Mann’s Screw Up #6 – Altering Data

Before I begin, I want to congratulate everyone who has made it through the series thus far. We didn’t cover all the problems in Michael Mann’s 1998 paper, but we made it through. Congratulations!

I’m afraid we’re not done though. Now we have to move onto Mann’s 1999 paper. This paper is a fun one. Its results depend entirely upon Michael Mann adjusting data from 1000-1400 AD to fix a problem he says exists in 1800-1900 AD.

Mann’s 1999 paper (referred to as as MBH99), extended the temperature reconstruction of the 1998 paper (referred to as MBH98) back in time 400 more years. The result (referred to as MBH) was a temperature reconstruction extending from 1000 AD to 1980 AD. This result is the iconic “hockey stick”:

hockey_stick

We can verify the MBH99 results were simply spliced onto MBH98, meaning MBH98’s results were unchanged, by comparing the two reconstructions. MBH98’s data can be found here. MBH99’s data can be found here

MBH-Comparison

The two are clearly identical wherever they overlap. I stress this point because in MBH99, there is extensive discussion of an adjustment they made to their data. Specifically, they adjusted NOAMER PC1, a proxy we’ve previously established was one of only two proxies which showed a hockey stick. In reference to this proxy, Mann says (in his book, page 51):

In a 1993 article, Graybill and idso had shown that these very trees might be expected to exhibit a positive growth response to increasing atmospheric carbon dixiode levels.

Sound familiar? It should. It’s the exact same point I made in reference to the series when it was used in MBH98. That means Michael Mann acknowledges one of the two key proxies to his 1998 temperature reconstruction is unsuitable for his purposes. It was so unsuitable he decided the data needed to be adjusted, saying (page 52):

By correcting for that carbon dioxide effect… we could indeed now meaningfully reconstruct Northern Hemisphere average temperatures over the entire past millennium.

But wait, “that carbon dioxide effect” is the idea human emissions of CO2 caused trees to grow more. Human emissions of carbon dioxide only started taking off ~1800. Mann makes it clear the divergence he believes it caused (page 52):

was almost perfectly correlated with the exponential rise of atmospheric carbon dioxide since the early nineteenth century associated with the Industrial Revolution. The disappearance of the divergence in the twentieth century was consistent with warmth once again returning as the key factor controlling the growth

Let’s assume CO2 began causing problems in 1800. Let’s also assume those problems vanished in 1900. Why in the world did Michael Mann only adjust data prior to 1400? Why would a problem which only existed from 1800-1900 manifest itself before 1400 but not anywhere else, including 1800-1900?

And more importantly, why in the world did thousands of scientists from all over the world accept the argument anthropogenic influences on the planet’s climate traveled back in time to magically manifest themselves in 1000 AD? It’s insane, and they made it the figurehead for the global warming movement!

Over to you Mr. Mugatu:


You may have noticed I didn’t discuss the validity of the adjustment insofar as Michael Mann made it. I’ll be doing that in the next post. For both a visual demonstration of the problem and a preview of that post, here’s a gif (created by the awesome commenter Jean S.) showing three different versions of the NOAMER PC1 series Mann considered using.

Remember, the difference between each version is only in how one corrects the data for a problem Mann says exists only from 1800-1900.

Advertisements

4 comments

  1. The red part on the end appears to be the spliced in instrumental data – worth noting. (If you scroll to http://climateaudit.org/2010/02/03/the-hockey-stick-and-milankovitch-theory/#comment-219355, a portion of the caption is legible for what I think is (one of) those same plots.)

    And Jean S.’s moneyline is: “However, the real surprise came when I tried the algorithm with the original Mannian PC1, i.e. without any “fixing”. The RE scores are 0.38 and 0.33, so based on these “skill metrics” there is no reason to “fix” the PC in the first place!”

  2. Al, the image in the comment you link to is just a modified version of the first image in this post. It has the same caption, which you can see the entirety of in the copy I posted. The red is definitely the instrumental record (it’s labeled as such), but it isn’t spliced on. The two graphs are just overlaid.

    As for that quote, it’s actually one of the points I’ll be covering in the next post. I find it amazing such a random and ad hoc “fix” is a large part of the basis for what is arguably the most important image in science of the last two decades.

  3. Offhand, I believe I have the unadjusted version, the final adjusted version, plus one of the unused versions. I might have the fourth one as well, but I can’t recall. I’ve gone through a number of computers in the time I’ve followed the hockey stick debate, and I have never maintained any sort of archive of things. It’s poor practice, but I never expected to follow the topic for 10+ years, much less to be actively involved in the discussion of it.

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