the Air Vent

Because the world needs another opinion

Online Experiment With the Latest Hockey Stick

Posted by Jeff Id on September 20, 2008

Ok, two days ago I published a report on the data called The All Important Blade of the Stick Uses Less Than 5% of the Data This report detailed what I consider powerful evidence that Mann and his group may have intentionally manipulated data in creating this paper.  I don’t say this lightly.

It is very clear that the proxies, which no one is sure represent temperature, were in fact extended using questionable statistics and then correlated to demonstrate that they match measured temperature. The addition of an up slope to 90% of the data was pasted on according to less than 5% of the data.  I am very suspicious that this 5% was chosen specifically because of its up slope in recent times. Think about that only 55 series determined the outcome of the majority of the 1209 series data!!

What really tipped it though was that Mann and crew made graphs like this one below.  I don’t care how good of a statistician you think you are, the PINK extension of this graph is obvious CRAP and has absolutely nothing to do with temperature. A four year old child can pick out that the graph below isn’t quite right!

So, my suspicions are so great now that I am putting my butt on the line here.

There were 148 proxies available to Mann which were discarded for unknown reasons.  These were accidentally posted on an NOAA server as original data for this paper.  As far as I can tell these series were not discussed in his paper.  Of the 148 – 64 had data in the current years which would have been very useful for calculating the pink extension. Why didn’t he like these 64 series?

It is my contention that these series will have a negative slope or more negative slope than the 55 proxies which were used.  I have no way of knowing that, but looking at the piles of BS data I have previously noted and considering the pressures on Mann (internal and external) I believe they may have taken the step of intentionally eliminating data which did not support their conclusion!!!

My belief (based only on what I say are suspicious methods) is that the critical end of the latest hockey stick has been created from cherry picked data!

The experiment!

PART 1

I am going to plot first the 5% which was used to put the end on the 1083 data series and I WILL POST IT FIRST. The critical portion of this 5% will be the last 50 years because this is what was used to insert (INVENT) data at the end of 1083 of the 1209 proxies.  These proxies (after modification) will have an up slope at the end 50 years!! I have never seen one plotted by itself let alone an average of all of them.

PART 2

I will plot the average of the 64 series which were long enough to be used for extending (INVENTING) data and I will lay it on top of the first. It is my contention that the up slope in the last 50 years of the rejected data will be visibly less than the up slope of the used data!!

Some proxies (series) will need to be cut from the 64 series group that Mann rejected, i.e. boreholes and Luterbacher simply because he stated he wouldn’t use any data with temperature or historical information for pink colored proxy extensions either way. — Just following the paper.

PART 3 – What happens if I am wrong.

If I am wrong I will  take the time to actually look up and name the whole Mann group here, list all of them by name and apologize for being so suspicious of them that I even ran this online experiment.

Lets do this thang.

———————————————————————————————-

Results.

Ok, step 1 is complete.  I plotted the first graph below.

Holy crap.  Look at that.  Jeff Id steps out to an early lead.  Jeeez,  I’ve never actually been mad at a graph before this.  Look at the huge upslope in the last 100 years.  How about that nice warming period from 400 to 800 ad.

This is the 5% data series used to paste an end on 90% of the data Mann’s group used.

I didn’t expect this extreme of an example when I started this post!

It’s not over yet though, I still have to plot the data which wasn’t used.

The final data!!!!

How did I guess.  This experiment could have gone horribly wrong for me.  The pink line represents the average of the not used data.  Of the 64 series, there were a bunch of Luterbacher datasets so the 64 turned into 46 series to generate the purple line.

There will be no apology today!

Summary

Mann used a select few data sets to paste an up slope on the end of 90% of the data in his latest paper.  The series used to make the pink line above could have been used in the process but were conveniently eliminated from the paper without mention.  Mann accidentally posted these series on a national server and cannot eliminate them easily.

It was my guess that these series would have a reduced up slope compared to the data selected in the last 50 years,  I had no idea the selected data would demonstrate such an extreme slope so I could have lost.  I can’t prove it yet but this now looks like deliberate manipulation of data to me!


26 Responses to “Online Experiment With the Latest Hockey Stick”

  1. Matt Y. said

    Thanks for the hard work. I recently found your website and have really enjoyed it. Going from allegations of incompetence/sloppiness to fraud/malfeasance would be a big leap. It will be interesting to see how this plays out. I am especially interested in hearing the original authors explanation.

  2. masmit said

    Yes, thanks for taking the time to do this.

    It occurs to me that it might be helpful to see what a plot using both sets of data would look like.

  3. Pete said

    Couldn’t Mann say that he threw out those for other reasons and he was just sloppy in not clearly saying that in his paper?

    Also, to be more thorough shouldn’t the average be done via a global surface averaged/interpolated method which might diminish the relative importance of some of the more downward trending series?

    Also I agree with Masmit that it would be great to see the overall plot with all the series, but do you have access to the global averaging code? Is Mann using Hansen’s method?

    Doesn’t look good for Mann.

    He is talking at U. or Rhode Island (USA) on 23 Sep at 07:30PM (Honors Colloquium) and I can’t stop by to listen, but I wonder if anyone will ask questions or if there will be no Q&A?

