One engineers perspective on global warming 33

In forum1554 MVP members of eng-tips.com develop articles for publication on the Internet. This informal peer-review process has resulted in several articles being published thus far. Early indications make this seem to be an effective way to make information available to the engineering profession.

One of the articles was on my perspective on global warming. It was published today.

I would be very interested to participate in a discussion here about the article or the subject in general. There were several conversations that were started and then cut short in the Engineering Writers Guild because that group is intended to promote publication, not debate points of view. If anyone want to take up those conversations, this is the place to do it (although it is too late to correct formatting, grammar, or punctuation errors).

The article is available at: One engineer's perspective on global warming

Thank you to all of those who worked hard on this article to keep me more or less honest. I had a lot of help, but at the end of the day any errors and/or omissions are mine. I'll share the credit for the good stuff, but I own the mistakes.

David Simpson, PE
MuleShoe Engineering

Law is the common force organized to act as an obstacle of injustice Frédéric Bastiat

 

[2dye4]

Lets just put up what Ritson found WRT Mann and McKitricks complaint.

""
Section 2 of the M&M GRL analyze results from simulations based on trend-less red-
noise and conclude that the MBH analysis produced hockey sticks “some of which bore a
quite remarkable similarity to the actual MBH98 temperature reconstruction”. The units
used by M&M to compare between MBH results and simulated noise results were arbitrary
and unrelated. To check the significance of this M&M claim we followed similar procedures
and simulated an ensemble of seventy-member data sets based on lag-coefficient 0.5 AR1
noise and analyzed them via short centered ( 1 Y). (m = 1 was used to bring our results
into correspondence with M&M’s). Analysis of the noise simulations finds hockey-sticks
heights randomly distributed up or down and thus with an ensemble mean height of zero,
The standard deviation of the heights about the zero mean is ∼ ±0.025. Our figure
1 A contains a ‘typical’ simulated hockey stick in units identical to those of the real
MBH signal. At first sight the hockey-stick appears to be a ‘zero-baseline’. Only when
multiplied by ∼ 50 does it become comparable to the MBH98 data.
""

So lets clear up GTTofAK's confusion. Ritson was responding to M,MK claim that the methods produced hockey sticks
when the climate data was exchanged with red noise. Ritson found that using the same methods as MBH98 the
hockey sticks so discovered ranged between + 1/50 th to minus 1/50 th of the results in MBH98 with random
distribution in the interval. To a reasonable person this is not relevant.

GTTofAK said
""That Mann's short centered PCA selects and equal number of positive and negative hockey sticks simply does not matter to the final result. His method is insensitive to the sign of the predictor. ""

Utter nonsense, this 'selection' was only a characteristic of the particular realization of red noise used in the very limited method test and had no
bearing on the results when applied to the real data.

The whole post is an exercise in misunderstanding what MBH98 attempted. Intentional or accidental ??

 

[rb1957]

it's not "Mann and Mckitrick", it's McIntyre and McKitrick ... geeze

have you read McIntyre's side of things ?

Quando Omni Flunkus Moritati

 

[GTTofAK]

“So lets clear up GTTofAK's confusion. Ritson was responding to M,MK claim that the methods produced hockey sticks when the climate data was exchanged with red noise. Ritson found that using the same methods as MBH98 the hockey sticks so discovered ranged between + 1/50 th to minus 1/50 th of the results in MBH98 with random distribution in the interval. To a reasonable person this is not relevant.”

Simply put neither you nor Dr. Ritson have any clue what Mann’s method actually does. The next step after calculating the PC1 is rescale the PC1 and apply that result in the regression step. The original scale of the PC1 means nothing. Dr. Ritson’s comment about the original scale is totally ignorant of the method involved. There is a reason we haven’t heard from Dr. Ritson in a very long time. He realized that the totally screwed up on many levels.


“GTTofAK said
"That Mann's short centered PCA selects and equal number of positive and negative hockey sticks simply does not matter to the final result. His method is insensitive to the sign of the predictor. "

Utter nonsense, this 'selection' was only a characteristic of the particular realization of red noise used in the very limited method test and had no bearing on the results when applied to the real data.”

The only utter nonsense is the argument you just made. Dr. Ritson’s claim was clear as day. There is an equal number of positive and negative hockey sticks and their aggregate is 0. This claim has been made by many others and is a totally ignorant statement that just proves and abject failure to even try to understand the nature of Dr. Mann’s statistical meat grinder. The fact that Dr. Mann for 3 years let these statements sit even on his own blog without correction knowing full well that is method was insensitive to the sign of the predictor, so positives and negatives do not average to 0, shows him to be a very dishonest. He has pushed claims of others that he knows full well to be wrong because they suit his objective.

 

[2dye4]

What you posted has absolutely nothing to do with ( McIntyre and McKitrick )'s complaint about noise making hockey sticks.

""The next step ""

There is no next step in what was being discussed which was the substitution of red noise for the proxy data and IT making hockey sticks.

Then you seem to be completely unable to grasp that this experiment with red noise has absolutely NO bearing at all on the results when
the original proxy data is used and processed as claimed.

Please do explain in detail to this poor clueless wanna be engineer what you are talking about.

Again you start with this..

""There is an equal number of positive and negative hockey sticks and their aggregate is 0""

It only refers to the noise experiment and actually proves the method used to be unbiased in a statistical sense "unbiased estimator"

""
The next step after calculating the PC1 is rescale the PC1 and apply that result in the regression step. The original scale of the PC1 means nothing
""
Yes that is true but irrelevant. PCA attempts to isolate a common signal in a group of signals and it makes total sense to normalize their variance
and subtract their mean before inclusion. Why do you think otherwise.


Do try to break from skeptic tradition and post coherent and well explained thoughts that flow together to make a good case.
I know I am not as smart as you so I will admit you post seems like strung together bits of emotional persuasion.

 

[rb1957]

"Then you seem to be completely unable to grasp that this experiment with red noise has absolutely NO bearing at all on the results when the original proxy data is used and processed as claimed." ... you mean you can't appreciate that if the temperature data is replaced with a noise input and this is processed and produces a hockey stick, then it's the analysis method creating the hockey stick, not the input ?


Quando Omni Flunkus Moritati

 

[GTTofAK]

"There is no next step in what was being discussed which was the substitution of red noise for the proxy data and IT making hockey sticks."

You are proving more and more that you don’t know what you are talking about. There definitively is a next step in Mann's method. He rescales the PCs then runs that into the regression. Dr. Ritson’s claim that the hockey sticks were too insignificant misses this step. Once the a red noise PC1 is scaled using Mann's method it is very significant. Dr. Ritson missed a step because he obviously did not read either paper.

"Then you seem to be completely unable to grasp that this experiment with red noise has absolutely NO bearing at all on the results when the original proxy data is used and processed as claimed."

That doesn't even make any sense. Red noise matters because red noise also produces hockey sticks. So hockey sticks are an artifact of the method. As M&M showed quite easily the method will mine the data for hockey stick series. As long as some hockey stick or sticks exist in the data sets a hockey stick will dominate the reconstruction. Inability for a method’s output using the data to be distinguishable from its output using noise falsifies the method. This is what R2 verification is.

“Yes that is true but irrelevant. “

What the hell? Your first sentence said that there is no rescaling “There is no next step”. Now you say that yes there is a rescaling, in the same post no less.

“PCA attempts to isolate a common signal in a group of signals and it makes total sense to normalize their variance and subtract their mean before inclusion. Why do you think otherwise.”

That is centered PCA. You just got done arguing that PCA doesn’t need to be centered, quite to the objection of those who developed PCA BTW. It also has nothing to do with Mann’s rescaling.

Your posts are getting more and more incomprehensible. You really have no idea what we are talking about do you?

 

[GregLocock]

2dye4- you are aware that Mann's hockey stick generating code is available and that you can repeat the red noise experiment for yourself?

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[HenryOhm]

OK, I'll bite as one who isn't up to speed on most of this. Who are Mann, McIntyre, and McKitrick?

OK, well, Mann is a climate scientist, McIntyre is a Canadian mining exploration company director (?!), and McKitrick is an economist (?!, this stuff just writes itself):

http://en.wikipedia.org/wiki/Michael_E._Mann

http://en.wikipedia.org/wiki/Stephen_McIntyre

http://en.wikipedia.org/wiki/Ross_McKitrick

OK, I need to fill myself in more on the hockey stick:

http://en.wikipedia.org/wiki/Hockey_stick_controversy

The hockey sticks main inquisitors are:

http://en.wikipedia.org/wiki/George_C._Marshall_Institute#Global_warming

http://en.wikipedia.org/wiki/Fred_Singer#SEPP_and_funding

http://en.wikipedia.org/wiki/Soon_and_Baliunas_controversy#Funding_controversy

http://en.wikipedia.org/wiki/Competitive_Enterprise_Institute#Environmental_policy

...and the above Canadian mining exploration company director and economist.

Well, of course, there are a bunch of people who support the hockey stick:

http://en.wikipedia.org/wiki/List_of_large-scale_temperature_reconstructions_of_the_last_2,000_years

Why does it always have to be some oil industry lobby or Koch Brothers think tank that funds the skeptics or skillfully behind the scenes creates exaggerated controversy?

The science of global warming may not be the most clear cut science, but the best of the skeptics attempts to undermine it at times is simply, well, unscientific. In a few cases, they come off as just pathetic. I really tried to educate myself (Salby, satellites that aren't even in orbit yet, this hockey stick "controversy"), but as before I end up just shaking my head.

 

[rb1957]

who is more up to speed on statistical analysis, Mann or McKitrick ?

if you want to poke holes in the analysis that M&M did on Mann's published work, then challenge that. don't say "oh they aren't climate scientists so 1) they don't have a valid comment, and 2) they must be in the pay of "Big Oil".

as have as I know M&M didn't say "therefore there isn't any climate change" they just said "Mann's analysis, producing the hiockey stick, is crap". They addressed his statistical analysis, something they have an understanding of.

Quando Omni Flunkus Moritati

 

[IRstuff]

Mann is, of course, THE lightning rod of climate skeptics, but one of the reasons that he's STILL the lightning rod is that those that came after him are much more careful with their data and analyses, and are substantially less assailable, statistically. Mann's work is >15 yrs old, and that were the only thing to look at, then there would indeed be a host of issues with the state of the art.

However, there have been additions to the body of work, and they are all consistent with Mann's suspect work. For a nice, technical summary, see:

http://www.skepticalscience.com/broken-hockey-stick.htm

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

Of course I can. I can do anything. I can do absolutely anything. I'm an expert!

 

[rb1957]

surely if they (the subsequent studies) are consistent then there was an underlying "truth" in Mann's work, and M&M are just pesky theoreticans fussing over details.

or if essentially the same data has essentially the same things done to it and producing essentially the same result, consistently; and one of these is proven to be false, invalid, then are others are ...
a) false, invalid, or
b) true, valid ?

Quando Omni Flunkus Moritati

 

[GTTofAK]

2dye4- you are aware that Mann's hockey stick generating code is available and that you can repeat the red noise experiment for yourself?

If he looked at the code he would also see the rescaling step of the reconstructed principle component to the observed principle component(measured temperature) plain as day. Not that he would ever understand it. This rescaling makes making Ritson's argument both moot and frightfully ignorant. He obviously read non of the supporting information before making his comment.

 

[rb1957]

"Not that he would ever understand it." ... now, now; play nicely.

Quando Omni Flunkus Moritati

 

[GTTofAK]

"Mann is, of course, THE lightning rod of climate skeptics, but one of the reasons that he's STILL the lightning rod is that those that came after him are much more careful with their data and analyses, and are substantially less assailable, statistically. Mann's work is >15 yrs old, and that were the only thing to look at, then there would indeed be a host of issues with the state of the art.

However, there have been additions to the body of work, and they are all consistent with Mann's suspect work. For a nice, technical summary, see: "

A shotgun approach to science is meaningless. If you set out to create a hockey stick you will probably create a hockey stick. Statistical analysis that starts at the conclusion and works its way back is meaningless.

 

[2dye4]

GTTofAK

"" Once the a red noise PC1 is scaled using Mann's method""

1 I don't find this method called out in MBH98 from a quick reading.
2 Ritson used PC1 ONLY to duplicate what McIntyre and McKitrick did to form their complaint.

from Ritson

"" analyzed them via short centered ( 1 Y). (m = 1 was used to bring our results into correspondence with M&M’s) ""

So it appears that it was really the two Mc's who got it wrong and Ritson was only responding to THEIR error.

3 Mann appears to use a number of principle components to do the reconstruction selecting them based on a least squares fit
to the calibration period. So where do you get the idea that they did otherwise.


Much of this criticism of MBH98 is secondhand, thirdhand.... N'thhand retelling of the story with errors added each time to suit the teller.

It makes it convenient for critics to regurgitate and sound like they know something so I want to look at the core of it.

One question I have.

Is there a significant difference between the methods outlined in MBH98 and those actually used. IE. coded up in the software
used directly for MBH98. Just so I don't waste my time posting first about the paper and then have to deal with complaints that
the actual method used was not that outlined in the paper.


Another question on second though.

"" He obviously read non of the supporting information before making his comment.""

What supporting information..Do you mean source code ??

 

[GTTofAK]

"OK, I'll bite as one who isn't up to speed on most of this. Who are Mann, McIntyre, and McKitrick?

OK, well, Mann is a climate scientist, McIntyre is a Canadian mining exploration company director (?!), and McKitrick is an economist (?!, this stuff just writes itself):"

McIntyre is a degreed mathematician with a specialization in statistics. He has used his statistical expertise in to make a good career in minning. McKitrick is an econometrician. I'm currently working with a group of econometricians to parse through years of 2 second power data. They are absolutely excellent at time series analysis and can apply their expertise to any field.

SO WHAT IS YOUR POINT?

 

[rconnor]

The only argument more trite than the “pause” is the “hockey stick”. rb1957 has just nicely wrapped up the decade of squawking from skeptics in this one sentence:

surely if they (the subsequent studies) are consistent then there was an underlying "truth" in Mann's work, and M&M are just pesky theoreticians fussing over details.

Could not have said it any better (and certainly not with such brevity!).

(rb1957, with regards to your “or” statement, see Wahl and Ammann 2007 or Mann et al 2008 or NOAA Paleoclimate Reconstruction Network.)

On top of all this, the “hockey stick” is really just the “it’s natural” argument phrased differently. Again, if recent climate change really is “natural” then one needs to explain how all natural drivers should be leading to lower temperature and energy uptake, not higher. Since the 1950’s, Solar activity is down, volcanic aerosols are up and anthropogenic aerosols are up but global temperature has risen and OHC has increased, both at very quick rates.

By whatever metric you want to use, the climate is seeing a positive total energy uptake. But is this due to increases in incoming energy, redistribution of energy, “generated” energy or decreases in outgoing energy? Solar and aerosol trends lead to reduced incoming energy. Energy redistribution seems unlikely as geothermal flux is far too weak (flux of 47 +/- 2 TW = 0.09 W/m^2 compared to an extremely conservative TOA imbalance value of 0.58 +/- 0.15 W/m^2 or solar radiation at 341.3 W/m^2) and far too consistent (minor changes on geological timescales, let alone the past 50 years) (see Stein and Stein, 1992 or Davies and Davies, 2010) to explain the recent trends. Energy “generated” (really redistributed) by waste heat of human activities is even less significant at 0.03 W/m^2. It’s almost as if there’s a decrease in outgoing energy. Hmmm….

 

[GTTofAK]

"GTTofAK

"" Once the a red noise PC1 is scaled using Mann's method""

1 I don't find this method called out in MBH98 from a quick reading.
2 Ritson used PC1 ONLY to duplicate what McIntyre and McKitrick did to form their complaint.

from Ritson

"" analyzed them via short centered ( 1 Y). (m = 1 was used to bring our results into correspondence with M&M’s) "

Yet more proof that Ritson is an ignorant boob who didn't read the papers or any of the supporting material before making his comment. M&M carried their PCA all the way to PC4. So m=1 is wrong on its face. He clearly never read any of the papers or supplementary materials. Almost everything Dr. Ritson says from the number of PCs used, to what PCA is sued for is on its face wrong. That man is clueless.

I don't know why you keep trying to publish this turd. There is a reason we haven't heard from Ritson in almost a decade. Everything he said in his comment was wrong.

 

[GTTofAK]

"Solar activity is down"

Turn a gas stove to high, put a pot of water on it, heat for 2 minutes. Lower to 7 continue to heat. Does the pot stop warming?

Seriously are we engineers?

 

[2dye4]

""Turn a gas stove to high, put a pot of water on it, heat for 2 minutes. Lower to 7 continue to heat. Does the pot stop warming?""

It certainly could.

Do you have a point ??