October 18, 2004

Medieval Warm Period...

Tim Lambert and William Connolley think that my colleague Richard Muller has been snookered by McKitrick and McIntyre, who believe that Mann et al.'s data normalizations artificially enhance the influence of series that show an uptrend since 1900. But Connolley argues--I think correctly--that McKitrick and McIntyre are simply confused: the normalizations diminish the influence of series that show a recent uptrend.

Deltoid : In this column, Richard Muller claims that McKitrick and McIntyre have shown that the hockey stick graph is an “artifact of poor mathematics”. If you have been following the global warming debate this claim should look familiar, because McKitrick and McIntyre made the same claim last year as well. So what’s new? Well, last year they claimed that the hockey stick was the product of “collation errors, unjustifiable truncations of extrapolation of source data, obsolete data, geographical location errors, incorrect calculations of principal components, and other quality control defects.” Now they are saying that the hockey stick is the product of improper normalization of the data. This is an improvement on their previous claims, since it seems that it will be reasonably simple to test. William Connolley has looked at the data and thinks M&M are probably wrong:

But (having read their paper) I now think I understand what they think the problem is (aside: they complain about data issues with some series but I think this is beside the point: the main point they are talking about is below), and I think that they are probably wrong, based on reading MBH’s Fortran (aside: Fortran is a terrible language for doing this stuff, they should use a vector language like IDL). But anyway:

Lets for the moment assume for simplicity that these series run from 1000 (AD) to 1980. MBH want to calibrate them against the instrumental record so they standardise them to 1902–1980. 1902–1980 is the “training period”.

What M&M are saying (and Muller is repeating) is (and I quote): the data

“were first scaled to the 1902-1980 mean and standard deviation, then the PCs were computed using singular value decomposition (SVD) on the transformed data…”
they complain that this means that:

“For stationary series in which the 1902–1980 mean is the same as the 1400–1980 mean, the MBH98 method approximately zero-centers the series. But for those series where the 1902–1980 mean shifts (up or down) away from the 1400–1980 mean, the variance of the shifted series will be inflated.”
This is a plausible idea: if you take 2 series, statistically identical, but when one trends up at the end where the other happens to be flat, and you compute the SD of just the end bit, and then scale the series to this SD, then you would indeed inflate the variance of the up trending series artificially. But hold on a minute… this is odd… why would you scale the series to the SD? You would expect to scale the series by the SD. Which would, in fact, reduce the variance of upwards trending series. And also, you might well think, shouldn’t you take out a linear trend over 1902–1980 before computing the SD?

So we need to look at MBH’s software, not M&M’s description of it. MBH’s software is here, and you can of course read it yourself… Fortran is so easy to read…

What they do is (search down over the reading in data till you get to 9999 continue):

  1. remove the 1902-1980 mean
  2. calc the SD over this period
  3. divide the whole series by this SD, point by point

At this point, the new data are in the situation I described above: datasets that trend upwards at the end have had their variance reduced not increased. But there is more…

  1. remove the linear trend from the new 1902-1980 series
  2. compute the SD again for 1902-1980 of the detrended data
  3. divide the whole series by this SD.

This was exactly what I was expecting to see: remove the linear trend before computing the SD.

Then the SVD type stuff begins. So… what does that all mean? It certainly looks a bit odd, because steps 1–3 appear redundant. The scaling done in 4–6 is all you need. Is the scaling of 1–3 harmful? Not obviously.

Perhaps someone would care to go through and check this. If I haven’t made a mistake then I think M&M’s complaints are unjustified and Nature correct to reject their article.

The fact that McKitrick, at least:

  • ...exported the original raw data to Excel but somehow exported 159 columns of data into a 112-column spreadsheet. M&M failed to compare the spreadsheet to the original data and thus produced a "correction" that was riddled with errors.
  • ...do calculations and get degrees and radians mixed up, [and] get the wrong answer.
  • ...claim[s] that physics provides no basis for defining average temperature and: “In the absence of physical guidance, any rule for averaging temperature is as good as any other. The folks who do the averaging happen to use the arithmetic mean over the field with specific sets of weights, rather than, say, the geometric mean or any other. But this is mere convention.” Physics does, in fact, provide a basis for defining average temperature. Just connect the two systems that you want to average by a conductor. Heat will flow from the hotter system to the colder one until the temperatures are equalized. The final temperature is the average. That average will be a weighted arithmetic mean of the original temperatures. Which is why the folks doing the averaging use weighted arithmetic means rather than the geometric mean...
  • ...claims that temperature is not a thermodynamic variable that lends itself to statistical analysis, nor does it measure a physical quantity...
  • ...claims that : “The main problem in the debate over what the Global Temperature is doing is that there is no such thing as a Global Temperature. Temperature is a continuous field, not a scalar, and there is no physics to guide reducing this field to a scalar, by averaging or any other method...

doesn't lend one confidence in McKitrick's results.

Posted by DeLong at October 18, 2004 06:25 PM | TrackBack
Comments

I love Fortran. I didn't know anybody still used it, though.

Posted by: Bernard Yomtov at October 18, 2004 07:32 PM

(aside: Fortran is a terrible language for doing this stuff, they should use a vector language like IDL

IDL yuk! I agree vector oriented languages are better for this kind of thing, but why IDL instead of MATLAB? Or better yet Mathematica. Even MATLAB has its problems. Try taking the determinant of a 13x13 Hilbert matrix. The last time I tried it MATLAB bombs, giving a random negative answer, while Mathematica gets it right.


Posted by: A. Zarkov at October 18, 2004 07:33 PM

The problem is that temperature isn't simply definable ... the temperature of an ideal gas is proportional to its average kinetic energy and this is well defined ... however the molecular structure of extended solids carry interatomic potentials and binding energies which generally make the calculation more complex ...

A better definition of temperature would be that it is a physical quantity proportional to the mean collision pauli exclusion velocity for valence electrons assuming there is no phase change.

Whatever the complexity of the definition it is a physically measurable quantity and in fact is a Relatavistically invariant Lorentz scalar. The evidence of this is that any blackbody has a temperature characteristic emission spectrum. However under any Lorentz transformation the measurement of the blackbody emmission curve will be appropriately redshifted or blueshifted. However in the rest frame of the object it's temperature will remain unchanged.

Because of this, like mass and length we can see that temperature is a basic physical quantity. The fact that the temperature is a value averaged over a large ensemble does not make it any less physical, because the average temperature represents the peaked entropy maximization value of the underlying occupied micro-states.

In a typical average you could have many events clustered around zero and many events clustered around 100 and have the average lying at about fifty. However this doesn't happen with temperature, because at equilibrium values are tightly bunched around the given value. So the temperature value gives a real physical information about interactions rather than representing an weighted "center" to a spread of values. There is of course some variance but the action of entropy makes this relatively small in a tightly grouped trailing distribution.

Posted by: oldman at October 18, 2004 07:44 PM

The problem is that temperature isn't simply definable ... the temperature of an ideal gas is proportional to its average kinetic energy and this is well defined ... however the molecular structure of extended solids carry interatomic potentials and binding energies which generally make the calculation more complex ...

A better definition of temperature would be that it is a physical quantity proportional to the mean collision pauli exclusion velocity for valence electrons assuming there is no phase change.

Whatever the complexity of the definition it is a physically measurable quantity and in fact is a Relatavistically invariant Lorentz scalar. The evidence of this is that any blackbody has a temperature characteristic emission spectrum. However under any Lorentz transformation the measurement of the blackbody emmission curve will be appropriately redshifted or blueshifted. However in the rest frame of the object it's temperature will remain unchanged.

Because of this, like mass and length we can see that temperature is a basic physical quantity. The fact that the temperature is a value averaged over a large ensemble does not make it any less physical, because the average temperature represents the peaked entropy maximization value of the underlying occupied micro-states.

In a typical average you could have many events clustered around zero and many events clustered around 100 and have the average lying at about fifty. However this doesn't happen with temperature, because at equilibrium values are tightly bunched around the given value. So the temperature value gives a real physical information about interactions rather than representing an weighted "center" to a spread of values. There is of course some variance but the action of entropy makes this relatively small in a tightly grouped trailing distribution.

Posted by: oldman at October 18, 2004 07:44 PM

Anyone who claims that temperature is not a thermodynamic variable must be such an ignoramus that I wouldn’t waste my time looking at his Fortran source code. While global temperature is well-defined concept, we don’t have a reliable long-term measurement record as most of the weather stations were in the northern hemisphere. It’s true that at certain spots we have surrogates for a long-term temperature record, but there’s still the problem of calculating a global average.

Posted by: A. Zarkov at October 18, 2004 07:47 PM

I'm surprised some wag hasn't sent a copy of Reif to McKitrick.

Posted by: chris_a at October 18, 2004 08:36 PM

Mann et al stand by their analysis.

http://www.johnquiggin.com/archives/001888.html

There is a reason why peer review exists.

Posted by: bakho at October 18, 2004 08:37 PM

It sounds like these guys are similar to creation scientists. I can't remember their new name, but they always blather on about irreducible complexity and the like.

Posted by: anon at October 18, 2004 08:39 PM

Are these guys worse than creationists?

http://holocene.evsc.virginia.edu/Mann/EandEPaperProblem.pdf

Posted by: bakho at October 18, 2004 08:44 PM

"America Warming Towards Mauthausen"

William Calvin, in 'Nascent Minds', http://williamcalvin.com/talks/2004%20MindsBigBang_files/frame.htm
shows that if you take a slightly
longer climate view, say 50,000
years, when the majority of the
time the climate was -20F to -40F,
the recent "global warming" episode
is "real", in that the fluctuations
have largely disappeared, but not
really "real" in the sense things have
warmed up due to industrial society,
or that the latest series of runups
can thus be considered more than the
effect of sunspots and solar insolation,
changes in the global undersea currents,
or any number of other "natural" causes.

As one scientist ironically noted, the
bacteria of the world outweigh us 1B
to 1, and the methane they produce far,
far exceeds the carbon dioxide we exude.
Just from the guts of termites alone,
their production of "global warming gases"
far exceeds the automobile.

But, as Bush noted, "History? We'll all
be dead by then."

Let's get back to modern times.

http://www.bushcountry.org/pdf_files/newsoldier_intro.pdf

Even now, 40 years later, Kerry's transcript on Viet Nam leaves a chill, if you were alive then. Read it again, while you ponder Afghanistan, and then Iraq and Iran, Bush's version of Nixon's Cambodia and Laos.

Will they parade all of US before the mountain of Islamic bones, like they paraded the German people through the Nazi death camps at Dachau before the Mauthausen War Tribunals?

http://www.scrapbookpages.com/DachauScrapbook/DachauTrials/Mauthausen01.html

It's got to stop. Now.

Posted by: Tante Aime at October 18, 2004 08:46 PM

"The crimes and atrocities were not single or unconnected, but were the inevitable outcome of the basic criminal conspiracy of the NeoCon Party. This conspiracy, based on the NeoCon doctrine of racism and totalitarianism, involved murder, terrorism, and the destruction of peaceful populations in violation of the laws of war. A conspiracy is criminal either because it aims at the accomplishment of lawful ends by unlawful means, or because it aims at the accomplishment of unlawful ends by lawful means.

Therefore, such technicalities as the question whether the extermination of fellow Americans by NeoCons perpetrated before there was a state of war, would be unimportant, if you recognize as the basic crime the NeoCon conspiracy which required for success the killing of dissident liberal Americans and the extermination of American (and non-American) Moslems before and after the war had begun."

Posted by: Harry Possue at October 18, 2004 08:51 PM

You guys are econ dorks.

Posted by: Sarah at October 18, 2004 08:51 PM

Isn't it strange how people's factual beliefs on matters like this, matters that they are completely unqualified to judge, tend to follow their political beliefs? Because someone believes in protecting the environment and taking the steps necessary on a global basis, this causes them to take a certain position on obscure minutia of statistical manipulations.

Doesn't it strike anyone as a bizarre coincidence that their ideological beliefs, which after all were not at all influenced by whether to multiply or divide by standard deviation in a certain analysis, would nevertheless turn out to be able to make accurate predictions about the matter?

Posted by: CP at October 18, 2004 09:54 PM

Violating my own dictum, I did look at the Fortran code. I asked myself: why did they use Golub’s original coding for the SVD from 1969 when they could have used the SVD routine from LINPACK, which is in the public domain? Then again, why would they use Fortran when it would have been some much easier and faster to use MATLAB? Is cost is a problem there is a free version of R available for all major platforms. The whole thing looks like the work of amateurs to me.

Posted by: A. Zarkov at October 18, 2004 10:26 PM

Be all that as it may be,the simple fact is that,over the last 800K year, climate's default position has been COLD.

Posted by: Brian Boru at October 18, 2004 11:44 PM

Well, Zarkov, people use the environment that makes them comfortable. You like matlab, some like FORTRAN or Mathematica, I like IgorPRO (used to use S, suppose I should learn R). These are like flavors of ice cream: all are sweet and fattening, but some don't like pistachio flavor. It's a matter of taste.

Posted by: CD318 at October 19, 2004 12:00 AM

Yes, the whole thing looks like the work of amateurs to a lot of people.

The amateurness must be of a very high degree if it appears here, methinks...but does this point out a physics-econ rivalry?

D

Posted by: Dano at October 19, 2004 12:00 AM

What happened to Today's Reason to Not Elect George W. Bush? Or is this it?

Posted by: ogmb at October 19, 2004 12:44 AM

I hate to ask, because it seems almost insulting, but instead of arguing over where these people came from their conclusions, why hasn't anybody simply adjusted the Fortran and rerun it using the different coding they suggest. Fortran is still available, and it shouldn't be hard to get it to work with a later version, and it should be immediately obvious whether they have a point. I don't know the statistics to do it myself, or I would, but switching to another language seems completely beside the point, although either approach should work.

It just seems like people are falling into the same error they accuse these two scientists of, which is to jump to conclusions first rather than just testing the theory with detachment. Even if these guys are bozos, if there is some feeling that they have a point, why wouldn't the theory be tested, since it should be quick and inexpensive?

Posted by: Ben Langhinrichs at October 19, 2004 04:43 AM

> While global temperature is well-defined
> concept, we don’t have a reliable long-term
> measurement record as most of the weather
> stations were in the northern hemisphere.

Might also want to consider that there have only been thermometers and recorded readings for a small fraction of _recorded_ human history, which is not only a tiny fraction of the relevant period (say, last 100 million years?) but also a small fraction of just human history (30,000 years?).

Cranky

Posted by: Cranky Observer at October 19, 2004 05:17 AM

"Doesn't it strike anyone as a bizarre coincidence that their ideological beliefs, which after all were not at all influenced by whether to multiply or divide by standard deviation in a certain analysis, would nevertheless turn out to be able to make accurate predictions about the matter?"

CP, that's exactly the reason that I initially gave these global-warming-skeptic guys a lot more credit than they deserved; I didn't want my ideology driving what I thought about physics, statistics and numerical modeling.

Lambert's work, though, made it clear to me just how meretricious this stuff really is. I'm satisfied that if somebody's letting their ideology run away with their critical thinking skills, it's not me.

Posted by: Matt McIrvin at October 19, 2004 06:20 AM

I'm not interested in the details. The fact that Nature rejected the article tells me all I need to know. Even scientific journals like scoops.

Posted by: Jonathan Goldberg at October 19, 2004 07:06 AM

FYI FORTRAN is still used a lot in biological modelling.

Because a lot of old biology farts are into modelling.

Posted by: Carol at October 19, 2004 07:10 AM

It turns out that you need to measure temperature at remarkably locations (thermometers, proxys, whatever) to get reasonably accurate measures of global temperature. The reason for this is that temperatures at many locations, even those separated by large distances, are highly correlated. If you then average over time (yearly, five year, etc), you end up with a system having 5-15 degrees of freedom. You want more stations for better statistical certainty, precision, etc.

Certainly we have enough measuring stations since 1860 or so to accurately determine a global surface temperature.

Posted by: Eli Rabett at October 19, 2004 09:16 AM

http://www.nytimes.com/2004/10/19/science/19leaf.html?pagewanted=all&position=

Those Brilliant Fall Outfits May Be Saving Trees
By CARL ZIMMER

As trees across the northern United States turn gold and crimson, scientists are debating exactly what those colors are for.

The scientists do agree on one thing: the colors are for something. That represents a major shift in thinking. For decades, textbooks claimed that autumn colors were just a byproduct of dying leaves. "I had always assumed that autumn leaves were waste baskets," said Dr. David Wilkinson, an evolutionary ecologist at Liverpool John Moores University in England. "That's what I was told as a student."

During spring and summer, leaves get their green cast from chlorophyll, the pigment that plays a major role in capturing sunlight. But the leaves also contain other pigments whose colors are masked during the growing season. In autumn, trees break down their chlorophyll and draw some of the components back into their tissues. Conventional wisdom regards autumn colors as the product of the remaining pigments, which were finally unmasked. In other words, autumn leaves were a tree's gray hair.

But in recent years, scientists have recognized that autumn colors probably play an important role in the life of many trees. Yellow leaves get their color from a class of pigments called carotenoids. Another group of molecules, anthocyanins, produce oranges and reds. Trees need energy to make carotenoids and anthocyanins, but they cannot reclaim that energy because the pigments stay in a leaf when it dies. If the pigments did not help the tree survive, they would be a waste. What's more, leaves actually start producing a lot of new anthocyanin when autumn arrives.

"The reds are not unmasked-they are made in autumn," said Dr. David Lee, a botanist at Florida International University.

Evolutionary biologists and plant physiologists offer two different explanations for why natural selection has made autumn colors so widespread, despite their cost. Dr. William Hamilton, an evolutionary biologist at Oxford University, proposed that bright autumn leaves contain a message: they warn insects to leave them alone.

Dr. Hamilton's "leaf signal" hypothesis grew out of earlier work he had done on the extravagant plumage of birds. He proposed it served as an advertisement from males to females, indicating they had desirable genes. As females evolved a preference for those displays, males evolved more extravagant feathers as they competed for mates.

In the case of trees, Dr. Hamilton proposed that the visual message was sent to insects. In the fall, aphids and other insects choose trees where they will lay their eggs. When the eggs hatch the next spring, the larvae feed on the tree, often with devastating results. A tree can ward off these pests with poisons.

Dr. Hamilton speculated that trees with strong defenses might be able to protect themselves even further by letting egg-laying insects know what was in store for their eggs. By producing brilliant autumn colors, the trees advertised their lethality. As insects evolved to avoid the brightest leaves, natural selection favored trees that could become even brighter.

"It was a beautiful idea," said Marco Archetti, a former student of Dr. Hamilton who is now at the University of Fribourg in Switzerland. Dr. Hamilton had Mr. Archetti turn the hypothesis into a mathematical model. The model showed that warning signals could indeed drive the evolution of bright leaves - at least in theory.

Another student, Sam Brown, tested the leaf-signal hypothesis against real data about trees and insects. "It was a first stab to see what was out there," said Dr. Brown, now an evolutionary biologist the University of Texas. He studied 262 tree species, noting the leaf color and number of aphid species specialized on them. Dr. Brown found that trees with bright autumn leaves tended to be the victim of more specialist aphids. The correlation supported the leaf-signal hypothesis. Dr. Hamilton did not argue that the evolution of leaf signals would make all trees brilliantly colored. Instead, he said, only species that were under heavy attack experienced this evolutionary pressure.

Posted by: anne at October 19, 2004 10:38 AM

There's been a lot of harrumphing about whether temperature is or is not a thermodynamic variable, and, regardless of the issues with the rest of their analysis, I'd like to clarify the issue.

My take on what McKitrick means by this, very poorly explained, is that the concept of "mean temperature" is, *theoretically*, somewhat meaningless. Suppose you have a block of air all at temperature T0, and a block of water also all at temperature T0, well fine, the two are in thermal equilibrium, nothing much to say. Now suppose the block of water is at temperature T1. At this point what sort of useful meaning can you give to an "average" temperature of the two?
If you simply say I'll add T0 to T1 and divide by 2, well clearly this ain't gonna get you anywhere useful. However that's not the only thing you can do. You could, for example, weight each temperature by the heat capacity of the blocks of air and water. As a general principle this is likewise equally useless, but as *practical* measure, if the difference between temperatures T0 and T1 is small, this starts to tell you something useful about how much energy will flow between the the block of air and the block of water.

Reverting to the original issue, we now have multiple blocks of air, multiple blocks of water, and multiple blocks of land.
While a graph of a function giving the difference between the temperature at each 3D point of the earth at some canonical time in the earth's past, and the current temperature, would be ideal, it likewise seems to me a valid approximation to condense all that information into one number, the mean change in temperature, *provided* that the mean is calculated using appropriate weighting by heat capacities, and provided that the spread in temperature deltas does not have some strange behavior (bimodal, grossly assymetric or whatever) that renders such a collapse to a single number misleading.

Now where things become interesting is exactly what do the proponents of the "mean temperature of the earth has risen by xxx K" mean? The simplest mean would be only over the low-lying atmosphere, and this does not really need to bother with heat capacity since it's going to be pretty much the same everywhere. But this ignore the oceans, with a much larger heat capacity than air, and it ignores the issue of latent heat releasd when ice melts.

So in summary, I think McKitrick is justified in saying that, in the absence of further definition and explanation, the concept of global mean temperature is meaningless. What is not justified is his implied extrapolation from that to the claim that no further definition and explanation would make it meaningful.
As for the proponents of this number, if they do have a carefully thought-out explanation of what their number means, they should do a better job of publishing it, because I have never seen such an explanation. And if they do *not* have such an explanation, shame on them --- their calculations are as meaningless, as stupid, and as amateur as the calculations one routinely sees in political discussions of economics that routinely ignore such basics as units and that can't seem to tell the difference between a lump sum and a per-time-unit flow.

Posted by: Maynard Handley at October 19, 2004 12:34 PM

Maynard, before you harumph too much you might try and figure out why what is measured is not temperatures but temperature anomolies, which pretty much answers your harumphs.

Posted by: Eli Rabett at October 19, 2004 01:11 PM

"I think McKitrick is justified in saying that, in the absence of further definition and explanation, the concept of global mean temperature is meaningless."

Uh, no. The problems with your analysis are that

1) Ice ages happened. We have considerable historic reasons to think that these physically linked systems (atmosphere and oceans) have tended to change temperature synchonously.

2) We have obvious physical reasons to think that there is considerable heat-flow between different parts of the 'climate system' (atmosphere and oceans).

Obviously the mean global temperature is a rough gauge of a feature of an overall system that actually contains a lot of complex inner detail. But that doesn't make it meaningless.


meno

Posted by: meno at October 19, 2004 03:39 PM

> Certainly we have enough measuring stations
> since 1860 or so to accurately determine a global
> urface temperature.

And I guess my implication was that it is going to be hard to use that data to explain the warming trend around AD 1000 that allowed colonization of Greenland or the cooling trend around 1750 that allowed cannons to be hauled across rivers that don't freeze these days - and presumably didn't freeze in 1000 either.

I know that there are tree rings, snowpack samples, etc., but that is not the same as having actual temperature data.

Cranky

Posted by: Cranky Observer at October 19, 2004 04:44 PM

Why cranky, of course you need that data to calibrate the proxy records. Of course some of the instrumental data goes back to 1650 or so.

BTW, you are throwing a lot of red herrings out there, global climate is not what happens in your neck of the woods only. Is there no cliche that goes unrepeated?

Posted by: Eli Rabett at October 19, 2004 05:39 PM

......
Uh, no. The problems with your analysis are that

1) Ice ages happened. We have considerable historic reasons to think that these physically linked systems (atmosphere and oceans) have tended to change temperature synchonously.


2) We have obvious physical reasons to think that there is considerable heat-flow between different parts of the 'climate system' (atmosphere and oceans).

.....

HUH? Do people lose all the reading comprehension skills when confronted by something unexpected?
In the above I do not
* deny that the earth has changed temperature in the past (ie ice ages)
* deny that the atmosphere and ocean are coupled.
These are both pretty freaking obvious points.

What I am saying is that a naively defined "mean temperature" is a meaningless concept. It's fine to want to capture in one number the idea and the extent to which temperature has changed over the entire globe during some span of time, but doing so is a non-trivial problem, and the naive definition of how to do so, a simple area or volume-weighted average, is probably meaningless.
This does not deny that a useful definition could not be found, but it is not clear to me that such a useful definition has been found. Or, to spell it out in words you will understand:
When people talk about a global mean temperature, how do they calculate this thing? And what physically, do they think it means?

Posted by: Maynard Handley at October 19, 2004 10:11 PM

>>When people talk about a global mean temperature, how do they calculate this thing? And what physically, do they think it means?<<

Do you want a scientific paper, a Googled definition, or something from a textbook? I can - and will be happy to - provide all three if you truly wish an answer.

See, there are people out there that do this temperature thingy for a living; they leave evidence of their efforts on 8.5 x 11 sheets of paper that contain text and figures. The paper piles get digitized, too, and those who choose to do so can access the piles of paper that way as well.

Some of the piles of paper are interesting to read, fascinating even. They're not, say, Ethan Canin novels, but still.

Best,

D

Posted by: Dano at October 20, 2004 12:35 AM

Sure, please provide all three, either here or to my email address.

Thanks,
Maynard

Posted by: Maynard Handley at October 20, 2004 12:31 PM

OK Maynard. I'll get that to you as soon as I can.

D

Posted by: Dano at October 20, 2004 04:43 PM

Journal calc:

A method is developed for estimating the uncertainty (standard error) of observed regional, hemispheric, and global-mean surface temperature series due to incomplete spatial sampling. Standard errors estimated at the grid-box level [SE2 = S2(1 )/(1 + (n 1))] depend upon three parameters: the number of site records (n) within each box, the average interrecord correlation () between these sites, and the temporal variability (S2) of each grid-box temperature time series. For boxes without data (n = 0), estimates are made using values of S2 interpolated from neighboring grid boxes. Due to spatial correlation, large-scale standard errors in a regional-mean time series are not simply the average of the grid-box standard errors, but depend upon the effective number of independent sites (Neff) over the region.

P. D. Jones, T. J. Osborn, and K. R. Briffa 1997. Estimating Sampling Errors in Large-Scale Temperature Averages. *Journ. Clim.* 10:10 pp. 2548–2568.

I also recommend Karl et al. 1988. Urbanization: Its Detection and and Effect in the United States Climate Record. *Journ. Clim* 1:11 pp. 1099-1123. Among the first robust papers to scale up and average temp over large scales (also rebuts old skeptic arguments about the temp record being skewed by UHIs).

-----
[ding-dang crappy comments & no HTML tags!]

Google: http://lwf.ncdc.noaa.gov/oa/climate/research/ushcn/ushcn.html#KWQB90

or

Global mean surface temperature is a first-order measure of what we consider to be "climate," and its change is roughly proportional to [Radiative Forcing] RF. The increase in mean surface temperature per unit RF is termed climate sensitivity; it includes feedbacks within the climate system, such as changes in tropospheric water vapor and clouds in a warmer climate. The RF providing the best metric of climate change is the radiative imbalance of this land-ocean-troposphere climate system—that is, the RF integrated at the tropopause. http://www.grida.no/climate/ipcc/aviation/070.htm


further expounded upon here: http://www.grida.no/climate/ipcc_tar/wg1/214.htm
-----

Textbook:

The most recent textbook I have is Bigg, 'The Oceans and Climate' Cambride U Press, 1st ed. I'm searching for the def in here, and apparently I'm too dense to find it def in the text. My apologies, can't provide with my time pressure.

Best,

D

Posted by: Dano at October 21, 2004 11:12 PM

Journal calc:

A method is developed for estimating the uncertainty (standard error) of observed regional, hemispheric, and global-mean surface temperature series due to incomplete spatial sampling. Standard errors estimated at the grid-box level [SE2 = S2(1 )/(1 + (n 1))] depend upon three parameters: the number of site records (n) within each box, the average interrecord correlation () between these sites, and the temporal variability (S2) of each grid-box temperature time series. For boxes without data (n = 0), estimates are made using values of S2 interpolated from neighboring grid boxes. Due to spatial correlation, large-scale standard errors in a regional-mean time series are not simply the average of the grid-box standard errors, but depend upon the effective number of independent sites (Neff) over the region.

P. D. Jones, T. J. Osborn, and K. R. Briffa 1997. Estimating Sampling Errors in Large-Scale Temperature Averages. *Journ. Clim.* 10:10 pp. 2548–2568.

I also recommend Karl et al. 1988. Urbanization: Its Detection and and Effect in the United States Climate Record. *Journ. Clim* 1:11 pp. 1099-1123. Among the first robust papers to scale up and average temp over large scales (also rebuts old skeptic arguments about the temp record being skewed by UHIs).

-----
[ding-dang crappy comments & no HTML tags!]

Google: http://lwf.ncdc.noaa.gov/oa/climate/research/ushcn/ushcn.html#KWQB90

or

Global mean surface temperature is a first-order measure of what we consider to be "climate," and its change is roughly proportional to [Radiative Forcing] RF. The increase in mean surface temperature per unit RF is termed climate sensitivity; it includes feedbacks within the climate system, such as changes in tropospheric water vapor and clouds in a warmer climate. The RF providing the best metric of climate change is the radiative imbalance of this land-ocean-troposphere climate system—that is, the RF integrated at the tropopause. http://www.grida.no/climate/ipcc/aviation/070.htm


further expounded upon here: http://www.grida.no/climate/ipcc_tar/wg1/214.htm
-----

Textbook:

The most recent textbook I have is Bigg, 'The Oceans and Climate' Cambride U Press, 1st ed. I'm searching for the def in here, and apparently I'm too dense to find it def in the text. My apologies, can't provide with my time pressure.

Best,

D

Posted by: Dano at October 21, 2004 11:22 PM

You will no doubt appreciate that neither of the two "definitions" you provide above actually tell one a damn thing about what is being calculated when one calculates a mean temperature, and what the meaning of that calculation is supposed to be.

I mean, this is not rocket science. I am not asking for a list of one hundred complicated corrections that have to be performed so as to deal with real world problems when calculating this number --- I am asking for a simple, assuming we have all the data we need and that it is error-free, definition of what we then do with the data and what the physical significance of the resultant number is.
And since the academic publishing industry does not deign to makes its wares available on the web, and since I have no access to such journals in paper form, telling me "read reference ..." is a cop-out. And it's unnecessary --- what I am asking for should easily fit into no more than perhaps ten lines of text, easily cut-and-pasted or retyped.

As I said before, you have my email address. Mail me when you have something.

Maynard

Posted by: Maynard Handley at October 24, 2004 06:23 PM

>>I am asking for a simple, assuming we have all the data we need and that it is error-free, definition of what we then do with the data and what the physical significance of the resultant number is.>and since I have no access to such journals in paper form, telling me "read reference ..." is a cop-out. And it's unnecessary --- what I am asking for should easily fit into no more than perhaps ten lines of text, easily cut-and-pasted or retyped.<<

2. I cut-pasted something for you. It should get you started. The structure of your original statement tells me you don't know how the planet-wide collection network is sorted, aggregated, and analyzed. 10 lines won't do it for you. Journal reading is required.

Start with Karl, T in Journal of Climate. Do an author index and start from there. Briffa, K and Jones, P will be next. Mann, M too.

It's complicated, and little stories of why temps are meaningless mendacicize away the reality.

Sadly for you and others, as a result of being a recipient of an unsuccessful virus attack in 2002, I don't provide my e-mail addy. Sorry. How about these URLs, then, from one of the leaders in the paleoclim. field:

ftp://holocene.evsc.virginia.edu/pub/mann/MannPark1993.PDF

This is the original paper that is being demonized on the indy-funded web sites: ftp://holocene.evsc.virginia.edu/pub/mann/mbh98.pdf

An analysis using stat techniques of multiple collection systems: ftp://holocene.evsc.virginia.edu/pub/mann/MannLees1996.pdf

An expansion upon one of my cut-pasted passages above: ftp://holocene.evsc.virginia.edu/pub/mann/MannPark1996.pdf
This is one way of showing how the data are analyzed to derive the mean.

Here is the US Historical Clim Network home page:
http://cdiac.esd.ornl.gov/r3d/ushcn/ushcn.html

they are the keepers of the data. Poke around in there and see what they do. Judge for yourself whether it is sufficient to create a mean temp.

http://lwf.ncdc.noaa.gov/oa/climate/research/monitoring.html

another.

And here's another:

http://www.ncdc.noaa.gov/cgi-bin/res40.pl?page=ghcn.html

The Global Historical Climo Network.

Good luck in your quest for knowledge,

D

Posted by: Dano at October 25, 2004 08:15 AM

But, a short answer:

The temps are massaged (as in above refs and explained in detail as ref'd above, [see esp. Tom Karl]), placed into gridded squares, massaged again (as in above refs and explained in detail as ref'd above), and a temp pops out.

The physical significance is the number is used for trending analysis.

HTH,

D

Posted by: Dano at October 25, 2004 08:24 AM

2724 http://www.blackjack-1000.com
PlLAY the best blackjack-1000.com
debt consolidation only.

Posted by: blackjack software at October 26, 2004 06:15 PM

6336 http://www.i--cialis.net
Offering Cialis with overnight delivery. Also, If your looking for generic cialis this is a good site to visit.

Posted by: Generic cialis at October 26, 2004 08:44 PM

5684 jobs online part time jobs work online work online from home work at home work from home online jobs work at home jobs work at home business jobs
Full time jobs dental plan dental
insurance
dental insurance
plan
discount dental
plans
cheap dental
plans
individual dental
plans

Posted by: jobs at October 27, 2004 01:59 AM


But, a short answer:
The temps are massaged (as in above refs and explained in detail as ref'd above, [see esp. Tom Karl]), placed into gridded squares, massaged again (as in above refs and explained in detail as ref'd above), and a temp pops out.
The physical significance is the number is used for trending analysis.


So, to summarize, what is calculated is a naive mean of the atmospheric temperature as a field, just as I discussed. And just as I discussed, this is not a valid thermodynamic operation --- adding two temperatures together is a meaningless concept.
Using this as a single figure of merit for trend-analysis is probably fine (not my field), but calling it a temperature of any sort, global mean or whatever, is very problematic. That is the whole extent of my criticism --- that these guys are calculating a particular mathematical quantity and labelling it as a temperature, when that is an invalid operation. This is the complaint that McKitrick is making that I said is justified.

Read for example http://www.ncdc.noaa.gov/oa/climate/research/ghcn/ghcnoverview.html
This gives lots and lots of detail as to how data was collected, outliers and bad data removed, gaps in the record dealt with etc etc. All very nice, but there is *nothing* discussing what they mean by "mean temperature" and why they believe it's valid. There's just as automatic assumption that you can take a sequence of temperatures, find a mean, and have something that's also a temperature. What you have is something that is *meaningful* yes, for certain purposes, but which is *not* any longer a temperature but simply a figure of merit, a summary statistic.
To give an example that may perhaps make the point more clear: you can't say things like "60% of white males and 10% of black males approve of GWB so therefore his total approval rating among males is 70%" --- adding together approval rating percentages is a meaningless operation.

Now this is easily enough dealt with simply by renaming this figure of merit as something other than a temperature; or by giving some explanation of why it is believed to be valid to call it a temperature.

Posted by: Maynard Handley at October 27, 2004 11:23 PM

For some reason I cannot reply from home to your answering to the short answer (like I knew you wanted).

Let's see if this goes thru...

D

Posted by: Dano at October 28, 2004 09:52 PM

No, no and no.

I gave a short answer, like what you wanted.

I tried to give you a longer answer (more correct one) and you didn't want to bother reading it.

The short answer is necessarily incomplete. Don't jump on that answer to make your point. Address the long answer and the journal articles or don't bother.

I pointed you to the journals to gain a richer understanding and you don't want to do the work of reading hard stuff.

Read the source material I gave you to understand why your post is incorrect.

Plus, your backing McKitrick when McKitrick can't do simple statistics and math is a big clue, also.

http://cgi.cse.unsw.edu.au/~lambert/cgi-bin/blog/2004/10#mckitrick8
http://cgi.cse.unsw.edu.au/~lambert/cgi-bin/blog/2004/08#mckitrick6


Good luck in your quest for knowledge.

D

Posted by: Dano at October 29, 2004 08:57 AM

4507 Kona Coffee Starbucks Coffee Jamaica Blue Mountain
Coffee
coffee maker gourmet coffee green mountain coffee kenya coffee organic coffee specialty coffee folgers coffee coffee brewers costa rica coffee Tullys Coffee Millstone Coffee coffee grinder http://www.coffee-delivered.com

Posted by: coffee at November 3, 2004 02:17 AM

8965
directv
direct tv directv satellite direct tv satellite directv dvr direct tv dvr direct tv tivo directv tivo directway
direcway
directway internet directway satellite direcway internet direcway internet free hbo free cinemax free dvd player satellite radio http://www.satellitetvboutique.com

directv
direct tv
directv satellite direct tv satellite directv dvr direct tv dvr direct tv tivo directv tivo directway
direcway
directway internet directway satellite direcway internet direcway internet free hbo free cinemax free dvd player satellite radio http://satellite-tv.cjb.net

Posted by: direct tv at November 10, 2004 10:31 AM

Very interesting point of view fdgfdg.

Posted by: angeln at November 10, 2004 01:52 PM

2 much spam in here :-(

Posted by: stadtplan at November 10, 2004 10:23 PM

Hi-Allll!

Posted by: Hi-All at November 12, 2004 03:16 AM

Online Loss Weight http://www.pharmacy-online-center.com/english/index.html

Posted by: Online Loss Weight, Online Phentermine loss weight pills at November 13, 2004 01:56 PM

Online Loss Weight http://www.pharmacy-online-center.com

Posted by: Online Loss Weight, Online Phentermine at November 13, 2004 01:58 PM

Online Loss Weight http://www.pharmacy-online-center.com/german/index.html

Posted by: Online Loss Weight, Online Phentermine diet pharma at November 13, 2004 01:58 PM

Online Loss Weight http://www.pharmacy-online-center.com/russian/index.html

Posted by: Online Loss Weight, Online Phentermine number 1 at November 13, 2004 02:00 PM

Online Loss Weight http://www.pharmacy-online-center.com/italian/index.html

Posted by: Online Loss Weight, Online Phentermine best at November 13, 2004 02:00 PM

Online Loss Weight http://www.pharmacy-online-center.com/french/index.html

Posted by: Online Loss Weight, Online Phentermine diet at November 13, 2004 02:02 PM

Online Loss Weight http://www.pharmacy-online-center.com/portugalian/index.html

Posted by: Online Loss Weight, Online Phentermine at November 13, 2004 02:03 PM

Online Loss Weight http://www.pharmacy-online-center.com/spanich/index.html

Posted by: Online Loss Weight, Online Phentermine pills at November 13, 2004 02:03 PM

Hi-Allll! I am a filthy comment spammer. See my site to learn how I spammed you: http://gir.dyndns.org:3400/

Posted by: Comment spammer at November 14, 2004 11:27 AM

Explorer will always be the best browser. Don"t you dare install Firefox or we will sue you.

IE Team

Posted by: IE Team at November 14, 2004 12:30 PM

All these are spmmaers- fuck them all!
ritazohar@012.net.il amirgans@hotmail.com crashoverrride@hotmail.com remove@nitzanhr.com zshen@shenmedia.com grass@lm.com futureofus@yahoo.com yanivp@pc.co.il e-commerce@shoplet.co.il kfir@vparty.co.il michal@vparty.co.il guy@vparty.co.il guyel111@hotmail.com gil_stern@yahoo.com gilad19@israsrv.net.il noam@work2you.net webmaster@sweethome.co.il amos@work2home.net amosref@walla.co.il amosref@yahoo.com http://www.shoplet.co.il/ http://quimka.com quimka@quimka.com info@taocomputers.com ran_tao@netvision.net.il http://taocomputers.com orders@d-d.co.il d-d.co.il@internic.co.il shlomi@e-shop.co.il arcadia@inter.net.il newsletter1@e-shop.co.il newsletter@e-shop.co.il sales@dynamic-studio.co.il arkadyr@dynamic-studio.co.il robin@dynamic-studio.co.il eduard@dynamic-studio.co.il iggi@dynamic-studio.co.il tempter@dynamic-studio.com sergi@dynamic-studio.com abuse@hot.ee etgar@69.56.135.2 abuse@69.56.135.2 etgar@hit8.hotlog.ru abuse@224.96.64.80 zvika@etgar.org.il

Posted by: Fuck spammers at November 14, 2004 11:02 PM

approved. largest supplies The get http://www.fast-carisoprodol.com AstraZeneca Bayer vitamins. 30 night free largest delivery Ortho-McNeil Order of over of Pfizer variety without days or FDA Carisoprodol The buy and Merck now

Posted by: Carisoprodol at November 15, 2004 08:22 AM

supplies night http://www.somasafari.com Order Soma Pfizer AstraZeneca largest now variety Merck The of Bayer get delivery and free of The buy Ortho-McNeil 30 days over largest FDA vitamins. without approved. or

Posted by: Soma at November 15, 2004 09:24 AM

1 2
3 4
5 6
7 8
9 10
2 much spam in here :-(

Posted by: Ben Maluchska at November 15, 2004 06:02 PM

nice site, was glade to see it...

Posted by: steff sine at November 16, 2004 02:53 PM

Online Loss Weight http://www.cigarettes-bestbuy.com/marlboro-cigarettes.html

Posted by: Online Loss Weight, Online Phentermine at November 17, 2004 12:59 AM

Online Loss Weight http://www.cigarettes-bestbuy.com/cheap-parliament-cigarettes.html

Posted by: Online Loss Weight, Online Phentermine at November 17, 2004 01:02 AM

Online Loss Weight http://homeloan.payday-loans-4us.com/

Posted by: Online Loss Weight, Online home loan at November 17, 2004 01:06 AM

Online Loss Weight http://www.payday-loans-4us.com/cheap-payday-loan.html

Posted by: Online Loss Weight, Online cheap payday loan at November 17, 2004 01:07 AM

Online Loss Weight http://www.payday-loans-4us.com/instant-payday-loan.html

Posted by: Online Loss Weight, Online instant payday loan at November 17, 2004 01:08 AM

Online Loss Weight http://www.cigarettes-bestbuy.com/camel-cigarettes.html

Posted by: Online Loss Weight, Online Phentermine at November 17, 2004 01:08 AM

Online Loss Weight http://www.cigarettes-bestbuy.com/

Posted by: Online Loss Weight, Online buy cheap cigarettes at November 17, 2004 01:10 AM

Online Loss Weight http://debt.payday-loans-4us.com/

Posted by: Online Loss Weight, Online debt consolidation company at November 17, 2004 01:10 AM

Online Loss Weight http://www.payday-loans-4us.com/

Posted by: Online Loss Weight, Online PayDay Loans at November 17, 2004 01:10 AM

Online Loss Weight http://www.cigarettes-bestbuy.com/cigars/

Posted by: Online Loss Weight, Online Phentermine at November 17, 2004 01:11 AM

Online Loss Weight http://debt.payday-loans-4us.com/debt-consolidation-counseling.html

Posted by: Online Loss Weight, Online debt consolidation counseling at November 17, 2004 01:11 AM

Online Loss Weight http://www.cigarettes-bestbuy.com/winston-cigarettes.html

Posted by: Online Loss Weight, Online Phentermine at November 17, 2004 01:11 AM

Online Loss Weight http://www.payday-loans-4us.com/payday-loan-online.html

Posted by: Online Loss Weight, Online payday loan online at November 17, 2004 01:11 AM

Online Loss Weight http://www.cigarettes-bestbuy.com/salem-cigarettes.html

Posted by: Online Loss Weight, Online Phentermine at November 17, 2004 01:12 AM

Online Loss Weight http://www.payday-loans-4us.com/military-payday-loan.html

Posted by: Online Loss Weight, Online military payday loan at November 17, 2004 01:16 AM

yeah nice site, was glade to see it....

Posted by: steffi sineo at November 18, 2004 05:34 PM

All of the DVD rentals are delivered directly to your mailing address by mail with a return postage-paid envelope from our distribution centers. The average shipping time is 3-5 business days for all movie rentals. Some DVD rental titles are specially ordered for members, which may cause a delay in the member receiving the movie rental. Step 1) Create a list online of DVD rentals by mail you want to see. Step 2) We will rush you 3 DVD movies from your list in 1-3 days. Step 3) No Late Fees! ? Keep each DVD as long as you want. Step 4) Return viewed DVD movie rentals to get new ones from your list using our prepaid return envelopes. We looked at several services offering Online DVD Rentals. Blockbuster Rentals are fast and convenient and are quickly becoming very popular. You can rent movies online from Netflix or a few other sites that deliver your movies free of charge for a monthly fee. You never get late fees with these GamezNFlix online rental services which saves ME a LOT of money on late fees.

Posted by: DVD Rental at November 20, 2004 06:27 PM

414 http://www.top-texas-hold-em.com

texas hold em

Posted by: texas hold em at November 22, 2004 03:44 PM

6993 Kona Coffee Starbucks Coffee Jamaica Blue Mountain
Coffee
coffee maker gourmet coffee green mountain coffee kenya coffee organic coffee specialty coffee folgers coffee coffee brewers costa rica coffee Tullys Coffee Millstone Coffee coffee grinder
http://www.coffee-delivered.com

You only get one set of teeth. Take care of them with a good
dental plan.
Dental
insurance is
money well spent. I sleep better since I signed up for my new dental insurance
plan.
Get yours at: http://dental-insurance-plan.freeservers.com/

individual
dental
plans
You only get one set of teeth. Take care of them with a good
dental plan.
Dental
insurance is
money well spent. I sleep better since I signed up for my new
dental insurance
plan.
Get yours at:
http://www.dental-plan-source.com
individual dental
plans
You only get one set of teeth. Take care of them with a good
dental plan.
Dental
insurance is
money well spent. I sleep better since I signed up for my new
dental insurance
plan. Get yours at:
http://www.e-dental-insurance-plans.com/
individual dental
plans

Posted by: individual dental plans at November 23, 2004 05:24 PM

Online Loss Weight http://realestate.world-datacenter.com

Posted by: Online Loss Weight, Online search online information about at November 23, 2004 09:53 PM

Online Loss Weight http://montecarlo.online-gambling-portal.cc

Posted by: Online Loss Weight, Online play online gambling at November 23, 2004 10:02 PM

877 http://www.top-texas-hold-em.com

texas hold em

Posted by: texas hold em at November 25, 2004 08:59 PM

Online Loss Weight http://antidepressants.pharmacy-online-center.com

Posted by: Online Loss Weight, Online Phentermine at November 25, 2004 10:59 PM

Online Loss Weight http://pills.pharmacy-online-center.com

Posted by: Online Loss Weight, Online Phentermine at November 25, 2004 11:02 PM

Online Loss Weight http://muscle.pharmacy-online-center.com

Posted by: Online Loss Weight, Online Phentermine at November 25, 2004 11:04 PM

Thanks for that insightful comment! It makes interesting reading, especially when I need a payday loans.

Posted by: payday loans at November 26, 2004 04:24 AM

NIce Feature Site

Posted by: Upskirts Mania at November 27, 2004 03:28 AM

1715 Very well said chappy.

Posted by: consoldiate debt at November 27, 2004 08:40 PM

You ar the the one I love honey where are you now just come back we will live very happily

Posted by: Wet Lesbian at November 28, 2004 12:42 AM
Post a comment









Remember personal info?