JoNova

A science presenter, writer, speaker & former TV host; author of The Skeptic's Handbook (over 200,000 copies distributed & available in 15 languages).


The Skeptics Handbook

Think it has been debunked? See here.

The Skeptics Handbook II

Climate Money Paper


Advertising

micropace


GoldNerds

The nerds have the numbers on precious metals investments on the ASX



Found: the hot spot? Not

The gap between real world data and thermometers is a make-or-break issue for the AGW theory. The models predict a hot-spot in the atmosphere above the tropics, but the weather balloons (called radiosondes) can’t find any sign of it. Most claims that the hot-spot has been found are not providing any new data, they are just massaging the same old numbers with a different statistical tool. Here are three variations (though the third is not a statistical-spin, it’s just nonsense).

1-Some AGW supporters claim that Santer et al has found the hot-spot. But his paper boils downs to a statistical reanalysis that suggests that due to noise and error, the hot-spot might be there. Santer hasn’t actually found the missing hot spot. He has a case, but it’s not a strong one. The statistical counterargument is at Climate Audit.

Even if Santer is right and the hot spot IS hidden in the noise, the most generous interpretation is that greenhouse gases must have a pretty weak warming effect…

2-Sherwood (another co-author of the Santer paper) published research suggesting that the hot spot is there – but he threw away the temperatures from weather balloons and used wind-shear measurements with mathematical analysis.

Because of the importance of this point, investigators have left no statistical stone unturned in their quest to find the hot spot. According to the Sherwood paper those same old radiosonde results going back to 1970 have been subject to “…statistical procedures, station metadata, various indicators of natural variability (such as volcanic eruptions, vertical coherence) and forecasts from a climate data assimilation system.” In their own words:

“Despite these attempts, most analyses of radiosondes continue to show less warming of the tropical troposphere since 1979 than reported at the surface.”
Allan & Sherwood. “Warming Maximum in the Tropical upper troposphere deduced from thermal winds”

Here Sherwood admits that despite an extensive all-out effort, almost every way you look at it with thermometers and statistics they just can’t find the results that the climate models predict. So he resorts to wind-shear. You have to award points for creative effort.

Crikey. The money we wasted putting thermometers on weather balloons…

3-Tim Lambert has a go at asserting that the hot spot has been found. But he confused himself with nice graphs and faulty reasoning. He shows a graph of the fingerprint of warming induced by CO2 next to the fingerprint of warming induced by solar irradiance. Both show a warm spot above the tropics. And what do you know? There’s no evidence that either fingerprint is occurring. (Tim, how is this supposed to prove the hot spot has been found?) Then he points to the cold bar at the top of the CO2 graph, keeps a straight face and suggests that because the weather balloons found some cooling here, that means that the hot spot is not missing. Hmm. So if you are a warmist, half a fingerprint counts as proof, even if it’s the cold half, and you’re trying to prove a warming event.

How are greenhouse gases supposed to heat the planet if they don’t warm some air somewhere? Cooling the upper atmosphere makes for a lousy heat pump.

A fingerprint is a not a fingerprint if half the pattern is wrong. And besides, if Tim looked further he’d note that other causes of global warming also cool the upper atmosphere, like ozone depletion.


The old radiosonde data is what it is. We can’t go back in time and measure where warming occurs until warming occurs again, which hasn’t happened since 2001. So we’re left just with what we have, which plainly does not show a hot spot. Measuring the temperature 10,000m up is not simple—witness all the compensatory factors researchers are working with just to take the air temperature: time of day, season, wind strength, equipment changes, and station relocation. Given how complicated this is, it boggles the brain to imagine the complexities involved in modelling the global climate.

If we can’t be sure we’re measuring the temperature correctly, what chance do we have to measure and manipulate cloud cover, air movements, rainfall, humidity, turbulence, surface reflectivity, AND temperature?

The world is not getting warmer unless you measure ground temperatures in car parks, or tropospheric temperatures with wind gauges. Clutching at straws anyone?

In the absence of better information, based on what we have, the simplest explanation is that greenhouse gases are not warming up the planet significantly.

See all posts tagged “Missing Hot Spot”

VN:F [1.9.22_1171]
Rating: 4.4/10 (8 votes cast)
Found: the hot spot? Not, 4.4 out of 10 based on 8 ratings

Tiny Url for this post:

16 comments to Found: the hot spot? Not

  • #
    Thomas

    “The money we wasted putting thermometers on weather balloons…”

    Weather balloons perform admirably at their intended purpose: doing measurements good enough for weather forecasts a few days into the future. Who cared if you changed instruments or procedures now and then that changed measurements by a degree or so? It didn’t affect forecasts anyway. It does, however, affect temperature trends a lot.

    You use “creative effort” as if it was an insult. In science it is a virtue.

    I’ve seen temperature estimated by measuring microwaves from the Earth and trying to sort out from which part of the atmosphere they derive, from time lag of GPS signals, from the rate atmospheric pressure changes with altitude, from boreholes, from the length of the growing season and numerous other methods. Have you studied wind shear enough that you can confidently say it isn’t a good proxy?

    Your “simplest explanation” is actually rather complex since it requires both that we find some reason why some very plausible calculations of the effects of CO2 are wrong and that there is some other, undiscovered factor that causes about the same amount of warming. Me, I think it is more likely that either there is some measurement error in this hot spot or other factors conspire to mask it.

    I know that Popper says that one measurement is enough to falsify a theory, but that isn’t how scientists work. They believed in aerodynamics even when it seemed as if it proved that bumblebees couldn’t fly, and the fact that several space probes doesn’t quite follow expected orbits doesn’t make them abandon General Relativity. You try to improve your understanding, but in the meantime you work with what you have.


    Report this

    00

  • #
    John Philip

    Huh?

    Based on our current best estimates of these observational uncertainties, there is no fundamental discrepancy between modeled and observed tropical temperature trends. In fact, many of the recently developed observational datasets now show tropical temperature changes that are larger aloft than at the surface – behavior that is entirely consistent with climate model results.

    from the factsheet released alongside the paper.

    http://tinyurl.com/5oys63


    Report this

    00

  • #

    John,
    Thanks for the link. This paper sounds strong, but all they’ve got is another reanalysis of the same old data. The raw numbers are just as far apart as they were, but if you allow for uncertainties then it’s possible to say they (the model predictions vs the real temp) are “not statistically different”. They still haven’t found the hot spot. They’ve found fog in the data. They’re saying the hot spot might be there. That’s not the same as saying it IS there.

    …there is no fundamental discrepancy between modeled and observed tropical temperature trends when one accounts for: 1) the (currently large) uncertainties in observations; 2) the statistical uncertainties in estimating trends from observations. These results refute a recent claim that model and observed tropical temperature trends “disagree to a statistically significant extent”.

    The best case scenario you can get from this for AGW is that the hotspot IS there but it’s too weak to see above the noise. It’s probably not the major climate driver it’s made out to be. Something else is causing the noise; it’s more powerful than carbon, and the climate models don’t know what it is.


    Report this

    00

  • #

    Thomas:

    “You use “creative effort” as if it was an insult. In science it is a virtue.”

    I’m impressed with Sherwood’s creativity. Really. I’m not convinced at his answer tho’

    “Have you studied wind shear enough that you can confidently say it isn’t a good proxy?”

    Thermometers ferrgoodnesssake are designed to measure the temperature. Convince me that a wind gauge that wasn’t, is accidentally better at it…

    “Your ‘simplest explanation’ is actually rather complex since it requires both that we find some reason why some very plausible calculations of the effects of CO2 are wrong and that there is some other, undiscovered factor that causes about the same amount of warming. Me, I think it is more likely that either there is some measurement error in this hot spot or other factors conspire to mask it. “

    Here’s simple: There’s a lab theory about CO2 that can’t be verified in the real world by any actual data. Based on that, I’m unconvinced that CO2 matters.
    I’m not required to come up with the causes that do control the climate in order to maintain my ‘unconvinced’ status about CO2.

    AGW supporters need to do complex mathematical statistical calculations to suggest there is still a possibility that their theory matters in the real world. It goes like this: “If CO2 matters, the evidence is hidden in noise; basic equipment (thermometers) are not accurate enough to find the signal; there are a multitude of other factors that are even harder to measure accurately, including feedback loops we are not sure of; yet we’ve created computer models that can predict things more accurately than our equipment has ever measured”. That’s complicated.

    “I know that Popper says that one measurement is enough to falsify a theory, but that isn’t how scientists work. “

    Sure we don’t stop researching carbon just because one result (or ten) calls the theory into question, but you don’t call the theory ‘proven beyond doubt’, or ‘agreed upon by mainstream scientists’. It isn’t.


    Report this

    00

  • #

    Skeptic’s Handbook: not novel, not right…

    Jeremy Jacquot has written a three part debunking of the claims in Joane Nova’s “Skeptic’s Handbook”: Part 1: increasing CO2 won’t make much difference, Part 2: warming has stopped and ice cores show that CO2 increases do not cause warming,……

    [My reply to Deltoid is here JoNova]


    Report this

    00

  • #

    [...] is not the hot spot but is in fact tropospheric warming and stratospheric cooling. What was her response? Read the rest of this post… | Read the comments on this [...]


    Report this

    00

  • #

    [...] is not the hot spot but is in fact tropospheric warming and stratospheric cooling. What was her response? [...]


    Report this

    00

  • #
    Crust

    You seem to be rather confused about what Lambert said and what the fingerprint metaphor means, etc. He lays it out here.


    Report this

    00

  • #

    JoNova: an excellent thread if I may say so. The Santer et 17 al paper raises some important issues, such as: How should we go about testing observations against model predictions? The Santer paper was published by the International Journal of Climatology, even though the main thrust of the paper has nothing to do with climatology, as it is almost entirely concerned with the methodology for testing model predictions against observations. That means the paper should have been submitted to Econometrica or Journal of the Royal Statistical Society, which have the in-house capability for determining the appropriate methodology for such issues, unlike the IJC. It seems to me that Santer et 17 al. by developing their own criteria for testing models are not even reinventing the wheel but venturing where angels unlike fools fear to tread. Here they follow in the tracks of MBH and their infamous hockey stick with its tree-rings exhibiting R2 of 0.001 invented a new criterion, RE, unknown to statistics textbooks, where a statistic of 0.0 is not a sign of zero correlation but one of perfect fit.

    Thus Santer et 27 al. avoid all mention of R2, and introduce a new stat., R1, which as with MBG conveniently validates their, in reality, spurious correlations. The truth is that there is no way the stats journals I mention would ever publish Santer et 17 al. If they had a case, they would not of course fail to publish their R2 of say .98 for their model forecasts of tropospheric temp. time series from 1978 to 1999 regressed against observations, but they cannot, so they do not.

    The other revealing absurdity of Santer et 17 al is that they castigate Douglass et only 3 al. for not allowing for non-anthropogenic (“natural internal climate variability”) in their tests showing that the Santer et 17 al. models fail to be confirmed by observations . Yet almost all of the Santer et 17 al. were co-authors of the IPCC AR4 WG1 which asserted again and again that there is no significant “internal climate variability”.

    PS I will also try posting this at Deltoid because of his scurrilous attack on you, but we should neither of us hold our breath!


    Report this

    00

  • #
    Chris O'Neill

    Here Sherwood admits that despite an extensive all-out effort, almost every way you look at it with thermometers and statistics they just can’t find the results that the climate models predict.

    Actually the hot spot is fed into the climate models. The models don’t have any choice in the matter. But models get the blame regardless.

    The radiosonde data is just suggesting that the tropical troposphere lapse rate might not always equal the saturated adiabatic lapse rate as is an assumed input to the models. I would expect radiosondes launched these days to carry hygrometers that could easily test the assumption of 100% relative humidity in the tropical troposphere. Thus it should now be very, very easy for scientists to test the saturated adiabatic lapse rate assumption. So if they are making that assumption you can can be pretty sure that they have got the observations to back it up. Scientists don’t make such obvious assumptions without hard evidence. Your whole hypothesis is that this assumption is wrong. Where is your hard evidence? Differences in radiosonde measurements 30 years ago from the much more accurate instruments of today are not hard evidence against the tropical troposphere 100% relative humidity assumption.

    Measuring the temperature 10,000m up is not simple—witness all the compensatory factors researchers are working with just to take the air temperature: time of day, season, wind strength, equipment changes, and station relocation.

    This should tell you something about the accuracy of radiosonde measurements from 30 years ago.


    Report this

    00

  • #
    kuhnkat

    John Thomas:

    from your link:

    “Using state-of-the-art observational datasets and results from a large
    archive of computer model simulations, a consortium of scientists from 12
    different institutions has resolved a long-standing conundrum in climate
    science…”

    Using the models you are trying to validate to reprocess data to validate the models DOES qualify as circular logic. How many other errors in scientific procedure will you allow to PROVE YOUR BELIEF SYSTEM???

    Chris O’Neill states:

    “Actually the hot spot is fed into the climate models. The models don’t have any choice in the matter. But models get the blame regardless.”

    HUH???? It isn’t the models fault and us deniers are blaming it on them????

    Earth to Chris, models are computer programs and exist based on what the programmers wrote. Generically, warmers say the models are right in a vain attempt at an appeal to authority, possibly thinking we will believe computer output and ignoring the fact they can only do what HUMANS have instructed. Deniers, like me, claim the models are wrong when we should be saying the programmers/modellers input the wrong algorerythms!!

    So, we agree on this, BUT, are we also agreeing that the output of the models are WRONG???

    As you stated, where is the radiosonde, or satellite, data to support the 100% relative humidity assumption?? Just saying that they wouldn’t make such an elementary error is pure…

    Here is a paper showing that it varies:

    http://ams.allenpress.com/archive/1520-0442/17/16/pdf/i1520-0442-17-16-3181.pdf

    and YES, BELIEVERS DO MAKE DUMB ASSUMPTIONS TO TRY AND PROVE THEIR POINT!!


    Report this

    00

  • #

    As I expected the incredible Lambert deleted my post once he knew it was there but not before some of his disciples had already commented (at least one deprecated the censorship!). It saeems he has also censored your own comments at Deltoid.

    I think my own point that papers proposing new statistical methodology should first be submitted to stats journals rather than climate journals is worth developing.

    In addition, I think the Real Climate pics that Lambert used to debunk you are bogus:

    1. It is dishonest to use blood red to depict temps at 300 Mb and less (i.e. 10 km altitude and above). When I fly, I note how temps decline monotonically as the plane climbs to as low as -50oC for most of the journey), so how come temps at 10 km up (c.35,000 feet) are depicted by Schmidt/Lambert as hotter than temps at ground level (1,000 Mb), about where most of us live? No doubt they would claim their pics show rates of warming, not absolute temps, but then where is the time variable? Absent that, the Schmidt/Lambert pics are dishonest (exactly to be expected).

    2. Their pics are also absurd in showing no change in temp at Lat. -60 at any Mb pressure, which is contrary to my experience when flying by BOAC (better on a camel) or even Qantas (queue and never taste a snack).


    Report this

    00

  • #

    Tim Lambert said (yet again) @ #301 on Deltoid, “Brian Martens is now permanently banned… Please do not respond to his trolling here, since I will delete your responses as well as his comments.”

    Posted by: Tim Lambert | January 5, 2009 10:52 AM

    My response (at Deltoid, Jan 6th 10.35 pm but about to be deleted by TL): “What right do you have to do that? at 10.52 am on 5th Jan 2009 you were supposed to be working for me (as one of the taxpayers funding UNSW where I believe you work; of course it is very possible that UNSW never requires any work to be done during working hours, if ever, which no doubt explains your stellar career there). If I am wrong, please verify that at 10.52 am on 5th January 2009 you were not using a computer purchased by UNSW, nor any server paid by UNSW, and that at said time you were entitled by your employer, part-funded by me, (a) to spend your time on the total rubbish that you disseminate on your site and (b) to prevent Australian taxpayers like me from posting at your site.


    Report this

    00

  • #
    Rod Smith

    Just a fast comment from and old (very old!) radiosonde/dropsonde guy.

    Radiosondes do not carry “wind gauges,” but wind direction and wind speed are calculated based on the drift of the balloon.

    I assume that late recent instrumentation may use GPS for better accuracy.


    Report this

    00

  • #
    ben

    I understand your criticisms of Sherwood’s analysis. But looking at the paper, it does show the hotspot very well, are you suggesting it is just a coincidence or is there a reason for this?


    Report this

    00

  • #

    [...] climate models predict. So he resorts to wind-shear. You have to award points for creative effort. http://joannenova.com.au/2008/10/not…-the-hot-spot/ It's not our problem that your cult hasn't informed you about the state of the science. And [...]


    Report this

    00