The “Hanno 2009” graph
Much ado about ... well, what exactly?
By Hanno Sandvik
One of the stranger experiences that have followed from my interest in climate, was a debate that has been immortalised in cyberspace as “the Hanno graph case”. A Google search for this phrase returns well above 300 hits. Of these, roughly 60 web pages will be able to tell you that I am “an obscure Norwegian ecologist”. Unpleasant as this experience has been (for me, that is), it taught an interesting lesson about the arguments, methods and desperation of climate change deniers.
The background was this: in September 2009, UNEP (the United Nations Environment Programme) issued a “Science Compendium” about climate change (follow this link for the Compendium’s current version; follow this link to see page 5 of the Compendium’s original version). The Compendium had the following image as its Figure 1.3:
Figure 1.3: Correlation between temperature and CO2. Atmospheric CO2 concentration and mean global temperature during the past millennium. CO2 levels (blue line, lefthand axis) are given in parts per million, temperatures (red line, right-hand axis) in degrees Celsius.
UNEP had taken this image from Wikimedia Commons (file name: CO2-Temp.png) and indicated its source as “Hanno 2009”. The explanation for this somewhat odd attribution was apparently that UNEP had downloaded the image in 2009, and that it had been uploaded to Wikimedia Commons by a user named Hanno. This Hanno was me. I had drawn the figure based on data from peer-reviewed scientific articles and uploaded it to Wikimedia Commons in December 2005. The temperature curve of the graph has long been known under the name “hockey stick”.
Within hours of the publication of the Compendium, a website called “wattsupwiththat” had started a campaign against it. Part of the campaign was a Kafkaesque guesswork regarding my person, i.e. bizarre speculations about who was hiding behind the “Hanno” pseudonym, and about my supposed personality traits. I was luckily unaware of both the Compendium and wattsupwiththat’s grotesque campaign until two days later, when my mailbox was flooded by e-mails that were as intimidating as they were cryptic.
During the following days I was contacted by a few journalists and interested lay persons, while several internet fora continued their blog-activism, and UNEP exchanged the Compendium’s Figure 1.3 with another image. After that, despite wattsupwiththat’s hopeful prediction that “The storm over ’Hanno 2009’ is very likely just beginning,” the interest in the story rapidly declined – which didn’t stop it from proliferating in the parts of the blogosphere controlled by climate change deniers for another month or so.
A revealing aspect of the campaign against the Compendium and my person, was that the focus of the attacks was constantly shifting. The following list gives an overview of the different objections raised. Although the list is non-exhaustive, I have tried to identify the objections that have been repeated most frequently (and/or perniciously):
- It was objectionable that I had provided the graph to UNEP. – As a matter of fact I hadn’t. Not that this would have had the slightest relevance, but I wasn’t involved in any of the stages of the writing, compilation or production of the UNEP Compendium. Which, in turn, led to the next accusation:
- It was objectionable that UNEP did not ask for my permission to use the graph. – Nobody needs to ask for my permission to use the graph. The very motivation of Wikimedia Commons is providing multimedia files that are not protected by copyright. When uploading the graph to Wikimedia Commons, I had released it under a “attribution-share alike” license, which means that everyone is free to use the figure as long as he or she attributes the originator and shares derivative works under similar licenses. The license does not entail that the originator has to endorse, agree in, approve of, or even be aware of, secondary use – to the contrary, originators explicitly give up such claims.
- It was objectionable that UNEP used Wikimedia as a source. – Using Wikimedia or Wikipedia as the only sources in science or decision making would indeed be poor scholarship. What UNEP was doing, however, was something quite different: the Compendium used the graph to visualise a statement that was made in the text and the validity of which was documented by the references cited there. No conclusions were derived from the graph itself. If (and only if) the authors of the Compendium thought that the graph’s source articles were trustworthy, there was nothing wrong in expressing these articles’ results graphically. The authors could have taken the original graphs from the original articles; they could have drawn the graph themselves, using the freely available metadata; they could have taken the graph from Wikimedia; or they could have taken the graph from any other source that provided it. None of the choices would have been more or less objectionable than any of the others, because the graphs would have been identical (except for their layout). Whether the authors of the Compendium were correct in thinking that the graph’s source articles were trustworthy, is another matter (and none that I am competent to answer). The latter issue is a question about science, not about who has downloaded whose graphs and from which website. To be sure, it would have been more correct of UNEP to attribute the figure to “Etheridge et al. 1998; Jones & Mann 2004; Jones et al. 2005; Hanno 2005”, in order to indicate the data sources in addition to the author of the figure.
- It was objectionable that Wikimedia provides “hockey stick” graphs. – Wikimedia is not a collection of truths, but a collection of multimedia files. No matter how controversial the “hockey stick” might be, there are exceedingly good reasons for Wikimedia Commons to provide a figure of it. If the graph is “true”, Wikimedia should have it so that people can see it. If the graph is proven wrong, Wikimedia should have it so that people can see how science advances. If the graph’s truth is still debated, Wikimedia should have it so that people can see what all the fuss is about. In short, the truth of the information conveyed by an image cannot be inferred from the fact that it is available at Wikimedia. A good start to find out how well-accepted the information expressed by an image is, is to read the Wikipedia entry that uses the image.
- It was objectionable that I had drawn a “hockey stick” graph. – I could have drawn the figure for any of the abovementioned reasons: that I believed in its message and wanted it to be known; that I doubted its message and wanted to expose it falseness; that I was agnostic to its message and wanted it to illustrate a recent debate in climate politics. My motivation for drawing the graph does, in fact, not have the slightest relevance, because my original motivation is entirely uncorrelated with the later uses to which other people may put the graph.
- It was objectionable that I had uploaded the graph to Wikimedia in spite of not being a climatologist. – Had I pretended that the graph was based on my own data, it would have been scientific misconduct to make it publicly available. However, the file description page of the graph states – in boldface and visibly to anyone who bothers to read – that I was the author but not the source of the graph. The sources, being peer-reviewed publications, were provided with complete references. My job as the author of the graph had been to make a curve out of the numbers listed in the sources. This can be accomplished by anyone who can read numbers and knows how to operate graphics editor software. No special knowledge of climatology is required.
- It was objectionable that I had combined a temperature curve and a CO2 curve in the same graph. – Any user of the graph has to figure out her- or himself whether its explicit contents and implicit messages fit the intended use. The two times that I had used the graph myself, I made it exceedingly clear that the “co-variation between CO2 and temperature which is evident [from the graph] does not in itself document any causal relationship. The latter is well documented by means of lab experiments, modelling and comparative field studies, however.” Other users have to answer for their use of the figure.
- It was objectionable that I had combined two different temperature datasets in the same curve. – In fact, this was the only objection which I have seen raised in professional fora. Furthermore, I perceive this criticism to be partly justified: had it been the intention to use the graph in a scientific publication rather than a popular account, the two parts of the curve (one based on proxy data, and one on direct temperature measurements) should have been clearly distinguishable, for instance by means of different colours for each of the data sources. However, I had not drawn the figure with this intention, and, again, other users of it have to answer for their use. While adequate and advisable for a scientific target group, use of different colours within the same curve would have been more confusing than illuminating for a nonscientific target group. Graphs intended for popularisations do not normally convey this degree of technical information.
To offer an admittedly subjective summary, many of the objections were more or less misguided (most of them rather more than less). More importantly, it seems to me that the real object of attack was an entirely different criticism, which however was seldom spelled out:
- It was objectionable that UNEP chose to use a version of the “hockey stick” graph. – My guess is that this was the implicit criticism motivating all the other attacks. It expresses a valid concern, and is, unlike the other objections, founded in a genuinely scientific question, i.e. an issue open to empirical investigation: is the “hockey stick” graph an adequate representation of the global temperature during the past millennium? The best answer I can give consists of several parts:
- The “hockey stick” graph, originally published by Mann and colleagues, has been heavily criticised for its methodology and use of source data. Most of this criticism has been put forth in various online fora, has not undergone peer-review, and is, as such, impossible to evaluate thoroughly. Some of the criticism, however, has been published in peer-reviewed scientific journals, and resulted in a correction by the original authors, stating that “the listing of the ’proxy’ data set in the Supplementary Information published with this Article contained several errors,” but that “None of these errors affect our previously published results.” Because I am a biologist involved in ecological climate impact assessment rather than a climatologist, I have sufficient knowledge of climatology to understand scientific publications within this area, while criticising their methodology or judging their validity is beyond my competence. I am therefore unable to decide who is in error and who (if anyone) isn’t, but think that, based on the relevant literature, it is fair to summarise that Mann et al.’s publication has not been retracted, neither by its authors nor the editor of the journal; that the results have not been refuted conclusively; and that some scientists think the study is flawed, while others don’t.
- A couple of other temperature reconstructions has been published by different authors, using different source data and applying different methodologies. While the results differ in a number of details (as one would expect), the general pattern of all these other reconstructions is similar to the “hockey stick”. Among other features of the “hockey stick”, these later studies have corroborated that the past 2000 years have not experienced any decade with a global temperature equal to or higher than the current one.
- A specific criticism raised against the “hockey stick” was that it “hides” (or doesn’t show) the two climatic phenomena known as Little Ice Age (a cold period around 1700) and Medieval Warm Period (a warm period around the year 1000). However, independent publications have confirmed that these phenomena were restricted to the Northern Hemisphere or the Northern Atlantic. In other words, while there doubtlessly was a Little Ice Age and a Medieval Warm Period, they will not necessarily be visible in reconstructions of global temperatures for the same time period.
- The scientific relevance and the political implications of the “hockey stick” have tended to be greatly overstated in some circles. By attacking the “hockey stick” graph and simultaneously emphasising that the globe has experienced earlier warming periods, it is often implied (a) that the current warming can have natural causes, too, and/or (b) that the human, social and ecological consequences of a warming period are negligible (or even predominantly positive). Such conclusions are logically and empirically invalid, however.
By repeatedly shifting targets and presenting ever new objections, which were not only irrelevant but also weak at best, the campaigners against the Compendium obviously hoped to discredit global warming as such. One could also describe this strategy as attacking global warming by proxy. However, a couple of other techniques was employed, too:
- Circulating lies. – At one stage in the campaign, the public was told that, “in an e-mail to TalkingAboutTheWeather.com[, ...] Sandvik politely distanced himself from the graph as the story unfolded.” Although this statement, too, was regurgitated by 60 other blogs or webpages, it was a plain lie, circulated by a certain Harold Ambler.
This person had sent me an enigmatic e-mail with the following contents (here displayed in its entirety, punctuation marks added): “MWP was real[.] LIA was real[.] nice hockey stick though!” As I found this message far from self-explicatory, I replied that I was “sorry to say that I d[id] not completely understand what [he was] hinting at.” In the following e-mail, Ambler presented a lengthy attack against the “hockey stick”, whereupon I explained that he was addressing it to the wrong person: any disagreement with the underlying data should be discussed with their originators, while my sole involvement in the entire “case” was drawing the graph and making it available to the public via Wikimedia. How this rather trivial clarification of my role in the tragicomedy could be interpreted as me distancing myself from the graph I had drawn, I cannot tell. (Perhaps needless to say, Ambler never presented himself as a journalist for “TalkingAboutTheWeather”. As a matter of fact, he didn’t present himself at all.)
- Naïvety. – Naïvety is no crime. If it is the basis for violent attacks, however, one can but wonder about the (absence of) personal judgment involved. For example, one of the anti-Compendium campaigners discovered that I had uploaded another paleoclimatic graph to Wikimedia, a graph that was based on a data set with a different geographic coverage and therefore did show the European Medieval Warm Period. Other things being equal, one might have expected that this made even climate change deniers understand that the main motivation for uploading images to Wikimedia is making knowledge available rather than censoring it. However, this was not the conclusion they drew. Instead, they summed up: “apparently ’Hanno’ contradicts himself with his own set of artwork”.
- Double standards. – Climate change deniers have repeatedly criticised climatologists for missing openness and/or truthfulness. This could also be seen in the “Hanno graph” web debates. At the same time, however, the very same blogs continued to spread lies both about science in general, about climatology in specific, about certain climatologists, and about me (see above). Likewise, the “journalistic methods” of some of the initiators of the campaign did not stand up to any standard of open- or truthfulness, neither scientific nor journalistic.
Another, almost funny, example of double standards at “wattsupwiththat” were the repeated attacks against the alleged anonymity of the ominous “Hanno” – presented by anonymous bloggers, and targeted at a Wikimedia user who openly had put a link to his real-world identity on his very user page.
- Attacking from two angles. – Climate change deniers often set up strange but effective alliances between two subgroups, one of which exploits the fact that scientific prose is often hard to understand for laypeople and thus easy to distort, while the other subgroup attacks every sign of popularisation as unscientific. The battle against this coalition between (real or alleged) naïvety and deliberate misrepresentation is almost impossible to win for science. If one presents the evidence in its overwhelming massiveness, a layperson has no chance of understanding it and will effectively ignore it. If one popularises the findings, however, one is an easy prey for scientifically educated but politically motivated people who see their chance to discredit climatology by attacking these simplifications.
For me, the “Hanno 2009” fuss was, at least after I had got over the initial sleepless nights, a rather interesting experience. I can even see some positive signs in it, viz. that climate change deniers become increasingly desperate. If the embarrassingly weak and far-fetched objections were the best they could come up with, it might not be impossible to get through with the science-based messages after all.
Notes and references
- ^ A search for "Hanno graph"+UNEP yields c. 111 hits, "Hanno 2009"+UNEP c. 315 hits (from which one has to subtract the c. 29 hits with both phrases; all counts as of September 2010).
- ^ Climate change deniers should not be referred to as “climate sceptics”. Scepticism is a philosophical school of thought valuing doubt, practicing academic openness and demanding convincing evidence. This makes any good scientist a sceptic (and, if you wish, any good climatologist a climate sceptic). Many climate change deniers, on the other hand, “invoke conspiracies, quote fake experts, deploy evidence selectively and create impossible expectations of what science can deliver” (Kemp, J., Milne, R., and Reay, D.S., 2010, “Sceptics and deniers of climate change not to be confused,” Nature, 464, 673; cf. McCright, A.M., 2007, “Dealing with climate change contrarians,” pp. 200–212 in Creating a climate for change, edited by S.C. Moser and L. Dilling, Cambridge University Press) and are, thus, “climate dogmatists” or “climate ignoramuses”, but certainly not “climate sceptics”.
- ^ For the years 1000–1880, temperature estimates were taken from P.D. Jones & M.E. Mann (2004): “Climate over past millenia”. Reviews of Geophysics, 42, article number RG2002. For the remainder, temperatures are based on instrumental records published on the web by P.D. Jones, D.E. Parker, T.J. Osborn & K.R. Briffa (2005) as “Global and hemispheric temperature anomalies – land and marine instrumental records”. In Trends: A Compendium of Data on Global Change. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, Tenn., U.S.A. [http://cdiac.esd.ornl.gov/trends/temp/jonescru/jones.html]
CO2 levels are based on historical carbon dioxide records from ice cores drilled at the Law Dome in Antarctica, published on the web by D.M. Etheridge, L.P. Steele, R.L. Langenfelds & R.J. Francey (1998) as “Historical CO2 records from the Law Dome DE08, DE08-2, and DSS ice cores”. In Trends: A Compendium of Data on Global Change. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, Tenn., U.S.A. [http://cdiac.esd.ornl.gov/trends/co2/lawdome.html]
[NB: These references are copied and pasted directly from the file description pages at Wikimedia, i.e. they have been clearly indicated together with the respective images since the latter had been uploaded in 2005.]
- ^ My personal favourite was the argument – based on an automatic Google translation from Norwegian to English of my Wikipedia user page! – that “my” English mistakes indicated a German origin. Speak about arriving at the correct conclusion based on a false premise!
- ^ obviously the time it took the “wattsupwiththat” bloggers-detectives to follow the link from my Wikipedia user page to my university home page
- ^ in an entry in Norwegian Wikipedia and an article in a Norwegian popular scientific magazine
- ^ I may add that, far from hiding the fact that the two parts of the curve have different sources, I had added the following caveat to the file description page of the temperature curve in 2005: “Caution is required in interpretation, however, because the data prior to 1881 are estimates based on proxies (such as tree rings and ice cores). This makes both the accuracy and the temporal resolution of the estimates poorer than for the last century.”
- ^ Consider, for instances, graphs about human population growth (such as this one), which do not normally tell the user that the most recent 50 years of the figure are built upon censuses, the 100 years before that on extrapolations, and the 10,000 years before that on proxies, anecdotal evidence and qualified guesses.
- ^ Mann, M.E., Bradley, R.S., and Hughes, M.K. (1998) “Global-scale temperature patterns and climate forcing over the past six centuries,” Nature (London), 392, 779–787
- ^ McIntyre, S., and McKitrick, R. (2005) “The M&M critique of the MBH98 Northern Hemisphere climate index: update and implications,” Energy and Environment (Brentwood), 16, 69–100
- ^ Mann, M.E., Bradley, R.S., and Hughes, M.K. (2004) “Corrigendum: Global-scale temperature patterns and climate forcing over the past six centuries,” Nature (London), 430, 105
- ^ Interested readers may consult the following figures, which compare different reconstructions and provide their respective sources: Fig 1, Fig 2
- ^ A summary of the relevant evidence is provided in Box 6.4 of IPCC’s Fourth Assessment Report
- ^ Norwegian readers may consult two of my popular articles for documentation (2008a, 2008b, and references therein). The short version is as follows:
- (a) Global climate can warm for many different, non-exclusive reasons. Earlier warming periods, of which there have been many through the eons, had necessarily natural causes. It does not follow that also the current global warming can be natural, however. To the contrary, there is unanimity among climatologists that natural causes, although they do play an important role, are not sufficient to account for the current warming. Should the “hockey stick” prove to be faulty and the underlying publication be retracted, this would have serious consequences for a couple of individual paleoclimatologists, but it would not have the slightest effect on the scientific certainty that global warming is (i) happening and (ii) to a large degree caused by human emissions of greenhouse gases. Both these conclusions are bolstered by a multitude of robust findings from a number of scientific disciplines. Among other things, the amount of human greenhouse gas emissions and their effect on the atmosphere are well-documented entities (and have been so for many years). The confidence in these empirical estimates is entirely independent of how warm the globe has been 500 or 1000 years ago.
- (b) Even assuming that the globe was warmer during the Viking age, this would not imply that the current warming is harmless. This conclusion might, at best, be justified other things being equal. However, other things are far from equal. To mention a few things that have changed dramatically, the current warming is happening at an unprecedented speed (leaving many species of plants and animals no time to adapt or migrate); it will continue at this speed for several decades (assuming, that is, the world community will be able to drastically and rapidly reduce greenhouse gas emissions – otherwise the speed of the warming will increase further); the Earth is inhabited at least 20 times more densely now; the ethical and legal frame conditions have changed. The last two points merit a short explanation: in the Viking age it was quite possible to leave an area that had become uninhabitable and settle or conquer another area. It was no exception during such migrations that the migrating or the conquered people became victims to genocide. Today, there are no uninhabited and inhabitable areas to move to, and genocide is no longer regarded as a legitimate tool for solving territorial conflicts.
- ^ Chapter 2 of the Intergovernmental Panel of Climate Change’s Fourth Assessment Report explains atmospheric forcing and provides the relevant references, chapter 1 gives a short historical account. See also “my” graph CO2-src.png (and the references cited on the file description page!) for the history of anthropogenic CO2 emissions.
- ^ http://wattsupwiththat.com/2009/09/26/united-nations-environment-programme-uses-unreviewed-graph-from-an-anonymous-wikipedia-author-for-official-report, and echoed verbatim in a handful other blogs
- ^ Gleick, P.H., et al. (2010) Climate change and the integrity of science. Science (Washington, D.C.), 328, 689–690.
- ^ A simple example is the use of the word proof. Scientists know that they cannot literally prove anything. Proof is possible in mathematics and other formal sciences, but is beyond the scope of empirical sciences. If pressed to answer whether global warming is proven, a climatologists can chose between answering “no” (which will be exploited by the naïve subgroup of climate change deniers to claim that there is no need for action) or “yes” (which will be exploited by the scientifically educated subgroup of climate change deniers to claim that climatologists are dishonest and do not openly communicate uncertainties). The correct answer, being substantially longer than “yes” or “no” and, therefore, unattractive to the media, has a good chance of being ignored by anyone who is not prepared to listen anyway.
[back / tilbake]