the Air Vent

Because the world needs another opinion

Wheeerrre’s Climate Waldo

Posted by Jeff Id on June 23, 2010

Ok, Steve M is back at it today with a cool post on another blog’s criticism of  Steve Mosher and somewhat collaterally, Climate Audit. In this post,  Steve describes how Arthur Smith has mischaracterized what has been said at CA about hide the decline.  Now I’m not here to defend CA -cause Steve does fine by himself but let’s look at this stolen without permission plot from CA.

Gee the green line is all from the same data, at least it is supposed to be—- I wonder, which one is the real green line?   Which one has meaning in the physical world. I wonder why the right side image didn’t exist prior to climategate?

Waldo is faster than we thought…tricky bastard him..

Update:  TTCA in comment 1

It all started as the same data, I think, but wound up being statistically molested into a barely recognizable state (more accurately, multiple barely recognizable states). The “Climategate AR4″ figure is the data as Briffa originally found it, which was then smoothed with padding to get the figure “Briffa AR4″, which mostly, but not entirely “hides the decline”. This already dubious at best, stunningly idiotic at worst, procedure was not misleading enough however. For the purposes of the figure “Mann TAR”, a different smooth and padding were not only used, but the data were scaled differently, literally reducing the variance. The “WMO” graph is the most baffling-Jones appears to have used an even more aggressive smooth than Mann, used either the original scaling, or slightly enhanced for Briffa’s data-and adds to this by apparently padding not with averages from observations, not averages from proxies, but the actual observed data, there by making it appear as thought the reconstruction of temperatures in the 20th Century was nearly perfect, which gives precisely the opposite impression one would get if one looked at Briffa’s original data (that the proxy’s performance was rather poor).

14 Responses to “Wheeerrre’s Climate Waldo”

  1. timetochooseagain said

    It all started as the same data, I think, but wound up being statistically molested into a barely recognizable state (more accurately, multiple barely recognizable states). The “Climategate AR4″ figure is the data as Briffa originally found it, which was then smoothed with padding to get the figure “Briffa AR4″, which mostly, but not entirely “hides the decline”. This already dubious at best, stunningly idiotic at worst, procedure was not misleading enough however. For the purposes of the figure “Mann TAR”, a different smooth and padding were not only used, but the data were scaled differently, literally reducing the variance. The “WMO” graph is the most baffling-Jones appears to have used an even more aggressive smooth than Mann, used either the original scaling, or slightly enhanced for Briffa’s data-and adds to this by apparently padding not with averages from observations, not averages from proxies, but the actual observed data, there by making it appear as thought the reconstruction of temperatures in the 20th Century was nearly perfect, which gives precisely the opposite impression one would get if one looked at Briffa’s original data (that the proxy’s performance was rather poor).

    I just find the more I read about this stuff, the more baffled I am by these guys. Assuming they aren’t doing this kn purpose might be less generous than I thought, since the alternative is incredible stupidity. That and enough gall to induce kidney failure.

  2. kim said

    None dare call it fraud.
    ============

  3. Layman Lurker said

    Deep Climate had an interesting post recently about the TAR version of the “trick”. He was making the argument that Steve’s “narrative” was all about blaming Mann for truncating Briffa’s data for the TAR. DC states that the emails show that Osborne sent a truncated version of his data to Mann and that Mann was only deferring to Briffa on this matter. IOW – no “trick”. DC’s post totally ignored the endpoint issue wrt Briffa’s series in TAR. It was not Briffa who created the illusion of endpoint coherence with other TAR spaghetti series. As Steve said on his post today, had the TAR caption describing the smoothing method for the graph been followed, the endpoint would have looked something like AR4 shown above. Therefore, the endpoint treatment of Briffa’s series in TAR is the key to whether a “trick” was actually played by Mann in TAR or not.

    At Arthur’s blog, DC does pick up on a point that Steve blogged about some time ago. Namely, that there seemed to be a slight re-alignment (upward) of Briffa’s series in TAR vs the published version. http://climateaudit.files.wordpress.com/2009/12/briffa_versions.gif This obviously must account for some of the “trick” effect in TAR. DC searched the emails for clues on this re-alignment and picked up on a discussion between Mann and Osborne on how to properly calibrate the series (an interesting aspect to this is that the base period for alignment of the TAR was from 1961 to 1990 – yet Briffa’s series was truncated at 1960). In the emails Osborne defends their choice of 1881 to 1960 baseline and Mann suggests 1931 to 1960. DC proclaimed that it was this re-alignment of Briffa in the TAR that accounts for the endpoint when coupled with “0” padding. Curiously, DC’s latest post at Arthur’s says that “recalibration” of Briffa in TAR was not done and is not the reason for the difference.

    IMO this appears to be where DC is going to go after Steve next. Pinning Steve down on whether it was Mann or Briffa who truncated Briffa’s data is not enough. In order for Mann *not* to have played a “trick” in the TAR, Briffa’s endpoint process must be explained, rationalized, and shown to have nothing to do with adding or padding with insturmental data.

  4. MikeN said

    Those right two graphs, do not look like AR4 6.10b, so what are they?

    Hiding the decline of divergence has an effect, but how much is the effect of using the 1945-1960 data for padding for Briffa in the third figure?

    I think the actual 6.10b figure at issue has a different trick, and while divergence is an issue, not that important in light of this trick. Not too different from the TAR chart where Briffa’s data disappears under the others, all of the proxies end early, with instrumental temperatures the last man standing. This gives the effect of making it appear that all the proxies are showing an uptick at the end.

  5. Steve McIntyre said

    The third graph is a direct re-plot of archived AR4 versions of the TAR series (MBH, Jones, Briffa, CRU) colored in TAR style. The fourth is the same as the third, using the complete Briffa reconstruction (without truncation).

  6. Genghis said

    MikeN – “all of the proxies end early, with instrumental temperatures the last man standing.”

    Fraud, pure and simple.

  7. Steve McIntyre said

    #3. In my Heartland presentation, I reviewed the chronology of the Trick in detail, observing that the first “bite of the poison apple” of deleting post-1960 data occurred in Briffa and Osborn 1999. I’ve consistently tried to place things in context – my point was that the trick email arose in the context of the aftermath of the 1999 IPCC Lead Authors meeting, not that any particular individuals were more or less responsible. Mosher’s been more involved in that narrative – he tries to place a bit of a white hat on Briffa during this period.

  8. andy said

    Whose work?

    What does it show?

    The rest is trying to muddy the water and get of on a technicality.

    It’s quite clear the purpose and aim of the ‘modification’ was.

  9. Layman Lurker said

    #7 Steve McIntyre

    Yes have read your presentation notes and other posts at CA where you speak to the fact that Briffa had truncated prior to TAR. DC speaks of the Briffa truncation which predates TAR to support his thesis that Mann is not to ‘blame’ – yet makes no mention of the fact that you were already on record in noting this chronology.

    I think DC’s argument breaks down with other inconsistencies in the big picture as well. For example, if Mann was just deferring to Briffa on truncating for TAR, why does he not defer on alignment and endpoint calculation? Osborne actually argues his baseline case to Mann in one email. IOW if the premise of ‘Mann as deferential’ does not hold for the alignment or endpoint issue, it does not hold *period*.

    BTW, I think that if you accept (for the sake of argument) that Briffa was justified in truncating at 1960, I agree with his method (published after the TAR) of using the complete series for processing the smooth before truncating at 1960. To me it shows that he is aware that the discarding of post 1960 data has (at best) a questionable scientific foundation and that the effect of endpoint padding would be misleading without such a foundation.

    There are other obvious circumstances with the TAR graph which in my mind do not allow the presentation of endpoint coherence to be explained away as some sort of innocent coincidence. Why was the caption of the TAR graph explaining the 40 year Hamming filter and padding with series means allowed to stand when it appears different smoothing methods were used for each series?

  10. Kenneth Fritsch said

    Take a poll. How many people anywhere think that AS or DC are interested in determining what is behind “hide the decline” as opposed to engaging in food fights as a distraction?

  11. MichaelM said

    #9 Layman Lurker: “Why was the caption of the TAR graph explaining the 40 year Hamming filter and padding with series means allowed to stand when it appears different smoothing methods were used for each series?”

    Excellent question clarifying yet another point overlooked by the blogs trying to exhonorate the Team.

  12. kim said

    Yah, KF, AS is deeply disingenuous and DC deeply paranoid.
    =======================

  13. GHowe said

    Thanks, Jeff et al, for being a reasonable voice in a crazy world

  14. Layman Lurker said

    DC has now posted along the lines of my speculation in #3.

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 142 other followers

%d bloggers like this: