Another attack on Michael Mann, but not because of some error he’s made in his science, but because he’s tired of being the target of bullshit attacks.

Michael Mann, the Father of the Hockey Stick graph — you know, the one where they had to “hide the decline” — seems to have a predilection for making implausible claims. Mann has sued the Competitive Enterprise Institute and Mark Steyn for libel, under the legal theory that he should not be subject to “public figure” rules, because he was “an involuntary public figure.” He explains the tragedy of his becoming a “target” in an essay published in The Scientist.

Aside from complaining that Mann is suing or libel, the author of this blog entry can’t even get the facts of a debunked denialist meme correct. He apparently still believes the ‘decline’ was of temperatures and used in the ‘Hockey stick graph’ even though it has been shown several times by different people that neither point is correct.

The ‘decline’ mentioned in the email was of tree ring densities. Tree rings are one of several proxies used to determine historical temperatures where we have no direct temperature information. They do not provide direct temperatures, the temperatures are derived by the change in density of the rings over time. Although the tree ring densities have tracked thermometer temperatures accurately  from the late 1880s, since about the 1960s the tree ring densities have correlated to falling temperatures. This is because additional atmospheric CO2 has changed the conditions the trees the tree rings have been taken from grow in, so we can no longer be sure the densities are only affected by temperatures.  this is called the ‘divergence problem’ and a journal article on it was was published in 1998.

The ‘trick’ Mann used was to add in instrumental temperatures to a graph he used in a 1999 paper.

Hockey Stick

Northern Hemisphere mean temperature anomaly in °C (Mann et al 1999).

The blue line is the reconstructed proxy data and the red is the instrumental data. Unless the proxy data is more reliable than temperatures taken from thermometers, then the most accurate data is in red. If you notice the grey area on the graph that is the uncertainty in the data. Not much uncertainty in the instrumental data but lots in the proxy data.

There is no ‘decline’ to hide.

Here is another graph where the ‘trick’ was used. It was in a paper from 2003. Mann et al (EOS 2003)


And a closeup.


The red line are actual temps taken from thermometers. All the others are reconstructions derived from different proxies. Only one line is truncated, that using tree rings. It is truncated because of the published and discussed divergence problem.

Replacing a questionable proxy time period with accurate instrumental temperatures, in the presence of 7 other closely matched reconstructions that are not truncated is hardly dishonest. Especially considering a published paper recommended exactly that.