  4. Jeff Id said

    Pete,

    I am absolutely certain that Mann and group will offer ‘reasonable’ explanations for the deletion.

    Mann has generously made his software available, I haven’t tried to run it yet so no global surface corrections.

    How about the entire concept of “infilling” data using other series, before determining if these proxies have anything to do with temperature?

  5. nanny_govt_sucks said

    Is there any reasoning given for pasting ends on to 90% of the data?

  6. MrPete said

    Jeff, SteveM has noted that Mann documented a filter requiring significant correlation with the (recent) temp record. That would be enough to cause this, would it not?

  7. Jeff Id said

    MrPete,

    I am not sure what filter you are referring to.

  8. david said

    Jeff,

    Thanks for all the hard work. If you are right, this is extraordinary. Given the scrutiny applied to the “original” hockey stick it is hard to believe that this could be “an accident”.

  9. Jeff Id said

    MrPete,

    I have been thinking what this blog will become. It is happening on its own but one thing I would like to do is help other people like myself understand some of the science. Some of my readers are very in tune with every detail of what is happening, others are like I was are just casual browsers. So I will reword my reply. I need to be more careful with my wording anyway.

    Can you give me some more detail as to which filter you are referring to? There are a couple of possibilities which may relate to your question.

  10. MugWump said

    Maybe Mann picked the desired slope that he wanted to “prove” and depopulated 95% of the data sets, by 90-100%, according to the degree to which such sets differed from the desired slope?

  11. JimmyBH said

    Am I correct in assuming that an average of the “final data” graph would need to be weighted 55/46 in favor of the dark blue line?

    At any rate, I would also be interested in an average of that graph.

    Interesting blog, I found it this evening via the “random hot post” feature on the WordPress dashboard.

    -j

  12. A Stoner said

    When it comes to global warming alrmists, I am sorry, but benifit of the doubt is not how to approach these human hating entities. Scientists have a responsibility to get to the truth, not to get to some predetermined end result. People in industry have responsibility to get to an end result. Scientists must aim to uncover truth and only truth.

  13. Raven said

    Icecap has you mixed up with Ian Jolliff: http://www.icecap.us/
    You might want to send him a note.

  14. Gary Plyler said

    Is Mann (et. al) getting federal funding for this potential fraud and/or malfeasance? Am I as a taxpayer paying for his shennanigans in the form of grant studies?

    When will the scientific community censure such potentially criminal conduct? When will an example be made?

  15. Jeff Id said

    Jimmy BH

    The weighting you suggest would be correct for my graph. However, the weighting Mann used is based on location (lat and long) so I need to do some more work to put corrected weighting on this.

    Gary,

    I don’t know the origin of Michael Mann’s salary but the problem is much bigger than the US government. His graphs are the cornerstone of the UN Intergovernmental panel on Climate Change. An organization who’s survival depends on climate change. Antrhopogenic global warming is more than a 100Billion dollar industry and it won’t go away anytime soon.

    Raven,

    I think it has been cleared up.

  16. Alan S. Blue said

    Jeff,

    If you really want to get steamed, do research on the shaft of the hockey stick. The elimination of the Medieval Warm Period is the egregious piece. Just plugging that into a search and following through history texts and summations of the period by historians gives a wildly non-flat impression of the temperature. It may be “regional”, but the local climate observations are pretty extensive and well-documented (if, admittedly, imprecise) from Greenland through central Asia and south into Africa. IOW: A pretty freaking big ‘region.’

  17. MattN said

    So this appears to be very nearly the same exact sleight-of-hand mathmatical trickery that he used before, yes?

  18. Diatribical Idiot said

    Good stuff, Jeff. Nice analysis.

  19. Owen McShane said

    At the onset of the Little Ice Age the Polynesians ceased their migration to New Zealand. (Maori)
    The Pacific is not normally regarded as part of Europe.

  20. […] Jeff Id (noconsensus.wordpress.com) provides us with some new perspective on the new hockey stick graph, Mann et al. 2008. It seems that the paper is not only a case of sub-prime science but an example of scientific fraud. […]

  21. paulm said

    Still looks roughly like a hockey stick, especially if you like hockey.

  22. […] Jeff Id (noconsensus.wordpress.com) provides us with some new perspective on the new hockey stick graph, Mann et al. 2008. It seems that the paper is not only a case of sub-prime science but an example of scientific fraud. […]

  23. […] Graphs To Dismiss Mann 17 10 2008 From The Air Vent’s Online Experiment With the Latest Hockey Stick: Data used by Mann for extrapolation Data used vs not used for extrapolation by Mann Mann used a […]

  24. Start Wall said

    great work if it you can prove it to be true

  25. Hello everyone!
    Valentines Day is within a few days and I want to purchase my gf something cute.
    Any ideas? Oh and anyone know where I can order cheap bouquets ? I want to preorder flowers for her for starters.
    I was also thinking of creating her a cd of all her favorite songs. Anyone have any free itunes codes? haha

    THANKS!

  26. […] From The Air Vent’s Online Experiment With the Latest Hockey Stick: […]

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

 
Follow

Get every new post delivered to your Inbox.

Join 148 other followers

%d bloggers like this: