is key: both the official texts, and the discourse of the article itself. You can assess the claims of influence of the publisher. You can investigate the sources of both its own theories and actions, and the theories and actions of those against which it shouts and rails. Most of all, though, time, thoroughness, patience to search for elements of a story which are not too far, but far enough, below the surface of the article.

I am sure there is more to discover here. I am in a way sympathetic to ClimateSci.org, because it does not feel to me that they are dissenters, denialists or sceptics with a specifically political or neo-liberal agenda to push (e.g. in the same way the Bush administration edited climate reports to assuage the oil and car lobby). Maybe just bad scientists? In terms of reputation, Grist does have far more of my trust. But then, admittedly, it is far more in line with my own inherent poltical values.

I also think that dissent has been quelled through what has been called ‘scientific reticence’ to question the details of climate change for fear that views could be picked up by the media and distorted, taken out of context, etc, as many scientists have said of Channel 4′s The Great Global Warming Swindle. As such, any dissent is often seen as detrimental to the overall direction in which the majority of scientists believe debate should progress.

However, is this case one of bias against EOS and Nature Precedings? I would think not. More a case of an article trying to shoehorn one argument onto data saying a different point, and neither with the methodological robustness that the scientific community are now demanding of the views on consensus on the causes, and magnitudes of those causes, of climate change.

Popularity: 2% [?]