False statement about paleoclimate studies

"Paleoclimatology" refers to the process of measuring so-called "proxies", such as  tree rings, making the unjustified assumption that they represent temperature, and carrying out an incorrect statistical analysis of the results (Mann, Bradley and Hughes, Nature 392, 779-787, 1998).

Page 9 of WG1 SPM contains the astonishing claim that:

"Studies since the TAR draw increased confidence from additional data showing coherent behaviour across multiple indicators in different parts of the world".

Let's look first at the "coherent behaviour". The relevant figure is Figure 6.10 in chapter 6 of WGI:

Do these different reconstructions show a coherent behaviour? Ironically, the only coherent part is the prominent medieval warm period at 1000 AD, which the IPCC then attempts to write out of the history books on the following page (box 6.4 page 468)! (Note that the reason all the curves agree in the late twentieth century is that they are forced to, by adjusting the baseline). 

Now for the amazing statement that confidence in such studies has increased since the TAR.

Firstly, the IPCC has implicitly acknowledged that this is not the case: in the TAR, the infamous and discredited 'hockey stick' picture from MBH98 was prominently displayed in the SPM, but in the AR4 SPM it does not appear.  The above diagram from chapter 6 does not resemble the hockey stick. If you ignore the misleading heavy black line of the instrumental record, the 20th century temperature rise in the proxy data is not unusual or unprecedented.

Secondly, several research papers and reports have criticized such paleoclimate reconstructions:

S. McIntyre and R. McKitrick,   Corrections to the Mann et al. (1998) proxy database and northern hemispheric average temperature series. Energy and Environment, 14, 751-771 (2003): They list numerous errors in MBH98 and show that when these are corrected, temperatures in the early 15th century exceed those of the 20th century.

S. McIntyre and R. McKitrick, Hockey sticks, principal components, and spurious significance. Geophys. Res. Lett., 32(3),
L03710 (2005): They showed that the MBH98 method can produce a 'hockey stick' picture even when fed with data in the form of random noise.

H. Von Storch et al, Reconstructing Past Climate from Noisy Data, Science 306,  679 - 682 (2004): They found that use of proxy data could underestimate past temperature variations by a factor of 2.

A National Research Council Report "Surface Temperature Reconstructions for the Last 2,000 Years",  was published in 2006. This report reached the somewhat ambiguous conclusions that:

  • It can be said with a high level of confidence that global mean surface temperature was higher during the last few decades of the 20th century than during any comparable period during the preceding four centuries.
  •  Less confidence can be placed in large-scale surface temperature reconstructions for the period from A.D. 900 to 1600. 
  •  Very little confidence can be assigned to statements concerning the hemispheric mean or global mean surface temperature prior to about A.D. 900

Unfortunately it is not clear what is meant by "less confidence", suggesting that there was no agreement among the panel.

A team of expert statisticians investigated this issue (the Wegman report). They stated that  "In general, we found MBH98 and MBH99 to be somewhat obscure and incomplete and the criticisms of MM03/05a/05b to be valid and compelling". They confirmed the results of MM05 that the 'hockey stick' could be generated by the MBH method even when the input data was random noise.


There are many good summaries of how these paleoclimate reconstructions have been completely discredited since the TAR (quite the opposite of the AR4 claim). See for example those by Ross McKittrickSteve McIntyre or Bishop Hill.  See also the Climate Audit post on this misleading statement in the SPM.