the Air Vent

Because the world needs another opinion

The Hockey Stick Data Hoax

Posted by Jeff Id on October 3, 2008

I said it again.  Hoax!  The more I learn about hockey stick graphs, the worse it is.

The graph below is from a very important group of data in the M08 latest hockey stick.   This represents an average of 105 of the 1209 data sets used in  M08 or 8.6%.  But wait, the “CALIBRATION” process eliminates bad correlation data series so only 484 of the 1209 were kept in the final calculation.  Of the 484 series which were kept 95 of them were from this very same set of proxies. Mann08 used 95 of these series in the final 484 or 20% of his data.

Proxies are not known to be temperature.  The process of calibration is used to determine if proxies are temperature by pattern matching to measured temperature in the last 100 years.  Well I like trees but they are lousy thermometers. There is a buzzword in the field called DIVERGENCE, everything in climatology is so politicized they can’t call a turd a turd.  Divergence means “doesn’t really match temperature”.  Well guess what 20% of the data right off the top doesn’t match temperature very well so our brilliant scientists who are saving us from CO2 plant food, did the honorable thing and CHOPPED THE DATA THEY DIDN’T LIKE RIGHT OFF THE END!

The last 38 inconvenient years of data for these series were truncated, chopped, amputated, excised and cut from the data.

This is not the end.  Mann08 feeling pretty cocky after pushing the other magic papers through the IPCC is emboldened in the face of the publicity they actually PASTED FAKE DATA ON THE END OF THE SERIES.

The new fake data has a nice smooth and not noisy temperature rise, they then compare the fake data to the temperature value and with great fanfare determined that over 90% of the measurement of tree hardwood series are actually TEMPERATURE.

Well Steve McIntyre has just yesterday after waiting for years (i believe) brought about the release of the REAL data to the public.  We didn’t know what the REAL data would look like but those who are suspicious of huge government funded studies that don’t show their calculations HAD A PRETTY GOOD IDEA WHY WE COULDN’T SEE IT.

Guess what, we weren’t disappointed.

The data to the right of the yellow line is the modified data in question. The light purple line is the real data, data which was easily available at the authors of M08 request but they didn’t want it.  The blue to the right side of the yellow line is the modified fake bull crap data M08 ACTUALLY USED.

What’s more, it was well known that these series had a divergence (ain’t really temperature) problem.  Here is my favorite section in the whole paper.

Because of the evidence for loss of temperature sensitivity after 1960 (1), MXD data were eliminated for the post-1960 interval. The RegEM algorithm of Schneider (9) was used to estimate missing values for proxy series terminating before the 1995 calibration interval endpoint, based
on their mutual covariance with the other available proxy data over the full 1850–1995 calibration interval.

This means exactly what I have been explaining, “loss of sensitivity” means “divergence” which means “ain’t really temperature”

Rather than use the real, available, easy for them to get data.  Data which was not easy for guys like McIntyre who don’t support the hockey sticks.  They used chopped off data which they added FAKE and KNOWN TO BE FALSE data on the end.

In the paragraph above when they use the term “loss of sensitivity” they are admitting they KNOW IT ISN’T TEMPERATURE. They know exactly what they are doing, if this were a criminal trial they would be going to jail.

Another point I have to make is that this end section is critical in setting the magnification values to determine how many degrees each fake tree unit represents.  A flatter set of data results in higher amplification to  match temperature, a steeper set of data results in lower magnification to match temperature.

Finally, 20% of the data series in the picture below is COMPLETELY FABRICATED! You have to look through the pretty colored red lines to find it but it is in there.

What a joke this science is.  We all owe a big thanks to Climate Audit for getting to the real data.

I need an asprin.


25 Responses to “The Hockey Stick Data Hoax”

  1. PaulM said

    Jeff, when you have had your aspirin and calmed down a little, you also need a spellchecker. Also, note that the paper you are referring to has 7 authors, so you should really refer to it as Mann et al 08 and say ‘they’ rather than ‘he’.
    The way they draw the thick red and gray lines over the data to hide the divergence problem is just amazing – why does the journal let them et away with that?

  2. Jeff Id said

    Thanks PaulM, The changes were made. I get a bit wound up if you can’t tell.

    What a joke, how many billions of dollars are we going to spend on this nonsense.

  3. Eric Anderson said

    Jeff, great site and keep up the great work. Just one minor suggestion: Although it is hard to not blow a top when one sees what is going on, I think a little restraint in allegations and attribution of motives might be in order. “Talk soft, but carry a big stick,” so to speak. Might give your message a more mainstream audience.

    Thanks again for all your efforts.

  4. Jeff Id said

    Eric,

    I feel like there is far too much understatement out there. This to me is intent.

    Climate audit and others can keep their nice image by understatement but someone needs to call a turd a turd.

  5. Matt Y. said

    Hoax is a good word. I know some of the more legal savvy skeptics are uncomfortable with the word “fraud”, but if it looks, acts, and quacks like a duck?

  6. chopbox said

    I notice, Jeff, that you are also like politics. It seems to me that you are taking on a role with this blog that is very similar to the attack dog role of the vice-presidential candidates in the US: that is, that you are playing Biden (or Palin) to Steve’s Obama (or McCain). The only problem is that even if you and Steve get some bite in the blogosphere, the “voting” actually takes place in the academic journals, and Steve, for all his prodigious output, doesn’t seem interested in writing it up there.

  7. vivendi said

    If all this is true, I wonder how much the time they had to spend to “fabricate” their spaghettis and whether they have removed all the mirrors from their home.

  8. Clark said

    well, they got away with simply truncation the data they didn’t like in previous papers. So why not go one more step and fake some new data to replace what they didn’t like. Science is all about progress!!!

  9. Raven said

    Jeff,

    The Team’s counter argument will always be that the “errors” don’t affect the outcome so it is not enough to point out the blatent errors – someone has to show that the results could not be obtained without the errors. The trouble is this paper has many dodgey series and eliminating one of them probably will not affect the results because the other will pick up the slack.

    IOW – debunking this paper will require that it be re-done after all of the dodgy series are identified and removed.

  10. Jeff Id said

    Raven,

    I should have an estimate of the net effect on the outcome soon also.

  11. Jeff Id said

    I just went back to the Tamino open mind blog just to see what those guys said when I told them about the infilling of data as a replacement. (open thread #6)

    Some idiot called Gavin’s cat who pretends to be a statistical genius, wrote in and demonstrated in some way that convinced himself that it’s ok to fill in the most important part of the proxy for correlation. I had some other guy say that it would take a huge study to find out if it was ok or not. Several others said I’m the crackpot so why even look. And a couple more suggested I bring it up with Mann et everyone in paleoclimatology.

    How can people be so dumb? I’ll never figure it out.

    It’s fake data, there’s no study required..Its fake. Jesus.

  12. Chris H said

    It does seem bizarre. If they dislike the data, then I *might* understand them removing it, and doing the analysis without it. But create fake data to replace the data they disliked – it’s just plain crazy.

    I assume that doing an analysis with sections of data missing makes it much harder, but I still don’t see that as a valid justification.

  13. Jeff Id said

    One of the main premises of the paper is that entire data set would be considered statistically random if less 13% was accepted (due to the very loose correlation limits). There were 71 Luterbacher series which were instrumental data and are not really proxies. These represent 95 other series which are faked for correlation.

    But don’t forget, over 90% of the 1209 series were infilled to some extent in the calibration period. These MXD proxies were just truncated before infilling because of known correlation problems.

    So what this paper really is is a group of people pasted non random data on the end of unknown sometimes ridiculous carefully chosen series and correlating to temp.

  14. Gary said

    Actually Jeff, I like this statement better from Mann 08 on divergence
    ‘Interestingly,
    although the elimination of all tree-ring data from the proxy dataset
    yields a substantially smaller divergence bias, it does not eliminate
    the problem altogether (Fig. 2B). This latter finding suggests that
    the divergence problem is not limited purely to tree-ring data, but
    instead may extend to other proxy records. Interestingly, the
    problem is greatly diminished (although not absent—particularly in
    the older networks where a decline is observed after 1980) with
    the EIV method, whether or not tree-ring data are used (Fig. 2 C
    andD).Weinterpret this finding as consistent with the ability of the
    EIV approach to make use of nonlocal and non-temperaturerelated
    proxy information in calibrating large-scale mean temperature
    changes, thereby avoiding reliance on pure temperature
    proxies that may exhibit a low-biased sensitivity to recent temperature
    change”
    My interpretation is that either the proxies are wrong or the surface station data is wrong. It’s not just the trees. And then if you apply the EIV manipulation you can change the proxies from a negative change over time to a positive change over time??? A red flag should have gone up here for the authors and reviewers.

  15. Jeff,
    “Well I like trees but they are lousy thermometers.” I think that comment deserves repeating (maybe several thousand times). The growth rate of anything, including trees, is not linear with temperature; a chart of growth rate vs temperature is close to an inverted parabola. The growth rate increases with temperature until it reaches an optimum range, then decreases when the temperature is too high. Why don’t we shout it: “WELL, I LIKE TREES BUT THEY ARE LOUSY THERMOMETERS.”

    Question: Is the effect of statistical detrending essentially drawing a trend line through a range and calculating the difference (plus/minus) from the trend line at each measured point. If the trend line has a slope, the result would flatten out any bulges. An example would be a trend line from about 900 A.D. through about 1850 A.D. This trend line would have a slight negative slope and taking the differences at each point would virtually eliminate the MWP and the LIA. Would this be a reasonable description for a non-statistician?

    I’m still working on the paper I mentioned.

  16. Geoffrey James said

    Over the years Scientists have found many people and companies to be fakes. A couple of recent Noble Prise winners found the hard part was telling everyone that a bacteria not stress or diet caused stomach ulcers, about 10 years from confirmed discovery to accepted information and an antibiotic cure. I wont say the drug companies didn’t like the idea that 7 antibiotic tablets, could cure the problem and remove a lifetime of medications such as antacids, protein pump inhibitors, tranquilisers such as Diazepam and Xanax, antidepressants through to deep abdominal surgery.
    Again we know the data is less than useful, the math done to reach the conclusion that the Earth is getting hotter because of the ‘green house effect’ would have had my degree cancelled and all credit for it removed if I had tried it on with my Prof.
    Take the bull by the cortex and do as the greenies do, just get vocal. Every interview mention the wrongs. Write brief articles when wrong is discovered and rather than just tell each other, get the 16 year old receptionist or your off spring to proof it so the people who are not from a sciences back ground understand what is wrong, why and what it means to them.
    As an example.
    In 20 years time under the current carbon fraud agenda a major war (WWIII) is most likely to occur. Think; what will happen when the poor starving people in South America find out that they are growing cane for sugar for car fuel. A large number of starving people with literally nothing left to loose could conceivably take on the North Americas where they are living life large rolling around of cane sugar powered cars.
    WAR? the guys joking? Have a look at what started the first and second world wars.
    The problem with modern science is that is it unsexy, invisible and for most people cold an boring. In Australia we have DR Karl Kruszelnicki who hits the high school circuit and puts science to those who are learning in a way that he holds teenagers attention. If you can’t get your findings understood by the public, then what’s the point of any anti green house blogs and pages?

  17. Louis Hissink said

    Jeff,

    good work – and while a little left field, I think it might be of interest to consider that tree growth might also be influenced by the earth’s electric field.

    So here is another possible cause for the observations.

    🙂

  18. Pete said

    Geoffrey,

    I agree. I’ve been trying to do it, but I see one big challenge in explaining the case against C-A-CO2-GW to non technical folks. That is that their eyes glaze over and they revert to their comfort zone of what they saw on CBS or NBC. For me the challenge is to only pick a nugget at a time.

    I am now trying to develop the tie to analogous cases like the wall street bailout or the ulcer bacteria story. But it needs to be concise. Perhaps we need some talking points.

    Even then it may take 10 years like the ulcer bacteria case for the public and then the policy folks (or vice versa) to change.

    Also, it seems that we may be witnessing a new problem that we have not seen before in human history. It revolves around analysis of complex and systems of systems problems, whether they be scientific, financial, legal/regulatory or all of the above. Will we ever adapt to these so we don’t continually get snookered by insiders? Perhaps auditors like Steve M. need some marketers to give their profession a makeover.

  19. Jeff Id said

    Douglas,

    The trend line of the offset and magnification are complex in shape. In the calibration range the offset has a high degree of slope and recovers at a rate dependant on the frequency of the noise. The amplification of the series is constant throughout. You are correct about the result though, any value away from the ‘offset zero value’ will be de-magnified by a constant multiplier.

    Sorry for the delay in replying, I got caught up in learning how to program in R. I think it will make my work quite a bit easier. This link shows the shape of the distortion quite clearly and it applies to every hockey stick reconstruction I know of.

    https://noconsensus.wordpress.com/2008/09/29/simple-statistical-evidence-why-hockey-stick-temp-graphs-are-bent/

  20. Jeff,
    Thanks for the reply. You don’t need to apologize for being late with a reply. I know how much effort it takes to host a blog. I think I’ll have to retire before I can spend the time to do what you, Steve McIntyre, and Anthony Watts do.

  21. Diatribical Idiot said

    Jeff, regarding your comment #11…

    I enjoy a good argument and debate as much as the next guy, but I simply can’t stomach the discourse at the misnomered site, “Open Mind.”

    I had a recent exchange over there regarding the fact that all short-term trends in every region of the globe are negative. I was more than willing to concede all the statistical caveats that short-term trends are not necessarily indicative of the long-term, and r-squared is generally lower, and potential distortions from ENSO effects, and yada yada.

    Well, I eventually suggested that, since everyone is caught up in the fact that only longer-term trends seem to matter, then can I at least get agreement on the fact that the 30-year temperature trend in the Antarctic region, based on the RSS satellite measures, had a negative overall trend.

    I mean, the data is right there for anyone to plunk into a spreadsheet and run a trend line against. You can provide all the excuses you want, but at least acknowledge that the raw data shows a negative trend.

    They could not physically bring themselves to acknowledge it. At least one person outright said “Nope. And since you’re obviously not open to discussion on why, then you’re not worth the time.”

    For all the talk of “deniers” not being open to science, it’s almost scary how unwilling people are to simply confront actual data.

    Keep up the good work. Not everyone is willing to dig into this information, so you do a valuable service by parsing it the way you do and trying to make sense out of it.

  22. jeff id said

    #21

    I was thinking of pointing out Tamino’s don’t be fooled again post to you. Is that where you commented, I would be interesting to hear from some open minds.

  23. Diatribical Idiot said

    That isn’t where I commented. It was actually an old post discussing ice data.

    I just went and checked out that post. He goes through a lot of statistical gyrations to state the obvious. It’s a valid post. And I agree completely that anyone who points to short-term trends as “proof” of cooling or “proof” against warming is overstating matters. I try not to leave that impression, but I find that by simply providing a chart that shows as much elicits the label of “denier.” I’m fine with that.

  24. Jeff Id said

    I agree that his post is valid, but his conclusions are a bit off. The way he states it, he makes it sound impossible that the current downward trend is anything other than red noise.

    He even refers to it as an “apparent” negative trend. Like somehow a negative trend is not really negative. Of course it is, it is only of a shorter term than the rest.

    I did notice he chose GISS data too for his analysis, and if you look into the corrections of the GISS data, they are as large as the signal in magnitude. The satellite data has some extreme corrections also. In both cases the corrections are necessary but poorly understood. (i.e. scientists are still negotiating more corrections)

  25. Eric Anderson said

    Douglas wrote:

    “The growth rate of anything, including trees, is not linear with temperature; a chart of growth rate vs temperature is close to an inverted parabola. The growth rate increases with temperature until it reaches an optimum range, then decreases when the temperature is too high.”

    True, but the situation is much more variable than that. First, the so-called “average” temperature for a particular year may not coorelate at all with growth. For example, a warm spring/summer and a cold fall/winter will likely have a very different growth profile than a cold spring/summer and a mild fall/winter, even if the “average” temperature for the year is the same. Further, for anyone who has more than one tree on his property, it is obvious that other factors can vastly outweigh the same temperature experienced by all the trees on the same piece of property: initial soil conditions, sun/shade, fertilizer, amount of watering, bugs/disease, and on and on.

    I can accept that a tree’s rings record growth; but recording growth and recording temperature are not at all the same. I would be happy to learn otherwise, but based on my current understanding, the idea that a relatively small number of tree ring proxies can yield a temperature reconstruction — with no evidence that the myriad of other growth variables are being held constant — is nothing short of ludicrous.

    You and Jeff are correct: Trees are great, but they make lousy thermometers.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

 
%d bloggers like this: