45 thoughts on “Busy times….Open thread.”

1. Enjoy your Thanksgiving dinners – and recommend some additional Vitamin C. See: Vitamin C and High-Fat Meals

   In the subjects taking vitamin C, the study authors could find no significant change in the flow-www.ed dilatation, indicating that vitamin C could protect against the negative effects high triglyceride levels have on blood vessels.

2. Not only a whiz with statistics, you bake pie too!

3. David– My mom made me enroll in 4-H in grade school. In the suburbs, they concentrate on domestic arts; in rural areas, animal husbandry. I lived in the suburbs. I can bake, sew, knit, crochet etc. (I should add that my younger sister always won the 1st prize ribbons. I only won in crochet.)

4. Indeed a well rounded education – my daughter is similarly in 4-H and my wife has trained her to take the blue ribbons in sewing whenever she enters. (We decided against the goats.)

   Thinking of climate statistics, you may wish to peruse:
   Falsification Of The Atmospheric CO2 Greenhouse Effects Within The Frame Of Physics by Gerhard Gerlich and Ralf D. Tscheuschner. In 114 pp, they explore the foundational physics involved and have a quite a few comments to say on the lack of rigor and physical basis in common climate “science”.

   In the first half, see modeling CO2 on a mass rather than volumetric basis. Averaging the fourth root of the irradiation rather than conventional methods. Lack of any physically well defined “global temperature” etc. etc.

5. David–
   There is a physical basis to the enhanced greenhouse effect.

   Coming from a engineering background in with a ph.d. in a topic in multiphase flow, I have no difficulties with climate scientists defining a “global mean surface temperature” (gmst). Speaking generically, various sorts of averages have uses. Area averages are used in multiphase flow. Admittedly, “gmst” isn’t important, but if area averages are meaningful in fluid mechanics, why would they cease to become meaningful in climate science?

   Admittedly, any average be mis-used, and there are some uses of “gmst” that one might criticize as over simplified or inappropriate. But saying “global mean surface temperature” lacks a well defined physical or mathematical meaning is going much to far.

6. I put up a post on wattsupwiththat which demonstrates a method for adjusting temperatures for the ENSO and the AMO. It is a simple least squares regression method but produces pretty good results. Spreadsheets are included for others to try it out and/or improve on it.

   http://wattsupwiththat.com/2008/11/25/adjusting-temperatures-for-the-enso-and-the-amo/

7. Hmmph. Even GISS agrees that surface temperature is “elusive”: http://data.giss.nasa.gov/gistemp/abs_temp.html

   “To measure SAT we have to agree on what it is and, as far as I know, no such standard has been suggested or generally adopted.”

   The air to solid/liquid interface surface on Earth is *fractal* (think of a forest). What’s more, it changes constantly: if I had a Stevenson Screen in my yard, it would be a foot closer to the surface in winter (given average snow cover) than summer. Unless we measure from the dirt, in which case the “surface air” would be inside the snow in some places!

   Sure, in theory, a rigorous definition could be made (it would be quite complex). But it hasn’t, and this allows the people doing the measuring to vary the methodology a lot and still claim they’re measuring “the same thing.”

8. Nice analysis of the sceptical categories 🙂

   I am firmly in the a) category (with among others D.Koutsoyiannis etc) if one replaces the word “noise” by the words “pseudo trends in chaotic dynamics” . “Noise” has a stochastical flavor and there is nothing random about deterministic chaos .

   Incidentally I will explain why spatial averages have with a high probability no relevant meaning for climate dynamics .

   When you describe the fluid dynamics in terms of the flow function , the PD equations you get are formally equivalent to hamiltonian equations describing a dynamical system in the phase space .
   The flow function is then equivalent to the hamiltonian .
   Follows that formally the solutions of the hamiltonian equations in the phase space have the same behaviour as the solutions of the flow equations in the ordinary space .
   In other words the spatial variables in fluid dynamics are equivalent to phase space variables in hamiltonian mechanics .

   Now it is a known thing that dynamical systems may be chaotic and present attractors and even fractal attractors in the phase space .
   The equivalence I mentionned predicts therefore that there will be SPATIAL structures in fluid dynamics that willl present exactly the same features (in some/many cases) .
   Surprise , surprise – they indeed do !
   Chaotic spatial structures fractal or not can be found everywhere – trees , clouds , Rayleigh-Benard flows , snow flakes etc .

   So now we can use this equivalence the other way round .
   It is a trivial result that “averaging” orbits in the phase space is mostly unphysical nonsense destroying information about the dynamics of the system .
   The equivalent of the nonsense in the ordinary space would be like making a spatial average of the velocities of a vortex , finding 0 and saying that the fluid doesn’t move or , indeed , spatially averaging temperatures and saying that the Earth is an isothermal body .

   I’d like to stress here that all that doesn’t apply to temporal averaging that is a completely different question related to quasi ergodicity assumptions .
   Time is never a coordinate of the phase space while it is a coordinate of space-time .

   One could object that sometimes one does spatial averages in fluid dynamics .
   Well yes – if there is isotropy and homogeneity what is rarely the case .
   Those rare cases are btw equivalent to dynamical systems that are not chaotic and as not everything is chaotic , it is not surprising that such cases exist too .
   Now it is not really difficult to get (qualitatively) convinced that the temperatures are chaotic . Look at the spatial structure of the isotherms and their dynamics – isotropy and homogeneity would be the last words one would choose to describe them .

   What would it take to “silence” the a) category ?
   Well as the weather is chaotic , the climate is it even more but only on a spectrum of time scales that are larger . Actually the climate should be much more complex than the weather because it plays on time scales where kick in processes that are neglected (because considered constant) when analysing the short time scales .
   So I ask for the proof that there exists some “window” in the time scale spectrum where the Earth’s dynamics magically stops being chaotic and becomes as well behaved as a laminar flow eventually slightly gaussian perturbated .
   Good luck for whoever would try (actually nobody does) because it would prove wrong most of what is currently established about dynamics 🙂
   Untill then , space averaging temperatures has to be considered as unphysical nonsense even if mathematically feasible .

   Caveat .
   I didn’t use once the word CO2 or more generally the “infamous” greenhouse gases .
   It is really trivial to establish qualitatively with a rather basic QM argument that increasing CO2 EVERYTHING ELSE BEING EQUAL increases the surface temperature .
   But this observation is not more or less relevant than the observation that winter decreases temperatures , that clouds stop sun’s rays or that freezing water liberates energy .
   What counts is what all those things do together and what they do together are things like ice ages and El Ninos that are pseudo periodical chaos .

9. Luci
   Appreciate your expertise in fluid mechanics. I have only dabbled in a bit of reactive CFD modeling combustion.
   Note the “slight” difference between “global temperature” and “global mean surface temperature”. The latter could be empirically measured to some degree of uncertainty. However, I expect that few if any have examined the full “uncertainty” of the measurements. The Type B systematic bias is probably very large due to lack of fully accounting for the Urban Heat Island Effect. See NIST Essentials of expressing measurement uncertainty

   On “global temperature” See: Section 3.7.7 in
   Falsification Of The Atmospheric CO2 Greenhouse Effects Within The Frame Of Physics by Gerhard Gerlich and Ralf D. Tscheuschner. pp 66-67

   3.7.7 Non-existence of a global temperature
   In the preceding sections mathematical and physical arguments have been presented that the notion of a global temperature is meaningless. Recently, Essex, McKitrick, and Andresen showed [169]:
   “that there is no physically meaningful global temperature for the Earth in the context of the issue of global warming. While it is always possible to construct statistics for any given set of local temperature data, an innite range of such statistics is mathematically permissible if physical principles provide no explicit basis for choosing among them. Distinct and equally valid statistical rules can and do show opposite trends when applied to the results of computations from physical models and real data in the atmosphere. A given temperature field canbe interpreted as both `warming’ and `cooling’ simultaneously, making the concept
   of warming in the context of the issue of global warming physically ill-posed.”
   Regardless of any ambiguities, a global mean temperature could only emerge out of many local temperatures. Without knowledge of any science everybody can see, how such a changing average near-ground temperature is constructed: There is more or less sunshine on the ground due to the distribution of clouds. This determines a field of local near-ground temperatures, which in turn determines the change of the distribution of clouds and, hence, the change of the temperature average, which is evidently independent of the carbon dioxide concentration.
   Mathematically, an evolution of a temperature distribution may be phenomenologically described by a differential equation. The averages are computed afterwards from the solution of this equation. However, one cannot write down a differential equation directly for averages.

   [169] C. Essex, R. McKitrick, B. Andresen, “Does a Global Temperature Exist?” J. Non-Equil. Thermod. 32, 1-27 (2007)

   Enjoy exploring the distinctions, and those of TomVonk (Comment#6942).

10. John Lang– I saw your post. I haven’t had time to read it in detail!

    Josh-

    Sure, in theory, a rigorous definition could be made (it would be quite complex). But it hasn’t, and this allows the people doing the measuring to vary the methodology a lot and still claim they’re measuring “the same thing.”

    Yep. The thing is: the claim that there literally is no physically meaningful definition of gmst, and the claim that what is measured is not the same as the gmst are different claims. The exact same difference complicates other fields!

    Tom–

    It is a trivial result that “averaging” orbits in the phase space is mostly unphysical nonsense destroying information about the dynamics of the system .

    Of course averaging of any sort destroys information. That’s true in everything– including fluid mechanics and climate science (which involves fluid mechanics.)

    But, once again, this doesn’t mean there is no physical definition of an average of something. It means that after averaging, we may have lost information required to predict things and consequently errors accumulate. In fact, that’s what happens when we do CFD– Reynolds Averaging of the NS equations kinda-sorta works, and kinda-sorta doesn’t.

    It’s recognized that the idea of averaging the PDE’s themselves is the problem. But that doesn’t mean averages don’t have physical meaning.

    David-
    Who literally uses “global temperature” to describe or predict anything? I’m sure there are some who use the two word term as short hand for something– always some sort of average. The average can be defined and that definition must be provided in context of someones particular predictive or descriptive model.

    However, one cannot write down a differential equation directly for averages.

    We do in engineering! We average the Navier Stokes we get “Reynolds Averaged Navier Stokes”. 🙂 TomV can tell you some of the problems while I vanish to eat turkey. But, we can write down differential questions for averages!

11. From various sources I understood the temperature of the troposphere to be the most important in distinguishing between CO2 driven AGW and the climate realists models.

    A comparison of tropical temperature trends with model predictions Douglass et al.
    <a href=”http://www.climateaudit.org/?p=3161March Radiosonde Data

    Global Warming as a Natural Response to Cloud Changes Associated with the Pacific Decadal Oscillation (PDO) Roy Spencer.

    Thus Ross McKitrick proposes his T3 Tax.
    Call their tax
    T3 Tax
    Ross McKitrick Publications & Papers

    The Troposphere temperature has the advantage of using satellite data and thus avoiding most of the problems of the Urban Heat Island Effect.

12. Lucia

    I believe that this is a kind of fine argument of what “unphysical” means with respect to “unmathematical” .
    What I mean is that for some mathematical transformations T of a physical variable X , T(X) is not measurable (as opposed to computable) AND if X obeys some natural law L such as L(X) = 0 then L(T(X))is not well defined .
    In this case I call T(X) “unphysical” even if T is mathematically well defined .
    Space averaging is an example of such T .
    Time averaging not necessarily .

    David

    Sigh . I can’t guess why Lucia would precisely point to me to talk about RANS while she is going to enjoy herself and eat things .
    But , well , yes she is kinda … right .
    It is actually rather easy and relates to the above comment .
    A natural law is generally a differential operator O acting on some function f such as O(f) = 0 .
    Now you may always define a transformation of f , T(f) = y – that’s called a change of variable .
    If you can inverse T , the result is trivial : f = T^(-1) (y) and you have O(T^(-1) (y)) = 0 what is a differential equation in y = T(f) .
    If you cannot inverse T (case when T is an average because by integrating f you lost almost all information about f) you will pay it by a certain number of arbitrary unknown functions Gi such as f = g[T(f) , Gi ]
    And then you can plug it again in the natural law O(f) = 0 and obtain O[g[T(f) , Gi ]] = 0 what is again a differential equation featuring T(f) and the Gi’s .
    That’s exactly what Reynolds averaging Navier Stokes does and you obtain differential equations of averages .

    However you have surely not forgotten that you still have to pay the price of the Gi’s .
    Obviously you have only one equation O[T(f),Gi] = 0 and besides T(f) some more new unknown functions Gi .
    So you actually made no progress .
    This is called the closure problem of RANS where the Gi’s are the Reynolds stress tensor components .
    From here on you stop any theory and become empirical .
    Assuming isotropy , symmetry , randomness , homogeneity and some more esoterical hypothesis you will put constraints on the Gi’s . If you find the same number of constraints as there are Gi’s , you will be left with 1 differential equation O and 1 unknown function T(f) (the average) which can be solved .
    You will not really have a theory but you will have a collection of practical cases where the empirical constraints put on the Gi’s are more or less consistent with what is observed .
    So you will know that RANS gives usable answers if the system looks like such and such and wrong answers in all other cases .

    CAVEAT
    RANS uses time averages . As I have already said in the previous post , time averaging is very different from space averaging .
    There are already serious problems with time averaging but if you try writing RANS with space averages you get much worse than only serious problems . You get actually useless numbers that I call “unphysical” .

13. TomV–
    In my turbulence texts, RANS uses “ensemble averaging”, that is, averages over some probability space. As a practical matter, those who actually runs codes based on RANS forget all about the underlying theory, and don’t worry whether it’s time, space or volume!

    From a theory point of view, volume and space averages can have difficulties with boundary conditions. Multiphase flow tends to use volume or time average. However, people discuss space averaged flow rates through planes all the time. I’m not sure why you have difficulties with this. If by “measureable”, you mean literally something that can be measured, it’s easy to measure area averaged flow rates (i.e. spatially averaged velocities.)

14. Lucia
    Yes CO2 absorbs/reradiates energy.
    Yes, the “enhanced greenhouse effect” “exists”.
    The 64 trillion dollar questions are:
    * What is it?
    * How large is it?

    Based on foundational physics, what is the net consequence of increasing CO2 and what is the uncertainty in those estimates and in those trends?

    I was surprised at the number of items I thought I “knew” that
    Gerlich & Tscheuschner examine and find are unfounded based on strictly known physics and formal equations, and that is unquantifiable from first principles.
    E.g., the basis radiation to the 4th power of temperature times a “constant”.

    So what can we actually rely on in terms of foundational physics versus empirical models?
    How sensitive are the models to the assumptions?
    And what models/assumptions are missing or wrong?
    The biggest unknowns seem to be the magnitude – and even the sign – of the water feedback mechanisms. E.g. Roy Spencer found at least one weather feedback mechanism is the negative rather than positive. Satellite data may also be revealing very low climate sensitivity compared to conventional paradigm. Global Warming: Has the Climate Sensitivity Holy Grail Been Found?

    Keep up the good work in examining what is statistically knowable.
    Look forward to your evaluation of the full uncertainty including systematic Type B bias. See: Evaluating uncertainty components: Type B NIST

15. Lucia

    I have mathematically no special problem with any averaging at all .
    I happen even to be aware how one measures flows in pipes and how , in the steady , incompressible case (or with well mixed phases) the average of an integral is the integral of an average because in this particular case only V plays a role and not some general non linear function of V .
    Now obviously this method and conclusion would be wrong if one wanted to measure the flow of a river just below a waterfall .

    My point didn’t deal with particulars of steady state well behaved flows but with space averaging of ANY dynamical parameter in the general case like the climate .
    And the problem with that is already described in detail in the previous post so only a cut/paste :

    When you describe the fluid dynamics in terms of the flow function , the PD equations you get are formally equivalent to hamiltonian equations describing a dynamical system in the phase space .
    The flow function is then equivalent to the hamiltonian .
    Follows that formally the solutions of the hamiltonian equations in the phase space have the same behaviour as the solutions of the flow equations in the ordinary space .
    In other words the spatial variables in fluid dynamics are equivalent to phase space variables in hamiltonian mechanics .

    Now it is a known thing that dynamical systems may be chaotic and present attractors and even fractal attractors in the phase space .
    The equivalence I mentionned predicts therefore that there will be SPATIAL structures in fluid dynamics that willl present exactly the same features (in some/many cases) .
    Surprise , surprise – they indeed do !
    .

    This a quite general result – averaging chaotic orbits in the phase space is useless , says nothing , is “unphysical” even if mathematically possible .
    Actually space averaging a simple closed non fractal curve gives a point – not even the dimensions are physically correct !
    Per equivalence space averaging dynamical variables is useless and “unphysical” as well IF there exist chaotic spatial structures .
    Btw chaos is not really necessary quasi periodical structures will do too .
    Both kinds of structures exist in the real climate .
    Ergo space averaging is forbidden .
    .
    Every single case where space averaging is (semi)legal is a case where no chaotic or pseudo periodical structures exist .
    Typically all examples discussed above belong to this category but these examples are not climate (look at the clouds or isotherms and compare with a flow in a pipe) .

16. Here’s my response to Rasmus posted at RC –

    ‘Why have we failed to convince?’
    I hardly know where to start. The rotating globes are very impressive, (although as pointed out above, they might be more impressive if they rotated the right way) but the same cannot be said of the content.
    The argument is not based on one data set as you claim. If you look at the four commonly used data sets (GISS HADCRU UAH RSS) since 1998 then three of the trends are negative. If you look from 1999 or 2000 then they are positive but if you start from 2001 or 2002 they are all negative.
    Most of those who say ‘global warming stopped in 1998’ do so tongue-in-cheek, perhaps to wind up you guys, but most reasonably objective people looking at for example http://data.giss.nasa.gov/gistemp/graphs/Fig.C.lrg.gif would think that there does appear to be a leveling off.
    And you won’t convince any skeptics by picking on one data set that doesn’t give the result you want, trying to find fault with it, feeding it into to one of your computer models for ‘re-analysis’ and magically creating a hot spot, and then plotting the results in a way to highlight the Arctic and hide most of the S Hemisphere.

17. Tom–
    I think your definition of “unphysical” and mine differ. “Useless” and “unphysical” are entirely different problems.

    I think people need to be very careful about how they use the term “unphysical” as it appears to mean entirely different things to different people.

    On the part in bold: With regard to whether or not spacial averaged values are unphysical, so what? As far as I can tell, all you are saying is the spacially averaged values mean something different from the local values, that we can’t use them interchangeably, and that, if we examine them from the point of view of chaos the average value can’t be expected to behave the same way. But that doesn’t make them unphysical.

    I’m not seeing what difficulty you see in computing the average of the flow through a surface below a waterfall. Computing the flow rate be a pain in the neck, it in might be pointless which means few would be motivated to bother. But there is not conceptual difficulty in computing the volumentric flow through any surface area, dividing and getting a “area averaged velocity”. The unsteadiness of the flow doesn’t make it impossible to do this. The resulting quantity has physical meaning.. (It can also be mis-used and/or used in some sort of approximation to estimate something else.)

18. PaulM–
    I too was puzzled by Rasmus’s claim that whoever it may be that is claiming global warming stopped is basing that on HadCrut. After all, GISS, HadCrut, NOAA, UAH and RSS all show negative trends since 2001. The Argos buoys show a negative trend since inception 4 or so years ago.

    Sure, HadCrut is more negative than GISS and NOAA, but so? And sure it’s possible the bits of the earth that aren’t easily measured my be warming. But if they aren’t measured, who knows?

    At least the “weather noise” argument admits the short term trend is down. It just insists that downtrend doesn’t mean much due to the stochastic nature of weather.

    It would have been nice if Rasmus identified the person whose claim he was disputing. Then we could see if he has addressed anything they claim. As it stands, he may have said a number of correct things I would not dispute, but he doesn’t seemed to have addressed anything anyone claims.

    There you go!

19. Apple pie, well here is mine (and you can make it in less that five minutes:

    cut 8 apples, toss in pan
    cover with fresh ground cinnamon and sugar to taste. Add some orange juice if you like.
    Quickly mix a mans fist worth of butter with 3 parts wheat flour and cover the apples. The butter must not get warm.

    Serve with home made Grande Marniere ice cream (heat 9 dl milk, 1 dl grande marniere, 6 egg yolks and two eggs and 3 dl sugar with a vanilla root, to 84°C and cool while stirring now and then.

    Sometimes this recipe is actually better than Cochran Orcutt 😉

20. Obviously, you bake the pie, at 225°C for 15-25 minutes.

    And if you don’t have an exact thermometer, blow the ice cream over a spoon. When it gets a rose pattern, its 84°C.

21. Avfaktare–
    I think what you describe is called a cobbler or crumble or something like that.

    I have half a bag of apples left, and crumbles are easier than pie. Plus, cobblers are very good! Hmm… maybe I’ll make that crumble tonight.

22. My english doesn’t stretch as far as differensing between a pie and a crumble but thanks for the education…

    You might also want to try the ice cream with Cointreau and lime instead of grande marniere (but not tonight!). Cut the lime in pieces and marinate with the Cointreau for a few days before cooking the ice cream (use only the liquid of the marinade). If you like your ice cream rich, exchange some of the milk for cream.

23. Avfuktare– we have many, many different words for slightly different foods! I’m sure it’s the same in Swedish– but I don’t speak Swedish at all. So. you are still ahead of me.

24. Well, in general I have two words for the entire swedish traditional cuisine: Bland and Blond 😉

    Once you think you learned a language you always stumble upon areas where your command is weak, in my case I’ve learnt this year that food/food preparation and disorders are areas where I come in short of being able to express what I want.

    It’s a bit like mathematics, there is always one area left to study 🙂

    Best!

25. Avfuktare-
    When I lived in Iowa, some friends at a Lutheran church joked that all the rule for food at Lutheran potluck dinners was it has to be white. (There is a large Scandinavian population in Iowa. There are other immigrant groups too, but the Irish are Catholic. 🙂 )

    I think food vocabulary tends to be the last thing one learns. No one bothers spending time explaining the difference between”batter” and “dough” during course work. Plus, lots of foods are popular in one country but not another.

    I spent a year in France. When I came back, I knew a French student who asked me what the word for broccoli was in French. He hadn’t ever eaten it or seen it growing up. I hadn’t see it in the dorms in France. Neither of us had any idea what the word might be. We decided that likely as not it was ‘broccoli’ but we really didn’t know. (Turns out, the French-English dictionaries say “brocoli”, so it’s missing the extra ‘c’.)

26. For the ice cream I would replace all the milk with half-and-half and beat the eggs and egg yolks with the sugar, beating the eggs until frothy and adding the sugar gradually, and continue beating until the mixture lightens and forms a ribbon. Meanwhile scald the half-and-half plus grand marniere and vanilla bean. Remove from heat. Add part of the hot liquid to the egg mixture with stirring to temper. Then add the tempered egg mixture back to the rest of the hot cream and heat with frequent stirring. 84 C seems a little warm. Alton Brown from Good Eats uses a maximum of 80 C, but the custard should thicken and either coat a spoon or form a rose on blowing. Cool. Freeze.

    This sounds like something to make for our New Year’s party to go along with the apple strudel.

27. Lucia
    One challenging question
    Watts has an interesting item
    Statistician debunks Gore’s climate linkage to the collapse of the Mayan civilisation
    Author MacKey finds Gore in effect supports natural oscillations controlling earth’s climate, not anthropogenic causes!

    MACKEY, R., 2007 refers to: Rhodes Fairbridge and the idea that the solar system regulates the Earth’s climate. Journal of Coastal Research, SI 50 (Proceedings of the 9th International Coastal Symposium), 955 – 968. Gold Coast, Australia, ISSN 0749.0208

    He raises the challenging issue of whether climate is ergodic or not.
    This appears to be foundational as to whether climate data can be averaged to reveal anthropogenic causes.

    Definition of Ergodic:“A stochastic process is ergodic if no sample helps meaningfully to predict values that are very far away in time from that sample. Another way to say that is that the time path of the stochastic process is not sensitive to initial conditions.”

    Mackay cites Douglass North (1999) on non-ergodic processes:

    NORTH, D. C., 1999. Dealing with a Non-Ergodic World:
    Institutional Economics, Property Rights, and the Global
    Environment. Duke Environmental Law and Policy Forum
    Vol 10 No. 1 pps 1 to 12. Professor North made this opening
    address at the Fourth Annual Cummings Colloquium on
    Environmental Law, Duke University, Global Markets for
    Global Commons: Will Property Rights Protect the Planet?
    April 30, 1999). The address is also available at
    http://www.law.duke.edu/journals/10DELPFNorth.

    So is climate ergodic or non-ergodic?
    How can we tell the difference?
    Does that impact testing uncertainties in the global surface mean temperature trends being within the IPCC projections?

28. Clearly, I am going to have to get on my Nordic track every day– just to deal with all these recipes!

29. Lucia, food in sweden nowadays are quite exciting, but the traditional recipes display mostly blonde as a colour. There are exceptions, like black pudding, but hey, exceptions confirm the rule… Our strange notions still remain though. We still have raw pickled herring for breakfast and think bread is best when dryer than biscuits 🙂

    DeWitt, I agree on the more elaborated procedure. However, i’m rather confident that 84°C is a very good temperature for making perfect ice cream, as I do like ice cream and is always sceptical about inherited wisdom in the kitchen. So I’ve experimented a lot.

    It is good to be sceptical though, since I learnt (the harder way 😉 ) that any recipe that uses an oven temperature of 175°C is to be misstrusted. I have yet to stumble upon something that is made best at this temperature. Most meats for example, is best cocked at two different temperatures: 140°C until coloured and then 80 – 90°C for as many hours it take to be cooked. Better yet is to fry the meat quickly and then bake it at 80 – 90 °C.

30. Avfuktare krypgrund (Comment#6968) November 29th, 2008 at 2:50 pm ,

    I’ll have to try 84 C for my custard the next time I make ice cream, well maybe 81 or 82 first. Your comment about oven temperature makes me wonder, though, if you’re cooking your custard on the stove top or in an oven. I think the minimum temperature my oven thermostat will accept is 200 F or ~93 C.

    I pretty much agree with your comments about cooking meat in the oven. But I do have a recipe for cooking a beef rib roast where you brown it at the end rather than the beginning.

31. Is it possible to be OT for an open thread.

    I just did a post on Dr. Craig Loehle’s paper discussing the non-linearity of tree ring data in temperature reconstructions, if people are interested.

    http://noconsensus.wordpress.com/2008/12/01/the-800lb-gorrilla-in-the-hockey-sticks-locker-room/

32. Re the Broccoli vegetable name comment: What we know in the USA as “Broccoli” is supposed to be an import from Italy, circa 1910.
    The fellow who introduced it was a greengrocer by that name, who became the father or grandfather of the Mr. Broccoli who produced the first James Bond movies..
    So the French word came from the USA-english.
    Cauliflower, Broccoli, and I believe Kale are supposed to be varients of the save species of cruciferous vegetable.
    Keep up the good work.
    Al S.

33. Al,

    They’re all in the cabbage family as are many other vegetables.

34. Al– That’s interesting! I like broccoli better than kale or cauliflower. 🙂

35. David Hagen

    So is climate ergodic or non-ergodic?
    How can we tell the difference?
    Does that impact testing uncertainties in the global surface mean temperature trends being within the IPCC projections?

    The climate is not ergodic .
    Actually nothing is ergodic as has been shown already 90 years ago .
    If you are familiar with measurable manifolds , I can give you the idea of the proof in 2 lines .
    What the climate (and perhaps other systems) may be is quasi ergodic .
    It is not known whether fluid systems are quasi ergodic .
    It is known that a set of elastic spheres in a perfectly reflecting container is quasi ergodic thus suggesting why it is that thermodynamics might work .
    But that’s about it . Ergodic theory is really not easy .

    You can tell the difference by redoing experiments . You can’t with the climate because you have only 1 experiment .

    The problem with IPCC projection is that it uses GMT what is a spatial average before using temporal averages .
    Spatial averaging is much more problematic than temporal averaging as I show above .
    So before one begins to bother about ergodicity that allows or forbids temporal averaging , one must justify spatial averaging that is generally meaningless .

    Lucia

    On the part in bold: With regard to whether or not spacial averaged values are unphysical, so what? As far as I can tell, all you are saying is the spacially averaged values mean something different from the local values, that we can’t use them interchangeably, and that, if we examine them from the point of view of chaos the average value can’t be expected to behave the same way. But that doesn’t make them unphysical.
    .
    That was not what I was trying to say so I missed some important point to explain somewhere .
    Let’s begin with the flow . A flow is physical . It is a momentum . As a momentum is a vector I can define m.V.dS . No problem .
    I can even calculate Q = Integral [m.V(x,y.z).dS] . If m is constant , I get Q = m Integral [V(x,y.z).dS] what is exactly the definition of spatially averaged V so Q = m .S.Vaverage .
    OK I lost about every interesting information about the flow but Vaverage makes a sense in this case .
    .
    Now under a waterfall we have foam , air and water distributed chaotically .
    So Q(t) = Integral [m(x,y,z,t).V(x,y.z,t).dS] .
    No relationship with Vaverage .
    Vaverage is unknown and unknowable .
    It can’t be measured directly or indirectly .
    There is no relationship connecting it to something physical .
    An infinity of measures would be necessary to compute it empirically .
    Its temporal behaviour is unknown and unpredictable .
    Well a parameter that has the characteristics described above is what one calls unphysical .
    .
    A shorter and more obvious example of this would be averaging an orbit in the phase space what is , I remind again , equivalent to spatially averaging a dynamical parameter .
    It gives 1 point in the phase space .
    This point is not on the orbit .
    So the system will never pass through this point which therefore doesn’t describe an allowable state of the system .
    A point that doesn’t represent a state of the system is called unphysical and on top forbidden , irrelevant , useless etc .

36. P.S

    I forgot to mention that velocity is a very special case that should not be confused with other dynamical parameters and shouldn’t be the focus of a discussion on space averaging .
    As per Noether’s theorem , symmetry of the equations by space translation defines an invariant quantity , in this case the momentum .
    As the conserved momentum is mV and as space translating is a sort of space averaging , one expects that of all dynamical parameters , the velocity would be THE ONE where the space averaging could make sense in particular cases .
    .
    This is absolutely not the case with the temperature that doesn’t correspond to any conserved quantity or equation symetry .

37. Tom Vonk,

    What about heat rather than temperature? Pielke, Sr. has long pushed heat content, particularly of the ocean, as a far better measure of radiative balance or imbalance than a global average temperature.

38. Area= Integral [dS].
    Q= Integral [V(x,y.z,t).dS]

    Vaverage=Q/A. This is the area average velocity.

    I don’t see why you think an infinity of measures is required to compute this empirically. (Do you mean an infinite number of measurements? Sure. But that’s true for many things. It doesn’t make anything unphysical.)

    The average velocity above not what I call “unphysical”. My understanding of “unphysical”, is it requires a violation of some physical law of the universe or it causes something physically impossible to happen.

    While Vaverage may be uninteresting, and we can find no constitutive law relating V to Q, and the exercise of computing Vaverage may be pointless, I don’t see what physical law of the universe is violates.

    I’m not seeing how anything you are describing suggests area averaged temperature is “unphysical” by the definition I generally see used.

    A point that doesn’t represent a state of the system is called unphysical and on top forbidden , irrelevant , useless etc .

    By my definition. a point doesn’t have to represent the state of a system to be called “physical”. You are showing that a particular thing might be irrelevant or useless. But something can be irrelevant or useless without being unphysical.

39. de Witt

    Of course . Imagine a planet half ocean , half continent .
    The continent’s average temperature increases by 2 and oceans decrease by 1 . Average temperature increases by 1 but the internal energy decreases .
    Now the continent’s temperature decreases by 2 and oceans increase by 1 .
    Average temperature decreases by 1 but the internal energy increases .
    So the variation of the average temperature can’t even answer the question if the planet is cooling or warming .
    One could also construct an example of a planet half ice and half water with ice mass variations where average temperatures would also mean nothing .
    Etc .
    .
    Lucia
    Your Q was m^3/s while mine was kg/s . Volume is not conserved while mass is . That’s why I said that an infinity of measures was necessary to compute Q (kg/s) because the density distribution under the waterfall is unknown and unknowable . As Vaverage = Q/S (Q in M^3/s) and this Q is neither conserved nor computable , I can’t say much about Vaverage .
    However as I mentionned in the PS , taking V as support to a general discussion about space averaging is not reperesentative of the problematic because of the Noether’s theorem .
    .
    OK I now see that for you “unphysical” means “violating a law” like a massive particle going FTL f.ex .
    My definition of “unphysical” is larger and related to the dynamical states of a system .
    A state being defined by a point on an orbit in the phase space , there are points in the phase space that will be NEVER visited by a given system (same general proof as the one showing that ergodicity doesn’t exist) .
    Yet the states represented by those forbidden points don’t violate any physical laws because you can conversely prove that there may be OTHER trajectories who pass through this point but just not the one that actually describes your system .
    I call a state that doesn’t describe a possible state of a given system as unphysical .
    So it is a broader defintion than only breaking some established law .
    It is a matter of operationnal efficiency – I prefer not doing physics on states that can’t exist .
    I have nothing against more restrictive definitions – after all there is no physical law forbidding the Earth falling into the Sun .
    It only happens that to the best to our knowledge , the orbit of the Earth in the phase space is chaotic but stable thus making the above state impossible .
    But people can of course spend their time looking for other virtual orbits passing through this state 🙂

40. Tom–
    I changed to Q from Mdot because we had previously been discussing area averaging, not mass weighted averaging. (Though, mass weighted averaging can be seen as a special case where we always weight area by density.)

    That said: Q is computable in a variety of ways. (Each depends on the precise application. But there is always, at least conceptually, a way.)

    One might make errors, but that applies equally to local measurements. The fact of measurement errors doesn’t make something less unphysical. Something doesn’t need to be conserved to be computable. Changing something to an mass weighted average doesn’t make any difference. Sure, one would need to know the local density to compute it correctly. But so?

    Anyway, we are at the point where it’s clear we don’t define “unphysical” the same way. I object to your use which appears to mean “unmathematical”. Your use also makes leaves us no word to distinguish between things that violate a law of physics, and those things which someone thinks violates some principle of mathematics. In papers, conferences etc., I see unphysical used to mean “violates a law of physics”. I recognize there are other ways for something to be impossible– but people should create or select correct words instead of destroying the meaning of a good and necessary word “unphysical”.

41. Lucia

    No the definition of unphysical that is often used in study of dynamical systems does NOT mean unmathematical .
    Something that is unmathematical is pure garbage so we don’t even to consider it if we talk science .
    You seem to think that hamiltonian mechanics is only mathematics .
    That is definitely not so and when one says that a state of a system is forbidden , it is a perfectly physical statement supported by sane mathematics what is generally a good thing too .

    There is no particular reason why unphysical should be restricted to laws only and not include states that cannot be observed .
    If you look at the use of the word “unphysical” by L.Motl and by other string and field theory physicists it appears clearly that they refer also to the states of the system that are possible (=physical) or are not possible (unphysical) while no laws are broken in either case .
    So the word unphysical is used by very many physicists in the same way I use it – that is btw the reason why I use it that way and will continue .

    In the end it is not so important how you define unphysical .
    I find your definition too restrictive but why not if you like it that way .
    What matters is to be aware where are the differences in definitions when people define the same word in slightly different ways .
    I find the question if space averaging temperatures in a chaotic system has any physical meaning more important than these matters of definition of unphysicality .

42. TomV-
    I don’t think hamiltonian mechanics is only mathematics. But, you aren’t just discussing hamlitonian mechanics.

    First lets go back here:

    What I mean is that for some mathematical transformations T of a physical variable X , T(X) is not measurable (as opposed to computable) AND if X obeys some natural law L such as L(X) = 0 then L(T(X))is not well defined .
    In this case I call T(X) “unphysical” even if T is mathematically well defined .
    Space averaging is an example of such T .
    Time averaging not necessarily .

    1) From the point of view of continuum mechanics, spacial average of velocity, density, temperature are all measureable. All these spacial averages are integrals (sort of sums) of measureable quatities defined at point.

    I realize you say that the spacial average of T is not measureable, but I have absolutely no idea why you think this. If the integral ( a sum) of T is not measureable, that would happen only because the T itself is not measurable at individual points. Saying the pointwise measurements aren’t measureable amounts to decreeing that we can’t take the continuum point of view when discussing heat transfer or thermodynamics.

    I guess that’s fine– but decreeing all of continuum mechanics “unphysical” means that “unphysical” methods can still have great value and give extremely useful results in many instances.

    If you mean that, then I would ask you: So, what, precisely, is wrong with something being “unphysical”? Continuum mechanics works quite nicely, and few people call it “unphysical”!

    If you look at the use of the word “unphysical” by L.Motl and by other string and field theory physicists it appears clearly that they refer also to the states of the system that are possible (=physical) or are not possible (unphysical) while no laws are broken in either case .

    Ok… maybe string theorists use this in some idiosyncratic way. But even ifthey do, no one says an averaged value is the “state of a system”. You haven’t shown that the definition of an average in and of itself somehow causes any incorrect specification of any state for any system.

    To predict incorrect states using averages, one must commit the additional error of replacing a correct primitive variable with an incorrect one in a formulation that requires un-averaged properties. It’s the inappropriate replacement of a unaveraged primitive variable with an unsuitable averaged one in a system of equations that require unaveraged variables that results in the unphysical behavior.

    The fact that one can make huge mistakes by mis-using averages does not render the definition of average of anything like T,V, and rho itself unphysical.

    Under any type of mechanics you like, to be sure of correct answers, you first compute using appropriate primitive variables. After you obtain a solution, you can define and compute any sort of average of any property you like. You can use these averages to describe results, characterize things or say whatever you like.

    No physical laws are violated. No unrealistic states are created by virtue of computing an surface average of temperature, velocity or flow after finding the correct solutions using methods that work.

    Averages are not unphysical!

43. Lucia

    My beliefs – because at the moment, all of us simply have our different beliefs, and those claiming otherwise should not be allowed to call themselves scientists – are as follows :
    – Humans emitting greenhouse gases do contribute to global warming. Just like standing by a quiet lake throwing coins into the lake do cause waves to form on the still water surface. (actually this one is not a belief, as there is little controversy on both CO2 being a greenhouse gas and human actities emitting it).
    – So if no other weather and climate effects existed, this warming would be measurable and significant.
    – But other weather and climate effects do exist, they are just not very well understood and modeled yet (and may never be from a “building the climate model bottom up”.
    – So in the same way as standing by the a stormy North Sea in winter throwing the same coins into the water create no really measurable and significant waves (in spite of the coin landing in the water having the same fundamental physical effect on the water with regard to wave generation), I fail to see that our miniscule – in the grand scheme of things – emissions have any significant influence on global climate.

    – Only when repeated analysis of empirical observations compared with model predictions prove that the climate model predictions can be trusted, will I start trusting models with so many built in assumptions as the climate models.

    I am a bit uncertain which of the “camps of people” you listed, this put me in.

44. Lucia 7015 > Averages are not unphysical!

    If demographics show the average couple has 2.2 childen, which body parts does the 3rd part-child possess or lack? And also, does the “average couple” has real referrent or not — and this is not a question about whether we can measure them in various ways, but whether we can pick them out of crowd and put them on prime time TV.

    It’s on this basic level of intelligibility that the science must eventually com down to if social policies are to be agreed. (Of course, one might take the view that democratic agreement for some policies is undesirable, but that is not entailed by the mathematics).

45. Luther– Ok… averages of continuum field properties that are not unphysical. Under the continuum assumption for mechanics, temperature, velocity, density etc. are not integer functions, and they are defined over space and volumes that doesn’t have sudden discontinuities.

    I agree that science must have a basic level of intelligibility if members of the voting public are going to use it to decide what policies they support.

Comments are closed.