Buy

Books
Click images for more details

Support

 

Twitter
Recent comments
Recent posts
Currently discussing
Links

A few sites I've stumbled across recently....

Powered by Squarespace

Discussion > Are Geological Paleo-Climate Records Relevant to The Climate Debate?

Our global paleotemperature reconstruction includes a so-called “uptick” in temperatures during the 20th-century. However, in the paper we make the point that this particular feature is of shorter duration than the inherent smoothing in our statistical averaging procedure, and that it is based on only a few available paleo-reconstructions of the type we used. Thus, the 20th century portion of our paleotemperature stack is not statistically robust, cannot be considered representative of global temperature changes, and therefore is not the basis of any of our conclusions. Our primary conclusions are based on a comparison of the longer term paleotemperature changes from our reconstruction with the well-documented temperature changes that have occurred over the last century, as documented by the instrumental record. Although not part of our study, high-resolution paleoclimate data from the past ~130 years have been compiled from various geological archives, and confirm the general features of warming trend over this time interval
Source: http://www.realclimate.org/index.php/archives/2013/03/response-by-marcott-et-al/comment-page-1.
See also: http://rogerpielkejr.blogspot.fr/2013/03/fixing-marcott-mess-in-climate-science.html

Apr 18, 2016 at 7:15 PM | Registered CommenterMike Jackson

We have thermometers for the 20th Century.

HTH.

Apr 18, 2016 at 7:23 PM | Unregistered CommenterPhil Clarke

And those thermometers would have shown the same dynamic resolution had they existed thousands of years ago. A fraudulent hockey stick is fraudulent whether it is produced by Mann or Marcott.

Apr 18, 2016 at 7:28 PM | Unregistered Commentermichael hart

Well, no, Entropic Man, the long term effects of warming are also highly net beneficial. We'd be storing blubber against the long freeze.
===========

Apr 18, 2016 at 7:53 PM | Unregistered Commenterkim

Marcott et al (2013) is the first paleoclimate paper I recall in which statistical methods were used to (blah blah).

Why are other geologists so reluctant to use statistical methods in paleoclimate studies?
Apr 18, 2016 at 5:45 PM | Unregistered CommenterEntropic man

EM - you have asked similar questions before and they have been answered before. We know that you are proud that you took "statistics 101" (as you once described it) "introduction to statistics for biologists", presumably. But it seems that the course did not cover the topic of when statistical analysis cannot be applied because of lack of knowledge or of data.

To answer the question again. Statistical methods cannot be used unless you have some sort of statistical model.

Why are other geologists so reluctant to use statistical methods in paleoclimate studies?

Presumably because they are well aware that statistical methods can only be used when you have a statistical model that can be verified, either by analysis of a physical model of the thing you are studying, or where you can regard it as a black box whose internals are unknown but you have adequately long records to construct and test your statistical models empirically.

The problem is that constructing statistical models to do useful statistical analysis on your computed results requires orders of magnitude more data than is available.

Making up a statistical model and claiming it is valid without the possiblity of validating it is a form of deception - of self and of others. Now, bearing in mind your question above, who does that apply to?


_____________________________________________

BTW - are Marcott et al geologists as you suggest? I thought they were - like Mann, and as described by somebody else in this thread, "climate scientists" - trying to extract historical information from data that they had obtained somewhere but essentially had no more understanding of it as geologists than you or me.

Apr 18, 2016 at 7:54 PM | Registered CommenterMartin A

Also, it is heartwarming to hear you admit the short term benefits of warming. Too few alarmists are honest enough to admit that.
================

Apr 18, 2016 at 7:55 PM | Unregistered Commenterkim

Heh, Marcott, whoooosh, an elevator to the top. He'll be sorry, he made a fool of Obama.
===========

Apr 18, 2016 at 7:56 PM | Unregistered Commenterkim

I think Andy Revkin caught on to a little of the corruption, there. You should see his interview with one of the co-authors.
=================

Apr 18, 2016 at 7:59 PM | Unregistered Commenterkim

EM

"Paleoclimate denies the paper's (Marcott et al(2013).validity. Why waste time linking him to evidence that his cognitive dissonance will reject out of hand?"

