Satellites Show Cooling

This is just one of dozens of responses to common climate change denial arguments, which can all be found at How to Talk to a Climate Sceptic.


Satellite readings, which are much more accurate, show that the earth is in fact cooling.

Some very old news here, that’s all. I wonder how long before this one stops being brought up?


There are a few advantages to the satellite readings, mainly the more uniform global coverage and the fact that readings can be taken at different altitudes.  However, it is in fact a very complicated process which uses microwaves emitted by the oxygen in the atmosphere as a proxy for temperature.  The complications arise from many things including decay of the satellite orbits, splicing together and calibrating records from different instruments, trying to separate the signals by the layer of atmosphere they originate from etc.  It is a little ironic that generally the same people who distrust the surface record so much happily embrace this even more convoluted exercise in data processing!

Anyway, it is already many years since the satellite analysis actually showed cooling. Until recently though, one of the several analyses of tropospheric temperatures did show only very little warming and was in direct contradiction to the model predictions which say the troposphere should warm significantly in an enhanced greenhouse environment.  Something had to be wrong, the observations or the model predictions.  Naturally the sceptics had no doubt that it was the models that were off.

However, it turns out that some additional errors were uncovered and the MSU Satellite temperature analysis now shows warming well in line with model expectations. Real Climate has a good run down of the technical details for those with the stomach for it!  In short, this long running debate turned out to be a great validation of the models and a real death blow to the "earth is not warming" crowd.

Beware of zombies!

(image from Global Warming Art)

This is just one of dozens of responses to common climate change denial arguments, which can all be found at How to Talk to a Climate Sceptic.

“Satellites Show Cooling” was first published here, where you can still find the original comment thread. This updated version is also posted on the Grist website, where additional comments can be found, though the author, Coby Beck, does not monitor or respond there.

68 thoughts on “Satellites Show Cooling

  1. An errata of my own:

    In my third and final point regarding Fu et al (2005), I mentioned that upward trend in surface temperatures (+0.17 K/decade) was larger in magnitude than the downward trend in stratospheric temperatures (which Fu reported as between -0.5 and -0.9 K/decade). I had a brain fart there. Mea culpa. Even the smaller (absolute) value of -0.5 is larger than 0.17. So, at least according to Fu, the stratospheric temperature trend is larger in magnitude than the surface temperature trend. However, more recent studies have shown that much of the negative trend in stratospheric temperatures was due to ozone losses, and now that ozone levels are recovering, the trend has actually reversed to positive. It is yet to be determined what the overall trend will be when ozone levels have fully recovered, but it is clear that the overall trend, over 1978 to 2004, is much smaller in magnitude (though still negative) than Fu what Fu claimed, in the neighborhood of -0.2 – -0.25 K/decade. This, of course, is still larger in magnitude than the trend in surface temperatures, so I was still wrong even with the updated analysis.

    Also, I said that the contamination of the T2 temperature reading by the surface temperature is positive, and that removing this contamination would result in a lower T2 trend. This is not necessarily true. IF the GCMs are correct, and the ACTUAL mid-tropospheric temperature trend is larger than the surface temperature trend, then the surface temperature trend negatively affects T2 temperature, and removing that bias would make the mid-tropospheric temperature trend more positive. However, if the actual mid-tropospheric temperature trend is smaller than the surface temperature trend, then my original point is valid; the surface temperature trend is a positive contamination of the mid-tropospheric temperature trend and removing it from the T2 trend would lower that trend. But the point is, we DON’T KNOW what the ACTUAL mid-tropospheric temperature trend is. That’s what Fu was trying to establish. He can’t just ASSUME that the models are correct in attempting to prove that the models are correct. That’s circular reasoning. What he should have done was included a term for the surface temperature in his regression analysis, the same as he did for the T4 stratospheric temperature, and FIND OUT what whether removing the surface temperature trend has a positive or negative effect on the T4 trend.

    (Actually, I suspect that Fu did just that, but discovered that including the surface temperature in the regression proved inconvenient with respect to what he was trying to show, in that either 1) it reduced rather than increased the mid-tropospheric temperature trend, or 2) it reduced the t-statistic for the T4 coefficient to complete insignificance, or 3) it reversed the sign of the T4 coefficient. Of course, Fu wouldn’t report any of these outcomes, but would instead throw out the surface temperature and run the regression again, resulting in what he eventually published.)


  2. (Correction)

    In last sentence of the second-to-last paragraph, delete
    “what” and replace “T4 trend” with “T2 trend”.


  3. “Right or wrong, however, once the radiosonde data has been thrown out, you can’t just go back and use it for something else. If it’s wrong it’s wrong, and you can’t use it anymore. If it’s right, it’s right, and there’s no need to use any other temperature record, let alone some dubiously-constructed amalgam of satellite data from two different channels.” – Trevor.

    Maybe that’s how deniers think science works, but in the real world, not so much.


  4. “Yes, Fu contradicts my conclusions. Because Fu CHEATED!” – Trevor.

    Tsk, tsk, more character assassination of scientists from those unable to accept reality.

    Above you accept that the stratosphere is definitely cooling – a fingerprint of the enhanced Greenhouse Effect, that the surface is warming and yet you seem to have difficulty accepting that the tropical troposphere is warming too?. Weird.

    For any interested readers, here’s a summary of the situation:


  5. The bottom line DW is that the models predict the upper troposphere should show a higher warming rate than the surface. As real world observations show (sat or radiosonde) this is not the case then the only logical conclusion that one can arrive at is the MODELS ARE WRONG.

    However i know that faith plays a large part in how you interpret the world you live in and no matter what empirical evidence there is you simply will not accept it and continue to believe the models are correct.


  6. Ah yes, I was just coming back to concede that “lying” was a little strong, but it’s more pithy than “mislead, on a tangent, and avoid updated knowledge”.

    The link that contains DW’s link to Santer et. al. is

    Obvious things to note, beyond Santer et. al.’s more careful consideration of statistics are
    1. The rate of surface warming in the tropics is lower than the global rate (the flip side of polar amplification). So the tropical mid-troposphere warming isn’t some super-strong signal that should be easy to find just because it’s 1.5 times the surface warming.
    2. As you belatedly realized, because the mid-troposphere warming is a maximum (less warming below, and less warming (and eventually cooling) above) it’s hard to see in the satellite channels, which cover such wide altitude ranges. Santer et. al. clearly claim to have found it in the satellite data though, and I believe this work has stood up … unlike so much of Spencer and Christy’s.
    3. The height and strength of the maximum can vary with weather, so you need a lot of measurements. That variability is what I meant when I said it was a theoretical maximum. There’s no guarantee you’ll find it without a lot of measurements over a climatologically significant time period.
    4. As you can see from the figure on the page I linked to, the sondes give a very wide range of results, but they do show a hotspot.

    So, the big picture is: the *surface* is warming at ~0.17C/decade. There is a mid-tropospheric region of faster warming, as predicted for any overall warming. It’s hard to measure but so far measurements are consistent with theory. There is definitely significant and easy-to-measure stratospheric cooling, the specific signature of greenhouse-gas driven surface warming. So … what’s your objection again? The eastern US had a snowy winter? Arctic ice briefly reached near-normal levels before falling back to the trend? Al Gore is fat?


  7. “However i know that faith plays a large part in how you interpret the world you live in and no matter what empirical evidence there is you simply will not accept it” – Crakar.

    Good to see you have a sense of humor. And seeing as you didn’t read the link I provided here are some relevant bits:

    “This Figure illustrates that both tropical surface and tropospheric temperatures
    have gradually warmed since 1979. Superimposed on this overall warming is
    climate “noise”, which in this case arises primarily from El Niños and La Niñas.”

    “Unlike Douglass et al., however, we found that most of our tests involving
    temperature trends in individual layers of the troposphere did not show
    statistically significant differences between models and observations. This result
    was relatively insensitive to which model or satellite dataset we chose for the
    trend comparison.”

    “The situation was a little more complex for tests involving the trend difference
    between surface and tropospheric warming rates. In this case, the statistical
    significance of the differences between models and observations was sensitive
    to our choice of observational datasets. When we used a satellite-based
    tropospheric temperature dataset developed at Remote Sensing Systems (RSS)
    in Santa Rosa, California, we found that the warming in the tropical troposphere
    was always larger than the warming at the surface.J This behavior is consistent
    with the behavior of the climate models and with our understanding of the
    physical processes that govern tropospheric temperature profiles. It is contrary
    to the findings of Douglass et al.”

    The real funny thing is Crakar, that Trevor nitpicks but doesn’t comprehend the implications of the warming surface temperature and cooling stratosphere, this despite all his flowery prose. Maybe he’s used to a credulous audience?.


  8. Are we reading the same thing DW? I dont think we are.

    The study i have in front of me states that the DCPS07 study shows the upper troposphere (10 to 12K above the tropics) is not warming at a greater rate than the surface. In fact the “hot spot” area had cooled compared to the surface.

    A concerted effort was then launched to prove this wrong and i quote “The general tenor of these
    assessments is that structural uncertainties in satellite- and
    radiosonde-based estimates of tropospheric temperature change are currently large”

    This resulted in CCSP; Karl et al., 2006 to report “advances in identifying and adjusting for inhomogeneities
    in satellite and radiosonde data had helped to resolve the
    discrepancies described above, at least at global scales”

    He continues

    “For month-to-month and year-to-year temperature
    changes, all satellite and radiosonde datasets
    showed amplification behaviour consistent with model
    results and basic theory. For multi-decadal changes, however,
    only two of the then-available satellite datasets (and
    none of the then-available radiosonde datasets) indicated
    warming of the troposphere exceeding that of the surface”

    Karl et al came to one of two conclusions, the first bieng the models are wrong and the second being “residual errors in many
    of the satellite and radiosonde datasets used in the CCSP

    So in essence DCPS07 found there be no evidence of the hot spot predicted by the models.

    Karl et al suggested structual uncertainties was the problem however this still did not find the missing hotspot. Left with little choice they then claimed the hotspot is there but we cant see it due to residual errors.

    Santer et al agree and once again i quote “In view of the large structural uncertainties in
    the observations, the consistency of model amplification
    results across a range of timescales, and independent
    evidence of substantial tropospheric warming (Santer
    et al., 2003, 2007; Paul et al., 2004; Mears et al., 2007;
    Allen and Sherwood, 2008a,b), this was deemed to be
    the more plausible explanation.”

    Santer then prattles on about why his study is better than theirs etc but essentially he along with Karl et al claim the difference in temp/rate is gobbled up in structual uncertanties and erroneous data.

    Did either Karl or Santer measure a temp rate greater than the surface? No they did not. Has the hot spot been found? No it has not. What ramifications does this have? Well its simple the hotspot is made up of all the water vapour which is supposed to build up in the atmosphere due to a +ve feed back from CO2. But yet it cannot be found, just like Trenberths missing heat which now resides at the bottom of the ocean which of course is something else we cannot measure.

    Rest assured DW your faith will see you through these difficult times.


  9. “The study i have in front of me states that the DCPS07 study shows the upper troposphere (10 to 12K above the tropics) is not warming at a greater rate” – Crakar.

    Keep up Crakar!, that’s the study whose flawed methodology was detailed in the link I provided!.

    Now you I can understand getting it all wrong, but those guys?, aren’t they supposed to be scientists?. The warming surface temperatures and cooling stratosphere should have been a dead giveaway, but those guys missed it.

    Crakar the mid tropospheric warming isn’t just a feature of the GCM’s, or of Anthropogenic Global Warming, it’s one of the fundamentals of atmospheric physics, i.e. it’s how we expect the atmosphere to respond to any surface warming. That’s why it’s surprising Douglas, Christy et al, didn’t thoroughly check their work & conclusions before publishing.

    A warming surface temperature will see an increase in water vapor and a corresponding change in the moist adiabatic lapse rate. In a nutshell, the increased water vapor is able to liberate more latent heat as air rises in the atmosphere and condenses at it’s dew point. Ergo both processes should see a rise in the mid troposphere temperatures, above that of the surface.

    “Did either Karl or Santer measure a temp rate greater than the surface?” – Crakar

    Santer et al 2008?, definitely. See link above. And text:

    “When we used a satellite-based tropospheric temperature dataset developed at Remote Sensing Systems (RSS) in Santa Rosa, California, we found that the warming in the tropical troposphere was always larger than the warming at the surface. This behavior is consistent with the behavior of the climate models and with our understanding of the
    physical processes that govern tropospheric temperature profiles. It is contrary
    to the findings of Douglass et al.”


  10. Peter Taylor is essentially saying the Earth is cooling in his book Chill. We need someone to take a long critical look at that book, as there don’t seem to be any critical reviews on the net. I think that Taylor will turn into another Ian Plimer under scrutiny.


  11. There are two types of temperature data. One is measured by thermometer. This thermometer data is flawed because this depends on where thermometer is located at. In last 40 years, at least 25% of thermometer stations have been shut down. Almost all of them used to be in extremely cold weather. There is a very decent youtube presentation for that Moreover, who can forget climategate, where fraud scientists were found to be deliberately manipulating temperature data? What makes you think that they stopped their fraud art now? The second temperature data is from satellite. This satellite data has already been manipulated by NASA. NASA has changed its computer programs to make global warming rook real. Here is the source


  12. So let’s get this straight shall we then Meghal.

    Thermometer temperatures are flawed because the locations are bad.

    Satellite temperatures have been manipulated by NASA.

    So, since we have no usable temperature data, we should not use it in our assessment of climate change. Fair enough? Then how can we say one way or the other whether the climate is changing or not?

    I know – let’s look around at any changes in the ecosystem.

    Oh look! Arctic sea ice melting. Greenland losing ice mass. Sea levels rising. Glaciers disappearing. Flowers blooming early. etc, etc, etc.

    What does that say to you? Or do you want to put your hands over your eyes and your fingers in your ears and go lalala and pretend you can’t see that as well?



  13. Yeah mandas, those pesky plants* and animals**, they must be in on the conspiracy too! I bet NASA signed them up when they promised not to put any dogs in orbit like the Russians did – I hear they signed a treaty with King Louis (the jungle VIP).




  14. There is also the movement of the gardener’s climatic warmer growing zones northwards. I guess that the gardeners of the world are in on the “scam” too.

    Well, it’s either that or the temperatures are changing climatically.


  15. “Why would you expect a regional phenomena to dominate a global signal?”

    Also, what do these people who look at regional historical data make of the fact that regions over the pole have warmed by over 6C? Compared to the maximum extent of the Central Europe readings given which were less than 4C, this warming is higher than the MWP, which, by the way, has been explained and fits within a forcing sensitivity that leads to about 3C per doubling of CO2 or equivalent GHG forcing.

    So if you accept MWP readings, you have to accept about 3C per doubling.


  16. “Am I wrong or does the linear regression of the data start a couple of years after the satellite data?”

    You can’t draw a line between one point. So your first point cannot be used. You can’t draw a line and ascertain a variation from only two points, you need three.


  17. Richard, to your gish gallop, the answers:

    1) False. Many proxies were used.

    2) False. Tree rings from some few trees in the north have been affected by other factors such as acid rain and other pollutants. And, since the IPCC do no research themselves, they would not have advices anything to Mann. Twice wrong. And the proxies work before heavy industrialisation based pollution affected the trees.

    3) False.

    4) False again (see #2, IPCC made no such assertion and #3, so you’re double-dipping here twice), they also didn’t exclude proxies from the MWP. They included the period the peak of the MWP existed for in europe.

    5) Since #4 was wrong, this is again double-dipping and again wrong for the same reasons.

    6) False, they used actual thermometers on the graph and clearly labelled them as such.

    7) False.

    8) False. The data was already available to McIntyre from the owner of the Yamal dataset. The pestering was merely to fabricate a story of hiding the data: the data was already there and available. And the “pink noise” only produced a hockey stick that was

    a) much smaller than the hockey stick from readings
    b) deleted any graphs that didn’t produce a hockey stick

    NOTE: funny how you don’t care about M&M hiding data and throwing away graphs they don’t like.

    9) And a rebuttal from several scientists was upheld: M&M’s concerns were not valid. Worse, their methodology was far more flawed.

    10) What caused astonishment was M&M’s paper was so badly written, so full of errors and yet still published. NOTE: published, not peer reviewed. This was the first time many honest scientists saw the paucity of argument against AGW.

    11) Wegman’s report was proven so full of errors it was nulled. It was false to such an extent that their work proved that a method that NOBODY USED wouldn’t produce reliable results.

    Just as well nobody other than Wegman used it.

    Oh, by the way, Wegman’s reports are being investigated for malfeasance and professional misconduct.

    12) Yes, the report was damning. It was, however false. Mann didn’t use the method Wegman used so therefore proved nothing about the Mann paper. It’s damning Wegman, mind with the investigations being done for fraud, IP infringement and misrepresentation.

    13) They asserted such, but their assertions were shown false. Blind test results with the data on professional statisticians produced the same results as Mann had. The statisticians knew nothing about the dataset except the values (leave out words “temperature” and “date” and you can’t tell what the data is about).

    14) Wegman’s report was written by an intern and used papers produced by a small circle of denialists such as McI. They also didn’t find any such evidence, though this lack of proof did not stop them asserting it. Nor you.

    15) GMU are dragging their feet about investigating Wegman, but he is being investigated for fraudulent misconduct. Again, the method Wegman used was not a method used by anyone else, so his paper could not be considered as reflecting upon Mann’s work. This is about the 9th time you’ve said this, you certainly like to repeat yourself.

    16) And Gerald North is wrong. This is not an unheard of occurrence. Several subsequent reports and investigations have refuted Wegman’s report. His finances also heavily rely upon contributions from the fossil fuel industries and associated lobby groups.

    17) A repeat of #16.

    18) No, they found that M&M had used the wrong analysis and had drawn conclusions impossible to state from the subset they decided to work with.

    19) Yes, the uncertainties of the M&M paper analysis were ignored.

    20) The bristlecone pine data was removed from the analysis by Mann and the results were robust. The shape of the hockey stick does not depend upon the bristlecone pine. Even though that’s what M&M and Wegman said, this is false.

    21) They can express all the lack of confidence they want. This is not any proof of anything other than their belief.

    22) Yet, despite all this “undertainty” there are scads of denialists CERTAIN that the MWP was warmer than today. Plus if the older data is so unreliable, why did you say older data was “proof” of fraudulent reporting in #4?

    23) You assert this but many subsequent analysis (as well as the blind test done mentioned above), with and without bristlecone data and extending even further have proven the hockey stick shape and proven that the current period (since the report in 98) is definitely warmer than the MWP.

    24) The data available from the decade-and-a-half has shown that the current period is definitely warmer. If you have any data to show a MWP in the south feel free to let people have it.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s