Monday, August 16, 2010

R.I.P., Hockey Stick

If there is a single icon for 'global warming,' it is the "Hockey Stick" (HS) created by Dr. Michael Mann.
The HS was featured on the cover of an IPCC report and was featured in Al Gore's movie, An Inconvenient Truth. Not only did the HS purport to show that recent temperatures are unprecedented but that the well-known Medieval Warm Period (MWP) didn't exist (see graph below) which, until the HS, was accepted by most meteorologists and climatologists.
From almost day one, the HS has been controversial. Steve McIntyre and Ross McKitrick published two peer-reviewed papers calling the HS into question. McKitrick commented,

"The Mann multiproxy data, when correctly handled, shows the 20th century climate to be unexceptional compared to earlier centuries."

In other words, the temperatures we are experiencing now are not higher than those of 900 years ago! 

If today's temperature levels have been previously reached naturally, before humans started driving up atmospheric concentrations of CO2, then the certainty that today's temperatures are due to CO2 goes away. It also tends to falsify the IPCC's hypothesis that CO2 is the driving force behind changes in climate. 

As you can imagine, this finding created a firestorm in the climate 'science' community. I put science in scare quotes in this instance because of the behind-the-scenes efforts (revealed in the Climategate emails) to keep McIntyre and McKitrick's work from being published  -- efforts which were the antithesis of science.

The firestorm rose to the level that Congress ordered an independent investigation via the National Research Council's Board on Atmospheric Science and Climate. Their report concluded there were errors in Mann's techniques and that there was low confidence in the temperature reconstruction from years 900 to 1,600 AD but found that Mann's basic work was correct. However, a major criticism of Mann's work remained unaddressed -- that the statistical techniques themselves used by Mann were not sufficiently rigorous.

In 2006, a team of university statisticians was created to review the HS at the request of Rep. Joe Barton and Rep. Ed Whitfield. The team was led by Dr. Edward Wegman, chair of the National Academy of Sciences' Committee on Applied and Theoretical Statistics. Some findings:

  • [Mann's papers] were found to be "somewhat obscure and incomplete" and the criticisms by McIntyre and McKitrick were found to be "valid and compelling."
  • The paleoclimate community is relatively isolated; its members rely heavily on statistical methods but do not seem to interact with the statistical community. Sharing of research materials, data, and results was done haphazardly and begrudgingly.
  • Overall, the committee believed that Mann’s assessments, that the decade of the 1990s was the hottest decade of the millennium and that 1998 was the hottest year of the millennium, cannot be supported by his analysis.

This wasn't all that was wrong with the HS. Over time, it became known that a "trick" was used to create the famous HS graph because the data used to estimate temperatures from years 1000 to 1950 showed falling (declining) temperatures after 1950 when thermometer measurements showed rising temperatures. This "divergence" in the data would -- even to the most unseasoned novice -- call the entire HS into question. So, they had to, in the words of the Climategate emails, "hide the decline."
The decline was hidden by truncating the data (all lines above except red) when the divergence in the HS data begins. The extreme zoom above reveals the technique.

You might think that all of this was more than enough to call the Hockey Stick into serious question. Yet, many in climate science brushed aside these numerous troubling aspects of the HS because of its importance to their crusade against CO2.

This brings us to the present. A new paper peer-reviewed paper is going to be published that subjects Mann's work to the long-overdue rigorous statistical examination.

The next issue of The Annals of Applied Statistics will publish a paper by Blakeley McShane and Abraham Wyner titled, "A Statistical Analysis of Multiple Temperature Proxies: Are Reconstructions of Surface Temperatures Over the Last 1000 Years Reliable?" (scroll down to last paper, click on the title to download the entire paper).
I have done a screen capture of the conclusions (above, click to enlarge) so our blog readers can read it for themselves.  "Our methods of estimating model ... accuracy are in sharp disagreement [with the HS]. ... we conclude unequivocally that the evidence for a 'long-handled' hockey stick (where the shaft of the hockey stick extends to the year 1000 AD) is lacking in the data."

Their paper indicates that there were not only warm temperatures 1000 years ago (the Medieval Warm Period) but that they were approximately the same as today's. The paper does confirm that there has been a relatively rapid runup from about 1850 to present (graph above reproduced from the paper, click to enlarge) but the "uncertainty bars" (cyan colored) indicate that the runup may not be as great as indicated due to problems with current temperature measurements.

Since this is a peer-reviewed paper, the preponderance of the evidence (i.e., this paper combined with McIntyre and McKitricks') is that the hockey stick is invalid.

What does all of this mean? My interpretation is:
  • There is no immediate global warming crisis. Certainly nothing that requires us to immediately turn the world's economy upside down and to subsidize inefficient energy sources (i.e., wind energy). 
  • As I have previously written, there is an urgent need to create an accurate record of earth's temperatures for the last 150 years and there is an urgent need to get a handle on ocean heat content. 
  • We are running a chemistry experiment pumping high levels of CO2 into the atmosphere and the results may be, on balance, unfavorable. We don't know.  Smart decarbonization makes sense while we sort out the answers.
  • We will likely know in 3 to 5 years whether the IPCC's hypothesis is correct. That is soon enough that we can wait to make these critical decisions whether widespread decarbonization is necessary.
  • Tom Fuller has an excellent proposal as to where we go from here as far as the science is concerned.
I suspect you will not be hearing much about this in the media because it is technical and it goes against the pro-GW narrative that dominates their coverage of this issue. Nevertheless, the McShane and Wyner paper is an important contribution to our knowledge of climate evolution.

UPDATE:  Statistician Matt Briggs agrees

1 comment:

  1. Brief yet concise, the best HS analysis I've read. Why isn't this blog more popular?