Quote of the week #9 – "negative thermometers"

negative thermometers do exist but it is the INVERSE thermometers that are only good on star trek!!!!

Positives and negatives are used to make averages. What is wrong with the absolute value approach? No smiley face on this keyboard.

Isn’t this a bit like the square root of a negative number?
Great in theory, but impossible….

I don’t really see the issue here. Or perhaps I’m missing something. On the unit shown it looks to me as if the lowest temp readable is -100 and the unit is constructed so as to be mounted vertically with the bulb up. What’s the big deal?

Aaron: I’m not sure you were serious. But if you are:
Mercury freezes at about -39C. That rules it out for -100 F. or C.
I am not sure what inverting a mercury thermometer would do. Liquids in very fine tubes do unusual things. But I suspect it would just ooze to the bottom.

I thought at first that it might make sense if the data series were all normalized. Take the series and subtract the mean of that series so all are zero average. Then to PCA / PLS to get T= c0 + c1 T1 + c2 T2 + … Etc.
But after thinking about it a bit you end up with something wrong if you allow negative coefficents. I guess you should set up and solve a quadratic programming problem to constrain the coefficients to be positive, assuming you are trying to model a set of data…

Perhaps just to look at this another way.
It is the fate of all organisations that their future will ultimately run through a cycle or if they are fortunate, run through a number of cycles, and this applies equally to the media including currently prestigious science publications like “Nature”.
Some manage to recover and go onto new heights.
Others disappear, never to be heard from again.
An organisation’s reputation is acquired over a considerable length of time by sticking to a set of very strong principles.
Then changes start to take place.
Sometimes there is not enough flexibility in the thinking of the principals to adapt to new situations and public mores.
Sometimes there are personnel changes to a leadership that is no longer capable of upholding the previous standards or who take an organisation down a route that proves to be a dead end.
With the publication of a number of unverified and unverifiable climate papers by Nature such as Steig’s and Mann’s papers and due in part to Nature refusing to uphold it’s own requirements for full disclosure of all relevant assumptions, calculations and data for these climate papers, what we may be seeing is the start of the long downhill decline of “Nature” as major and prestigious science publication.
This together with the refusal to publish climate papers, apparently due only to a fixed ideological position on climate change, that questions the current “consensus” on global warming and climate change is surely indicative that Nature has entered the nether world that leads to that long slow decline into irrelevancy.
In a couple of decades, Nature may no longer exist as a science publication even at the lowest common denominator.

Perhaps they measure latent temperature? Or dark temperature? Or missing temperature? Or magnetic temperature (good for your health!).
How about bull-o-meter?

This reminds me of the defence offered by the NAS panel of MBH98&99, the methodology might be ropey but the conclusion could well be right. Its not satisfying science but statistical analysis generally isn’t satisfying.
Hence I tend to hesitate when presented with arguments about “negative thermometers” as we’re not talking about the real world observations but the artificial treatment of statistical analysis. The significance of such practises is probably little I suspect.
I guess what I’m saying is that I don’t treat any of this as being of significant importance. I’m dubious of any particular interpretation being held as definitive. I would like to see a more open presentation of statistical analysis including access to all data and source code in order to facilitate a debate on how best to apply a limited interpretative skill, what assumptions are made in order to facilitate a statistical analysis and whether there are errors made in the construction of the analysis (such as general noise producing a particular result).

Negative weightings of temperature data does seem odd, but why wouldn’t you use a negative weight of normalized data IF you were compensating for a known warming bias at a particular station? Isn’t the issue with Steig et al’s paper their use of negative weights with no explanations?

Oh, come on people! Everyone knows that negative temperatures (in degrees Kelvin) are necessary to explain how the Dark Suckers in your house work. We only call them “light bulbs” because that’s what you get after they suck up all the dark.
;^)

@ Rich (05:33:55) :
Of course it is doctored. [snip – just a bit over the top, see my response to Rich above. Anthony]
/sarcasm

Although not applicable here, negative absolute temperature (in degrees kelvin) does exist. It is produced by a population inversion as seen in stimulated emission (lasers). The energy required is quite finite. It must be noted that negative temperatures are not below absolute zero, but above infinite temperature. At infinite temperature all levels are equally populated. This follows from a statistical mechanics definition of temperature and does not apply to a largely classical system as used in climatology. Here, temperature can be defined as internal energy per degree of freedom, in appropriate units, which will never be negative. You can, of course, have negative celsius temperatures which could be averaged in the same way as any group of numbers. Except that an average temperature is a meaningless concept! Temperature is an intensive and not an extensive parameter such as volume. If you can’t define a total temperature, you cannot define an average. I also assume that all celsius temperatures in antarctica are negative, so I have no idea what Steig et al are doing. Some sort of strange weighting, no doubt. Cheers.

Come on people, the thermometer is half full, ……… :^]

Pubic ridicule is the best antidote to poisonous lying pretending to be Divine Truth. Thanks to Anthony, Jeff Id especially, and Ryan O; now commenters can pile it on and readers around the world can join in with guffaws.
Personally, I like this one: John W. (05:54:19) :
Oh, come on people! Everyone knows that negative temperatures (in degrees Kelvin) are necessary to explain how the Dark Suckers in your house work. We only call them “light bulbs” because that’s what you get after they suck up all the dark.
;^)
And I subscribed to Nature for years; but I stopped a nmumber of years ago. I wonder how many others want a reputable science publication, not “yellow journalism”.
Finally, Steig’s pretense at measuring Antarctica’s “warming” with only a few, mostly penninsula, stations along with switching some others, I find to be most reprehensible — even with turning some thermometers upside down to get desired results. [snip]

My father was interviewed by an Italian organization. They put him comment to comment with Stephen Schnieder and they seem to give equal time to both. Perhaps the paradigm is shifting. I do not speak Italian, so I would welcome any translation and information as to whether this is as even handed as it seems. Too good to be true? http://www.avoicomunicare.it/#topvideo

I guess its like a proctologist looking for throat cancer.

George (06:50:22) :
Of course it is doctored. Do you think any member of the AGW church would even ‘touch’ a mercury thermometer? Touch alone is enough to kill you.
They won’t touch a mercury thermometer but a CFT light is perfectly safe. LOL

I just got in some big crazy battle about how it might be reasonable to have inverted thermometers. There is absolutely no hope that I’ll ever figure people out.

Someone on the air vent provided a layamn’s explanation.
The method used by Steig and the others is a form of weighted averaging. It is like trying to determine an average price for gold by taking the values in all markets and weighting them by the size of their own market market. This creates a valid average
If someone tried to do this and ended up with a negative weight somehow then it would be clear that their is something wrong with his method. Thus the Steig method is in error because it has negative weights. The explanation for the reason that it is wrong is that the method that he uses to determine the weights is sensitive to noise and by truncating the number of PCs at 3, he is introducing noise.
This is the explanation provided at the Air Vent as I “understand” it.
The rationalizations being given for possible negative covariances are wrong because this calculation has nothing to do with co-variances. This is a simple averaging.

Antonio San (12:45:24) :
…
Climate–carbon modelling experiments
…
OK. “Modeling experiments.” When I read that, The meaning I get is that they are running an invalidated, unverified model, and calling the run an experiment. Do I understand that correctly? Are these people really asserting, either explicitly or implicitly, that they don’t have to look at the real world because they can get all the information they need by simulating the real world? If this is what they are doing, could someone please explain in what way, shape form or manner their activity has anything to do with science?

Speaking of negative thermometers, in standard optical ray tracing software it is conventional to treat ordinary mirrors as being a boundary between two media with refractive indices of +1 and -1, so then if you do that, you can calculate the ray propagation by treating it as simply a refraction across a +1/-1 boundary.
So what if you have a mirror inside a piece of glass; so the refractive index before the mirror is +n, why not have the index following the mirror be -n, so the ratio of the two is -1 as before.
So I once asked a ray tracing software vendor, why his program would crash, if I entered a negative refractivew index (other than -1). He replied that it was not relevent since negative refractive index materials do not exist; so he refused to change his code (fortran) so it accepted -n just as easily as it accepted -1.
So much for him and his now off the market program; we n0ow do have optical materials with negative refractive indices, and the refracted ray, now lies on the same side of the normal as the incident ray.
You reject a good idea, and you go out of business.

That thermometer is used in experiments regarding Thiotimoline
http://wapedia.mobi/en/Thiotimoline

There are 2 places to put a negative thermometer and one is in the trash basket.

As if to add to those worries about negative temperatures, an Associated Press article today warns about the onset of negative winds. It cites a “preliminary” study authored by Sara Pryor, an atmospheric scientist at Indiana University, and Eugene Talke, a professor of atmospheric science at Ohio State University (sic). Their study “… will be published in August in the Journal of Geophysical Research.” It “…does raise a new side effect of global warming” by claiming that global warming is creating big drops in wind speed (maybe 10% over a decade) in places such as “Ohio, Indiana, Michigan, Illinois, Kansas, Virginia, Louisiana, Georgia northern Maine, and western Montana.” Worse yet, the article questions whether global warming therefore may make wind farms uneconomic in places. Imagine that! Is there anything left that can’t be blamed on “global warming?
Bob

Imaginary temperatures might not exist but they can be useful nevertheless, at least if you also allow imaginary $s. Suppose the complex temperature rise is C=RC+iIC and the complex cost of abatement per complex degree C is the complex dollar amount $=R$+iI$. The total real cost of abatement is the Real part of C*$. Since C*$ = RC*R$ + i(RC*I$+IC*R$)-IC*IR the Real part of $*IC is R$*RC-IC*I$. Since IC and I$ are the imaginary components we can make them anything we want without altering the Real parts of C and $. So if we choose the imaginary components such that IC*I$=RC*R$ the total real cost is zero, which is very cheap. As for the leftover imaginary component we can pass that on to the IPCC. Since the benefit of funding this community is imaginary let it be iIB per imaginary $. The consequent funding to the IPCC would therefore be -IB*(RC* I$+IC*R$). Noting that the imaginary components IB, I$ and IC can be anything at all, including negative numbers, this funding can be anything at all. The Linear Theory of complex temperatures and $s therefore agrees exactly with the possible future costs of abatement of “carbon” emmisions and funding of the IPCC:
Real costs = $0
Real funding to IPCC= $anything you want

Lots of fun comments but I think that the main reason they use Celsius is that it looks more impressive. When the worlds temperature goes from 16 degrees to 17 degrees that doesn’t sound like much, but compare to how it sounds going from 289 degrees Kelvin to 290 degrees Kelvin.
Think of how that chart would look going from 271 to 279.
A year or two ago I read on-line a newspaper article that claimed that local temperatures had doubled from +2 C. to +4 C. Oh the humanity!!!!!!!!!!!
PS: I probably got the link from here…

“The consequent funding to the IPCC would therefore be -IB*(RC* I$+IC*R$). Noting that the imaginary components IB, I$ and IC can be anything at all, including negative numbers, this funding can be anything at all. The Linear Theory of complex temperatures and $s therefore agrees exactly with the possible future costs of abatement of “carbon” emmisions and funding of the IPCC:
Real costs = $0
Real funding to IPCC= $anything you want”-DavidC
I hope there is some validity to this theory since I really like the idea of paying carbon taxes with imaginary money. Unfortunately, all our money is likely to be imaginary on our present track. Perhaps that is the necessary corollary to imaginary science.