Blog Archive

Tuesday, August 17, 2010

The Curious Case of the Hockey Stick that didn’t disappear. Part 1: The Police Lineup But who killed the Medieval Warm Period?


The Curious Case of the Hockey Stick that didn’t disappear. Part 1: The Police Lineup

But who killed the Medieval Warm Period?



by Joseph Romm, Climate Progress, August 16, 2010
Before we begin the investigation into the usual suspects, some background for people who those who don’t follow climate science closely, which certainly includes most of the disinformers and  apparently at least two statisticians.
  1. There is a high probability that the recent warming is unprecedented for 1000 years and probably much longer (see “Sorry disinformers, hockey stick gets longer, stronger: Earth hotter now than in past 2,000 years“ and here and here).
  2. This conclusion is based on an analysis of multiple proxies for temperature, which individually engender much uncertainty and collectively still engender a fair amount. It is a canard of Curry-esque proportions to assert that scientists have not clearly explained the nature and extent of these uncertainties. They have bent over backwards to do so.
  3. The temperature trend in the past millennium prior to about 1850 is well explained in the scientific literature as primarily due to changes in the solar forcing along with the effect of volcanoes, whereas the recent rise in temperature has been driven primarily — if not almost entirely — by human activity (see Scientist: “Our conclusions were misinterpreted” by Inhofe, CO2 — but not the sun — “is significantly correlated” with temperature since 1850 and Part 3 [to come]).
  4. Absent human emissions, we’d probably be in a slow long-term cooling trend due primarily by changes in the Earth’s orbit — see Human-caused Arctic warming overtakes 2,000 years of natural cooling, “seminal” study finds.
  5. Thus, the rate of human-driven warming in the last century has exceeded the rate of the underlying natural trend by more than a factor of 10, possibly much more.  And warming this century  on our current path of unrestricted greenhouse gas emissions is projected to cause a rate of warming that is another factor of 5 or more greater than that of the last century.  We are punching the climate beast — and she ain’t happy about it!
Back to the investigation of attempted murder — and the ‘innocent victim’ who may have been killed in the attempt.  The folks who don’t follow climate science closely have been trumpeting a new paper “A Statistical Analysis of Multiple Temperature Proxies: Are Reconstructions of Surface Temperatures Over the Last 1000 Years Reliable?” by McShane and Wyner about to be published in Annals of Applied Statistics.   Supposedly it is fatal to the Hockey Stick.

Here is the police lineup.  Take a look at three independent reconstructions of the past one to two  millennia and the new one by the statisticians — and see if you can pick out which one allegedly killed the others (with apologies to Deltoid):


mann1.jpg


Lake Tanganyika lake surface temperature

figure
mcshanewynerhockey.png

Did you figure out who the alleged killer was?  Yes it is the final reconstruction (of Northern Hemisphere temperatures) that is supposedly our Keyser Söze.  The helpful graphic overlay is by Deltoid from his post, “A new Hockey Stick: McShane and Wyner 2010” [the comments contain many gems].

While WattsUpWithThat thinks this paper is so important that he has been running a post on it at the top of his blog for days, he conveniently omits this rather remarkable statement from the authors:
Using our model, we calculate that there is a 36% posterior probability that 1998 was the warmest year over the past thousand. If we consider rolling decades, 1997-2006 is the warmest on record; our model gives an 80% chance that it was the warmest in the past thousand years.
Doh!

The judge has no choice but to dismiss the primary murder charge because the Hockey Stick still lives.

But there appears to have been some collateral damage.  The new model seems to have done away with that favorite of the anti-science crowd, the Medieval Warm Period, which I suppose could be accidental homicide, except that would have required the victim to have existed in the first place, which is at best an open question.

Indeed, this new reconstruction shares a similar form to the one immediately above it, of the Arctic, from the 2009 Science paper, “Recent Warming Reverses Long-Term Arctic Cooling.”  So let me excerpt some of the  analysis and commentary surrounding that paper:
Arctic temperatures in the 1990s reached their warmest level of any decade in at least 2,000 years, new research indicates. The study, which incorporates geologic records and computer simulations, provides new evidence that the Arctic would be cooling if not for greenhouse gas emissions that are overpowering natural climate patterns.
So reports the National Center for Atmospheric Research (NCAR), which coauthored the study.  The Washington Post story notes:
The analysis, based on more than a dozen lake sediment cores as well as glacier ice and tree ring records from the Arctic, provides one of the broadest pictures to date of how industrial emissions have shifted the Arctic’s long-standing natural climate patterns… 
It’s basically saying the greenhouse gas emissions are overwhelming the system,” said David Schneider, a visiting scientist at the National Center for Atmospheric Research and one of the Science article’s co-authors.
The same could be said about the entire planetary ecosystem — on our current path, we’re going to overwhelm the whole system (see “Intro to global warming impacts: Hell and High Water“).  Indeed, in some sense we already have, as a number of climate scientists have pointed out.  The NYT’s Andy Revkin interviewed Thomas Crowley, a climate specialist at the University of Edinburgh:

“I would say that this is another piece of evidence that strengthens the argument that humans are now capable of preventing the onset of a future ice age,” he told me. Another scientist holding this view is James E. Hansen of NASA, whom I interviewed about the timing of the next ice age in 2003.
The Arctic study suggests (again) that the Medieval warm period was limited to only a part of the Northern Hemisphere, and that recent human-caused warming is quite outside the boundary of the last two millennia:
Darrell Kaufman of Northern Arizona University, the lead author and head of the synthesis project, says the results indicate that recent warming is more anomalous than previously documented. 
“Scientists have known for a while that the current period of warming was preceded by a long-term cooling trend,” says Kaufman. “But our reconstruction quantifies the cooling with greater certainty than before.”
Some of our leading climate scientists say this Arctic reconstruction is an especially important paper, as the WP piece notes:
Mark Serreze, director of the National Snow and Ice Data Center at the University of Colorado at Boulder, said the study was significant because it helps confirm scientists’ current understanding of how the earth’s climate has changed over millennia. 
“It’s not that we don’t know how the climate works, it just we didn’t have anyone at that time measuring the climate forcing then,” referring to 2,000 years ago. “Climate doesn’t change all by itself for no good reason. Something has to force it.”
Precisely.  I will return to this point in Part 3.
Robert Correll, who chairs the Arctic Climate Impact Assessment, said the paper in Science will likely “in the long haul become a seminal piece in the scientific literature” because it allows other climate researchers “to set their work in a long time scale.”
And Revkin’s print piece underscores the danger:
Jonathan T. Overpeck, a study author and climate specialist at the University of Arizona, said the rising concentration of long-lived greenhouse gases guaranteed warming at a pace that could stress ecosystems and cause rapid melting of Greenland’s great ice sheet.
The fast rate of recent warming is the scary part,” Dr. Overpeck said. “It means that major impacts on Arctic ecosystems and global sea level might not be that far off unless we act fast to slow global warming.”
Bottom line:  It appears to remain the case that, with the appropriate caveats, there is a high probability the recent warming is unprecedented for 1000 years and probably much longer — and that the  current rate of warming is unprecedented in the recent paleoclimate record.

In Part 2,  I am planning to reprint a forthcoming forensic debunking of statisticians’ paper.  Feel free to post any critiques you have in the comments section.  It does not look like a terribly strong paper given how much it cites Wegman and McIntyre (talk about your ‘usual suspects’) – and the fact that it cites Al Gore’s movie  and three Wall Street Journal pieces,  which must be a first for that journal (and probably any other publishing pieces on paleoclimate reconstructions)!


Link:  http://climateprogress.org/2010/08/16/hockey-stick-paper-mcshane-and-wyner-statisticians/#more-31767

2 comments:

ccpo said...

I don't know why this never occurred to me before, but I may have a thought on why climate sensitivity calculations seems too low. When they say that anthropogenic emissions are overcoming natural cooling, I don't think I've ever read of anyone factoring the cooling that didn't happen. That is, if we continue the trend of the cooling that would likely be taking place if CO2 emissions had not interrupted it out from 1850-ish to present day, there's another .25 to .33 degrees that our emissions have countered.

So actual warming should count the absolute rise from 1850 to now AND the .25 to .33 cooling that was overcome. If orbital changes would have caused that much cooling, then that cooling had to have been compensated by addition of heat. Thus, warning since 1850 is not in .8C range, but is more like 1.05 - 1.15C.

It's such a simple thing, it must be factored in, right?

Tenney Naumer said...

I don't know what the calculations are based on -- Hansen had a higher number than was generally accepted for many years, but some are working closer to his numbers now. Might want to check his papers on this via his website.

But, for some reason, I'm thinking they do not use the item you mention - I think I remember that it is more related to the physics.

I mean, I think they would take out the orbital effects, as well as something like global dimming, if they went that route.

I am in way over my head here, so forgive any obvious ignorance in the above.