I did not deny the paper's validity. I pointed out that because of its failure to resolve a major and very well documented climate excursion many Paleoclimatologists did not consider it's conclusions to be robust.

I also questioned the claim that its resolution was 20 years.

Now you have explained that you base your claim of a 0.02 w/m2 energy imbalance over the 5000 years before 1850 CE on Marcott et al ( 2013)

Thank you.

Apr 18, 2016 at 8:02 PM | Unregistered CommenterPaleoclimate Buff

Paleoclimate buff

Marcott et al quantify the size and limits of the uncertainty in their ensemble data.

Very few other workers make any effort to do the same. You cannot describe the 8.2kya event as major and well documented when you have no idea if the techniques used have the temporal and temperature resolution to detect it.

Without a clear statistical idea of the uncertainties you cannot say whether the event is real, or just wishful thinking. That the first major meta-analysis of the proxies fails to find the 8.2kya event is an indication that it may not have happened.. At best it may have been a local event which only showed up in the North Atlantic proxies.

I would be willing to bet that you look at Holocene proxies and see what you expect. If you see a trough around 8.2kya it confirms your bias.If you do not see an 8.2kya trough there must be something wrong with the data.

Apr 18, 2016 at 8:34 PM | Unregistered CommenterEntropic man

7:56 kim I think Andy Revkin caught on to a little of the corruption, there. You should see his interview with one of the co-authors.

Yes. Revkin was caught on.

...for 11,300 years temperature has been doing a long slow slide down ...the last century they picked up - that much -..... and then BOOM we're just outside the elevator - you know? Up and out. Super hockey stick! Yes!

But, later, ...

...the 20th century portion of our paleotemperature stack is not statistically robust, cannot be considered representative of global temperature changes...


As I said to EM recently, once somebody shows themselves to be untrustworthy, their statistical calculations don't cancel that out. That is just how it is.

Apr 18, 2016 at 9:46 PM | Registered CommenterMartin A

I would be willing to bet that you look at Holocene proxies and see what you expect. If you see a trough around 8.2kya it confirms your bias there must be something wrong with the data.If you do not see an 8.2kya trough there must be something wrong with the data it confirms your bias.
Seems equally valid, EM. And in the rest of that post you seem to be arguing against yourself.

Apr 18, 2016 at 10:02 PM | Registered CommenterMike Jackson

Martin A

Are you going funny again?

Marcott et al included the full time period their data allowed and clearly said that the 20th century portion of their study was not statistically robust, because only a few proxies continued that far. Pointing out the weakness of some of your data if necessary is standard scientific practice. How is that dishonest?

Fortunately we have a continuos temperature record from 1880, so the statistical reliability of the Marcott uptick is not a problem. It has been confirmed by direct measurement.

This continuous assumption that anyone who does not agree with your delusion is dishonest becomes wearying after a while.

Apr 18, 2016 at 10:06 PM | Unregistered CommenterEntropic man

Exactly right EM. Some seem to have missed part of Marcott's answer above

Our primary conclusions are based on a comparison of the longer term paleotemperature changes from our reconstruction with the well-documented temperature changes that have occurred over the last century, as documented by the instrumental record.

Or did the warming of the last 100 years just not happen ?!

Apr 18, 2016 at 10:15 PM | Unregistered CommenterPhil Clarke

Mike Jackson

The whole point of that post is that we don't know how reliable individual proxies are, because workers like Paul Dennis do not measure their confidence limits. He talks about different degrees of uncertainty but has not quantified them.

Some individual proxies show an 8.2kya event, some do not.There is no statistical evidence for their reliability. We do not know how uncertain their results are, so cannot say whether the positives are seeing reality or noise.

When you put 70 of them together the positives and negatives cancel out and the event disappears. That ensemble does have a full statistical analysis, which indicates that before 1850 the data are statistically robust.

I don't know for certain whether there was an 8.2kya event or not. A phenomenon which becomes less likely as you look closer may well be a mirage.

Have you noticed the Irony? Usually it is me arguing that a climate phenomenon is real, while you lot protest that it is uncertain. In this case you all think the 8.2kya event is real, while I am uncertain.

Apr 18, 2016 at 10:35 PM | Unregistered CommenterEntropic man

EM,

I take great exception to your statement and insist you retract it because it is a blatant falsehood. In all my work I take great care to quantify the precision of my measurements. I have patiently explained to you many times how the precision of measurements can be taken through to confidence in temperature estimates. You are an autodidact who has very little understanding of how proxies work and live in a world of blissful ignorance.

As for any further discussion this really is the end. You are a blustering idiot.

Apr 18, 2016 at 10:47 PM | Unregistered CommenterPaul Dennis

Paul,


You are a blustering idiot.

You really do fit in here perfectly. That isn't a compliment.

Apr 18, 2016 at 10:54 PM | Unregistered Commenter...and Then There's Physics

In one of the many posts at Climate Audit discussing the Marcott hockey stick, one reader gave a good, simple, visual, aid for those unaware of the standard operating procedure to make a climate hockey stick.

Apr 18, 2016 at 11:02 PM | Unregistered Commentermichael hart

aTTP: and what do you know about astrophysics that I do not?

I know what your answer is going to be; while the words used may not be those you would use, what Paul Dennis said is the same answer.

Apr 18, 2016 at 11:15 PM | Registered CommenterRadical Rodent

Paul Dennis, some of us were wondering when you were going to notice that. Welcome to Entropic Man.

Apr 18, 2016 at 11:36 PM | Unregistered Commentermichael hart

Paul Dennis

Relax. I am not attacking the technical quality of your fieldwork or the precision of your laboratory technique.

Let me make myself clearer. In biology organisms are not all identical. Every measurement gives a slightly different value. To compensate for this you take repeated measurements and then calculate a mean. To judge the quality of the mean you calculate standard deviation and confidence limits.The problem is that precision and variation are not the same My instrument may measure to +/-0.1, but different measurements may vary n the range from 8.0 to 10.0

Thus a measured variable is not given as a single number 9.0, but as 9.0+/- 1.0.This indicates the mean and the uncertainty of your mean
These are included in data tables and shown as error bars on graphs.Any biological paper without such analysis fails peer review.

Am I correct that when you take a sediment core you take a series of oxygen isotope ratio measurements along the core and then use them to build up a time sequence?

I looked at the graphs in your January 2016 NAO paper. Each measurement was represented by a single point, so I do not know how many times you replicate each measurement or whether you take measurements from several cores from the same site.

You certainly know the precision of your technique, but do you know the variation? If you only take one measurement at each depth you do not know how much other measurements from the same depth or from different cores would vary. If you take multiple measurements from each depth you can calculate the variation in your sample and express your final temperatures with a numerical value for their uncertainty.

You earlier described the drop in temperature during the 8.2kya event as about 1.7K. Would you not prefer to be able to describe the cooling as 1.7K+/-0.5?

This is what I have been seeking from you throughout this thread, a quantitative measure of the uncertainty in your temperature estimates.

I would like to be able to look at a paleotemperature graph such as GISP2 and know how much confidence to put in each value. I want to know how big the error bars should be. Are they +/-0.1C, +/-1.0C or +/-5.0C?

Apr 19, 2016 at 12:10 AM | Unregistered CommenterEntropic man

How can aTTP play games of Jolly Hockey Sticks with Cook, Lewandowsky et al, if people keep breaking his stick?

Apr 19, 2016 at 12:16 AM | Unregistered Commentergolf charlie

EM, I think you may have just been talking to a person whose knowledge of statistical analysis is rather more profound than your own.

Apr 19, 2016 at 1:11 AM | Registered CommenterMartin A

Martin A, I expect EM can prove with his own statistics that he knows best.

Apr 19, 2016 at 1:29 AM | Unregistered Commentergolf charlie

Yes, we see, Phil, one understanding for the scientists, and an entirely different one for the President and the press. It's not nice to fool Mother Populus.

So why are you at it?
==============

Apr 19, 2016 at 1:37 AM | Unregistered Commenterkim