Satellite temperatures

There have been a number of interesting recent posts elsewhere about satellite temperature. Most of this has been motivated by claims that the satellite data is the best data that we have and that it should play some kind of pivotal role on our assessment of anthropogenic global warming (AGW). Even if the satellite data was excellent, this would be an odd thing to suggest given that it’s only measuring one part of the system; the troposphere. We really should be considering all the evidence, not simply one dataset associated with a small part of the climate system.

However, even those who work with satellite data regard it as less reliable that the surface temperature datasets. As Carl Mears says:

A similar, but stronger case can be made using surface temperature datasets, which I consider to be more reliable than satellite datasets (they certainly agree with each other better than the various satellite datasets do!).

This doesn’t mean that it should be dismissed, but does suggest that claims that it’s the best data we have are unjusitified.

Kevin Cowtan also has a very nice article that compares satellite datasets and surface temperature datasets. It essentially shows that the uncertainty in the satellite data is quite a bit greater than that in the surface data, and concludes that

on the basis of the best understanding of the record providers themselves, the surface temperature record appears to be the better source of trend information. The satellite record is valuable for its uniform geographical coverage and ability to measure different levels in the atmosphere, but it is not our best source of data concerning temperature change at the surface.

The latter seems pretty obvious, given that the satellite measurement are used to estimate temperatures in the atmosphere, not the surface.

Peter Sinclair has also produced a very nice video about the satellite data. I’ve included it below, but it discusses how actually determining temperatures from the measurements requires a model, points out that there have been a number of well-documented errors that have had to be corrected (typically changing the trend), and that really we should consider all the data when trying to assess AGW. It also points out, quite rightly, that it’s odd that those who regularly criticise adjustments to surface datasets, ignore not only that satellite data requires a great deal of model-based analysis, but has a history of corrections that make substantial changes to the trend.

Some people are not very happy with the video. John Christy is accusing those involved of smearing him. As far as I can tell, everything in the video is true, even if it does reflect poorly on some. On the other hand, he appears to have said this to James Delingpole, someone who has made a career out of smearing climate scientists. It’s hard to take John Christy’s concerns all the seriously, given that it’s being reported by someone with James Delingpole’s history.

At the end of the day, we should consider all the evidence, satellite too. However, we should also be aware of the complexities involved in analysing these datasets. No single dataset will be perfect, and no single dataset will be completely useless.

Advertisements
This entry was posted in Climate change, ClimateBall, Global warming, Judith Curry, Roy Spencer, Science and tagged , , , , , , . Bookmark the permalink.

127 Responses to Satellite temperatures

  1. lerpo says:

    Tamino shows RSS minus RATPAC here: https://tamino.files.wordpress.com/2015/09/rssminusrat.jpeg?w=500&h=332 . It looks like the satellite models track the RATPAC measurements fairly well up until about the year 2000. Then something happens to cause a divergence.

  2. Michael 2 says:

    Hello all. Testing what seems to be a significant malfunction in WordPress. Most of my comments just vanish; they used to redisplay saying “in moderation” so I knew the server had them. Now they just vanish. So I post this throw-away comment to see if it vanishes, goes into moderation or whatever.

  3. Michael 2 says:

    Well, that test vanished. Interesting. It doesn’t even send me a subscription confirmation email.

  4. M2,
    They were simply in moderation.

  5. “I’ve included it below, but it discusses how actually determining temperatures from the measurements requires a model”

    Shit.. I’ve been telling skeptics that for years. they dont like to hear it.

  6. Steven,
    Yes, well some people just don’t want to listen 🙂 Technically data analysis almost always needs something that one could reasonably describe as a model. Some of them are trivial, others more complicated.

  7. @DumbSci,
    I missed that exchange. I saw some between Gavin and Roger, but I hadn’t seen yours. Funny that someone who would agree to be part of the GWPF temperature review would accuse others of being politically motivated.

  8. Willard says:

    Very Serious Science from Senior over the Tweeter:

    But is Gavin honest?

  9. Still haven’t figured out what caused all the miscommunications in that exchange. It seems like he kept asking me to answer a question after I’d already answered it, then called me a troll after I asked him to answer his own question. Surely that can’t be right?

    I agree with ATTP about the video. As far as I can tell, everything in the video is true. In fact, those scientists basically seem to just be describing the same papers I’ve already described:

    Gary and Keihm 1991 showed that natural variability in only 10 years of UAH data was so large that the UAH temperature trend was statistically indistinguishable from that predicted by climate models.

    Hurrell and Trenberth 1997 found that UAH merged different satellite records incorrectly, which resulted in a spurious cooling trend.

    Wentz and Schabel 1998 found that UAH didn’t account for orbital decay of the satellites, which resulted in a spurious cooling trend.

    Fu et al. 2004 found that stratospheric cooling had contaminated the UAH analysis, which resulted in a spurious cooling trend.

    Mears and Wentz 2005 found that UAH didn’t account for drifts in the time of measurement each day, which resulted in a spurious cooling trend.

    It would be surprising if a well-published climate scientist hadn’t heard of those papers. So this ultimatum (presumably about that video) is also “funny”:

  10. Steven Mosher says:

    On the video.

    They missed an opportunity. Should have attacked the data and left Christy alone.

    Innuendo needs to be more subtle and left to the reader to go a bridge too far

  11. Willard says:

    21 hours ago:

  12. They presumed Dr. Christy was honest. Assuming he made those mistakes honestly might be more constructive than other potential assumptions.

    Pointing out that the UAH record has been adjusted more than surface records isn’t attacking data. Nor is pointing out that even UAH doesn’t show a statistically significant change in the warming rate. Nor is pointing out that the UAH record only represents ~1% of trapped heat. Nor is pointing out that NASA also has satellites which measure the ~90% of heat going into ocean heat content, if anyone wants to look at the big picture rather than cherry-picking the ~1%. This isn’t about attacking the data, it’s about defending the scientific conclusions of NASA, the National Academy of Sciences, the Royal Society, etc. from the endlessly repetitive misrepresentations of that cherry-picked data being spread by WUWT, Ted Cruz, Dana Rohrabacher, Lamar Smith, etc.

  13. curiousaboutclimate says:

    Nowhere in that video is Christy attacked or smeared. That is simply a fact. Shortcomings of the satellite data are pointed out, and not for the purpose of casting doubt on what they show, but for the purpose of explaining that they should not be unequivocally considered the “best data we have”. The video is quite clear on that.

  14. BBD says:

    Yes but ClimateBall.

  15. Paul Williams says:

    If the surface record is so good …

    … then why are they fixing it every few weeks. As in completely changing the historical record.

    It seems that the weather station operators of the past were always reading temperatures too high. And then every few weeks, our estimation of how much too high is just increasing.

    That is why there is a discrepancy now between the satellites and the surface records.

  16. curiousaboutclimate says:

    Every few weeks? What is this guy talking about?

  17. curious,
    I don’t know. Maybe Paul can clarify.

    Nowhere in that video is Christy attacked or smeared. That is simply a fact.

    Indeed. In fact, Carl Mears – Senior Research Scientist at RSS – is in the video. For some reason he was happy to be included in a video about the reliability of Satellite Temperatures, while Christy seems to think he’s been smeared.

  18. BBD says:

    Paul Williams

    I was under the impression that BEST put to rest the concerns about the existing gridded surface temperature products. It was a sceptical investigation but it confirmed GISTEM, HadCRUT etc. – it didn’t show them to be exaggerating the centennial trend.

  19. Harry Twinotter says:

    It’s a diversion of course. An attempt to change the subject from satellite data to who was and wasn’t “smeared”.

  20. Phil says:

    Pielke Sr’s cartoon (courtesy of Willard) actually inadvertently catches the essence of the issue:

    “It’s a satellite, it’s state of the art, it was expensive, very expensive. It has to be right!”

    “Err, but it doesn’t, you know, actually measure temperatures. It wasn’t designed to. You have to infer them with a model and correct the data …”

    “But it’s a satellite, it’s “state of the art”, it was expensive, very expensive. It has to be right!”

    In denier world “satellite” is a magic word.

  21. Seems that Radford Neal has something to say about the word of the year:

    Many people pointing to a pause look at the satellite temperature data from UAH, which starts in 1979.

    https://radfordneal.wordpress.com/2015/12/19/has-there-been-a-pause-in-global-warming/

    I rather like his conclusion, with caveat emptor and all (ibid.):

    I emphasize again that the exercise of looking for a `pause’ or `slowdown’ is really a crude way of looking for evidence of long-lag autocorrelation. The quantitative results should not be taken too seriously. Nevertheless, the conclusion I reach is that this data does not produce a definitive yes or no answer to whether there is a pause, even in the UAH data, for which a pause seems most evident.

    I like that conclusion because I can now speak of da-paws-that-may-or-may-not-exist and because everyone involved in this ClimateBall ™ episode seem to lose.

    PS: H/T opluso at Judy’s for the link. Reading Denizens helps. Go Team!

  22. Eli Rabett says:

    Dumb Scientist: Dealing with the Pielke’s is an art. One starts from the observation that they behave like entitled three year olds playing why daddy, and then shift ground rapidly. The conclusion is that you have to call them out for acting that way.

  23. Eli Rabett says:

    Eli and Tamino are on to something.

    It looks like either a) there is some aging effect in the AMSU receivers or b) the atmospheric/surface composition has shifted in the last 15 years or so (e.g. less ice/ more water vapor) in a way that biases the returns.

  24. semyorka says:

    If I understand correctly about 5000m the air pressure will be about 55kpa, this means there is physically less mass of air up there thus it is it takes less energy to warm (or cool) it. More over water vapour tend to become water again at these altitudes releasing energy, so where there is a lot of moist warm air getting pushed up from the surface near the tropics you will see a big surge in temperature while near the poles, even when warming you will see a lot less heat being pushed into the mid troposphere in the form of latent energy in water vapour.

    To me this looks like you will not and should not expect to see the same trends as at the surface. The mid troposphere will show exaggerated effects of low latitude ocean cycles and will tend to mask warming at very high latitudes.

    Its simply measuring some thing different. Thus 10 years of stronger la Ninas, cool water pulled up in the equatorial Pacific which should show an exaggerated lack of warming trend compared with the surface.

  25. BBD says:

    For the mix:

    Weng et al. (2014) Uncertainty of AMSU-A derived temperature trends in relationship with clouds and precipitation over ocean

    Microwave Sounding Unit (MSU) and Advanced Microwave Sounding Unit-A (AMSU-A) observations from a series of National Oceanic and Atmospheric Administration satellites have been extensively utilized for estimating the atmospheric temperature trend. For a given atmospheric temperature condition, the emission and scattering of clouds and precipitation modulate MSU and AMSU-A brightness temperatures. In this study, the effects of the radiation from clouds and precipitation on AMSU-A derived atmospheric temperature trend are assessed using the information from AMSU-A window channels. It is shown that the global mean temperature in the low and middle troposphere has a larger warming rate (about 20–30 % higher) when the cloud-affected radiances are removed from AMSU-A data. It is also shown that the inclusion of cloud-affected radiances in the trend analysis can significantly offset the stratospheric cooling represented by AMSU-A channel 9 over the middle and high latitudes of Northern Hemisphere.

  26. Brandon Gates says:

  27. opluso says:

    Leaving the clashing personalities and egos aside, is concern over satellite-derived temperature estimates more, or less, appropriate than concern over satellite-derived estimates of sea level rise?

  28. opluso,
    Your question would seem to indicate that you’re not really getting the point. It’s not a concern over satellite-derived temperatures, it’s a suggestion that there is little evidence to support the claim that the satellite data is the best we have and, even if that were the case, we should really be considering all the data and not simply focusing on one dataset. The satellites estimate temperatures in the troposphere. Satellites that measure sea level rise can be compared with other estimates for sea level rise.

  29. guthrie says:

    We can be concerned about more than one thing at a time. However what little I know of ocean height measurement leads me to believe that it is that it is far simpler, comparatively, than the temperature measurements. Moreover, nobody has been making loud noises about differences between tidal gauges and satellite observations, or rather, no seriou case has been made that there is a divergence and concommitant problem.

  30. > [I]s concern over satellite-derived temperature estimates more, or less, appropriate than concern over satellite-derived estimates of sea level rise?

    Leaving clashes of one-upmanship aside, we should be thankful for all the concerns.

  31. John Hartz says:

    More grist for the mill…

    A Response to the “Data or Dogma?” hearing by Benjamin D. Santer and Carl Mears, Skeptical Science, Jan 17, 2016

  32. snarkrates says:

    Opluso,
    Dude, at least the satellites are actually measuring sea level, whereas the satellites are NOT measuring temperature. Get it?

  33. To add one further point to ATTP’s response to opluso, the satellites measure 11 different zones in the atmosphere -you can see the corresponding temperature models from these channels here .

    Since their inception in 1979, all of these channels are consilient with AGW theory, and even after 1998, all but one of the channels are still consilient with AGW. So the issue isn’t just a matter of choosing satellites over other forms of temperature estimation, it is also the focus on less than 10% of even the satellite data.

    But there is a more general question here to do with how scientific reasoning should work. There has to be a hierarchy of reasoning, otherwise science is incoherent. What I mean is that if one kind of dataset is not consilient with the rest of the data and the theory, then it is the new data which should be questioned first, unless it is part of a better theory, which accounts for both the same evidence as the existing theory, and also the anomalous evidence which the existing theory cannot account for. Theories bring together multiple lines of evidence into a whole which is greater than the sum of the parts -that is why data without a strong theory should always be lower on the hierarchy of reasoning than data which fits into a strong theory.

    It really is an elementary part of the scientific method that if your measurements are not in accord with the prevailing theory, the first thing you should do is re-check your data, before you go claiming to have falsified the theory.

    As it stands at the moment, the enhanced response of the satellite models to the ENSO cycles, plus the many complexities associated with satellite data, plus the absence of a new theory which can incorporate the satellite data into a better theory than AGW, all mean the most reasonable response right now is skepticism of the satellite models.

  34. Eli Rabett says:

    Well actually there are a whole bunch of tide gauge fans out there. Scratch a Christy groupie and you will find one.

  35. opluso says:

    Personally, I think the ground-based temperature sets are pretty good, particularly considering the difficulty of the task. Satellites are another tool but I don’t consider them (or the interpretations of their data) infallible.

    I was curious whether folks felt that all microwave readings from satellites, filtered through complex models built upon a series of assumptions but “ground truthed” with comparisons to alternative measurement technologies, were potentially problematic. Clearly, the answer is “not always.”

  36. I agree opluso. It may be that there are systematic problems with the satellite data, or it may be the data is good, and there are processes taking place in the troposphere which have not been accounted for in the theory to date (there is serious research being done to investigate both questions). Sorting out either one would advance the science, in the sense that this anomalous data would be resolved into a better overall explanation.

    It’s also possible that the satellite data points towards a fatal flaw in the AGW consensus, but in the absence of a better theoretical ‘home’ to explain this data, it makes sense to treat it as a sign that ‘the tail of this dog needs some attention’, rather than ‘we need a new dog for this tail’, as Christy et al claim.

  37. JCH says:

    Oplusa is very sharp, and he asks rigorous questions of all sides.

    Discuss lines of evidence that ten to affirm there has been no warming of the surface in the last 18 years.

  38. Brandon Gates says:

    John Hartz,

    Thank you for posting the Data or Dogma hearing follow-up by Drs. Mears and Santer. I offer a critique of following paragraph:

    “Model input errors” and “different variability sequences” require a little further explanation. Let’s assume that some higher extraterrestrial intelligence provided humanity with two valuable gifts: a perfect climate model, which captured all of the important physics in the real-world climate system, and a perfect observing system, which reliably measured atmospheric temperature changes over the last 18 years. Even with such benign alien intervention, temperature trends in the perfect model and perfect observations would diverge if there were errors in the inputs to the model simulations,[h] or if the purely random sequences of internal climate oscillations did not “line up” in the simulations and in reality (23, 24, 30, 32-36).

    It is my understanding, going all the way back to Lorenz (1963), Deterministic Nonperiodic Flow [1] that the more appropriate way to think of the weather/climate system is that it would NOT diverge from previous behaviour if all the initial inputs were exactly the same. “Purely random” evokes the concept of a stochastic system where there is no such guarantee by definition.

    I understand that what Drs. Mears and Santer mean by “purely random” is that with the the real system, which is massive and complex, we do not have the observational fidelity OR computational ability to reliably predict short-term climate trends (i.e., weather) in advance due to the sensitivity a deterministc system has for initial conditions — therefore, it behaves as an “effectively random” system for the purposes of exactly timed, very precise prediction of future states.

    However, because it is an almost completely deterministic system, we can at least theoretically hope after the fact to suss out a causality chain for the various modes of internal variability and/or pertubations in external forcings which do contribute to constant change even absent our influences. I think this is a distinction which separates our argument from the magical thinking of the “climate is always changing (and nobody knows why)” crowd …

    … not that many of them see it that way since Lorenz (1963) is often abused as “proof” that climate cannot possibly be projected or predicted over the long-term because … chaos.

    —————

    [1] http://eaps4.mit.edu/research/Lorenz/Deterministic_63.pdf

  39. John Hartz says:

    Brandon Gates,

    If you have not alrady done so, I encouarge you to also post your comment on the SkS website.

  40. John Hartz says:

    ATTP: Once again you and Phil Plait seem to be channeling each other…

    No, Sen. Ted Cruz, Satellites Are NOT Our Best Way to Measure Global Warming by Phil Plait, Bad Astronomy, Slate, Jan 14, 2016

    I do, however, prefer Plait’s headline to yours — it has more bite and stickiness.

  41. Brandon Gates says:

    John Hartz, I have cross-posted my comment to SkS, thanks.

  42. “It looks like either a) there is some aging effect in the AMSU receivers or b) the atmospheric/surface composition has shifted in the last 15 years or so (e.g. less ice/ more water vapor) in a way that biases the returns.

    voting for the second one

  43. “It’s a diversion of course. An attempt to change the subject from satellite data to who was and wasn’t “smeared”.

    ya… they needed to play the innuendo notes much softer. no need to dog whistle so loudly.

    Mears and Dessler were all that were needed… connecting the dots is best left to the readers

  44. Steven,
    Do you really think it would have made any difference? Partly, I do think that some of the history is relevant but I also think that it’s unlikely that a slightly different framing would have made any difference to how the video has been received. It’s not really to convince the unconvinceable, I think it’s more to illustrate that the claims that the satellite data is the best we have are particularly silly.

  45. Brandon Gates says:

    Anders,

    I think it’s more to illustrate that the claims that the satellite data is the best we have are particularly silly.

    Address all complaints to Willard, this is all his fault:

    Yes, sadly, yes I am.

  46. jsam says:

    On teh Goddard and Climate Despot you will read that NASA admitted satellites were better in 1990. Expect to see the meme repeated.

    The claim is here.
    https://stevengoddard.wordpress.com/2014/01/07/nasa-1990-no-global-warming-surface-temperature-record-should-be-replaced-by-more-accurate-satellites/

    The acolytes say that the interweb has lost the source. I say not so.
    http://www.nasa.gov/home/hqnews/1990/90-045.txt

  47. jsam says:

    I’m rubbish at tags. 😦

  48. This is all I can find about the NASA report of no global warming. No prizes for guessing who was involved.

  49. BBD says:

    But the UAH product was borked. How can anyone turn the lamentable history of errors and revisions that hangs over UAH TLT into a claim that it is the best temperature data we have?

    ‘Tis nuts, i’faith.

  50. BBD says:

    DumbSci provides an excellent summary upthread:

    Gary and Keihm 1991 showed that natural variability in only 10 years of UAH data was so large that the UAH temperature trend was statistically indistinguishable from that predicted by climate models.

    Hurrell and Trenberth 1997 found that UAH merged different satellite records incorrectly, which resulted in a spurious cooling trend.

    Wentz and Schabel 1998 found that UAH didn’t account for orbital decay of the satellites, which resulted in a spurious cooling trend.

    Fu et al. 2004 found that stratospheric cooling had contaminated the UAH analysis, which resulted in a spurious cooling trend.

    Mears and Wentz 2005 found that UAH didn’t account for drifts in the time of measurement each day, which resulted in a spurious cooling trend.

  51. I see you tracked down the origins of the 1990 NASA report.

  52. jsam says:

    Indeed, and I also have the link upthread in this stream of posts.

  53. Steven Mosher says:

    Attp.

    Ya it would have made a difference.

    @bbd the skeptics generally prefer rss today.

    Rhetorically the best approach is too point out all the issues with rss…. And leave uha out of the discussion until they drag uha in…

    Then hit them with the uha issues.

  54. BBD says:

    @bbd the skeptics generally prefer rss today.

    Despite the fact that Carl Mears explicitly said that the surface temp data were more reliable…

    Nuts.

  55. BBD says:

    BTW Steven, are you ever going to let us know what’s wrong with Wilson16? You rather left that hanging on the other thread.

  56. BBD says:

    Ya it would have made a difference.

    Absolute certainty, zero evidence.

  57. Joshua says:

    ==> “Absolute certainty, zero evidence.”

    Shocker.

  58. verytallguy says:

    Rhetorically the best approach is too point out all the issues with rss…. And leave uha out of the discussion until they drag uha in…

    Then hit them with the uha issues.

    Climateball part match analysis. Always so fascinating.

    I’m intrigued though. “Best” in what sense, Steven?

  59. Brandon Gates says:

    So the (A)MSU estimates contain 5 times the uncertainty of HADCRUT4. No wonder Judith Curry thinks the satellite data are the best we have.

  60. Willard says:

    > “Best” in what sense […]

    When your opponent puts a card on the table, he can’t put it back.

    If you put it on the table before he committed to it, he might attack it instead of having to defend it.

    ClimateBall ™ takes discipline because managing commitments is hard.

    ***

    On the other tail, dogwhistling has strong limits:

    http://neverendingaudit.tumblr.com/tagged/dogwhistling

  61. Thank You willard.

    VTG… think of it as chess.

    you let them put RSS en prise ( monkton is a good choice)
    you have mears explain
    A) RSS uses a GCM to adjust the dirunal cycle
    B) They use modelled temps (NCEP) as a basis for their forward algorithm

    At this point the skeptic is arguing that the best data is adjusted by a GCM and bases its temperature on NCEP, a model.

    And then you talk about the uncertainties… again mears excellent work

    And you hope that the skeptic then puts UHA in play

    And then you take that down..

    in the end… a smart skeptic will say that all data is shit and we should start with a brand new observation system

  62. Steven,
    Except, that assumes that those who posted that video actually want to play that kind of game. They can also simply let the video stand on its own.

  63. verytallguy says:

    Steven,

    You seem to be arguing that you have a good strategy for convincing a smart and open minded sceptic of the facts.

    Have you spotted the problem yet?

  64. Windchaser says:

    Have you spotted the problem yet?

    Well, yeah. They go from “the satellites show no warming, so the surface is wrong” to, when that is disproved, “all data is shit and we should start with a brand new observation system”.

    At no point did they recognize that they only liked the satellite because it agreed with their idea that AGW is wrong; that they were being half as skeptical of the satellite data as the ground data, and why. They were patently biased, but instead of saying “hmm, maybe I’m not really looking at this issue fairly”, they fall back as little as possible.

    And in another month or two, they’ll have forgotten about this entirely, and they’ll follow whatever the new anti-AGW fad is. “Surface temperature data shows cooling since 2015”, “a new ice age is coming”, “look at this petition of engineers that says AGW is a fraud”, etc. Even if the new fad contradicts an old one. It’s all fair game.

    You can pretty easily tell which people are genuinely interested in an accurate understanding of the natural world, and which ones are ideologically motivated. The ideological ones are comfortable presenting cherry-picked data or multiple contradictory ideas to make their case. It’s about throwing stuff at the wall and seeing what sticks. And if it doesn’t hold up: oh, well, we’ll keep throwing.

  65. Windchaser says:

    So the (A)MSU estimates contain 5 times the uncertainty of HADCRUT4. No wonder Judith Curry thinks the satellite data are the best we have.

    Well, it fits well with this approach:

    “Look at all the uncertainty!
    … Okay, now ignore the uncertainty and focus on the central estimate, which shows we can ignore global warming.”

    It’s a nice piece of handwavery.

  66. Brandon Gates says:

    Windchaser,

    It’s a nice piece of handwavery.

    Heat must be getting to them. Spotted over at WUWT this morning:

    ——————

    Monday mirth – Old Reliable
    cartoonsbyjosh / 11 hours ago January 18, 2016

    Josh writes: As visitors here know there’s a video titled “How reliable are satellite temperatures?” with the usual suspects giving their expert opinions. You can watch the video here and read the post on WUWT here.

    ——————

    My response, shamelessly borrowed from Willard:

    In more serious news, I posed a question for Kevin Cowtan over at SkS this morning …

    http://skepticalscience.com/surface_temperature_or_satellite_brightness.html#115622

    I’m reading through Mears et al. (2011), and find Table 2 reports 2-sigma trend uncertainty estimates for TLT as 0.044 K/decade, whereas Figure 4 of this note puts the 95% CI for RSS TLT at about 0.13 K/decade. Explanation for that nearly 3x difference is not readily apparent to me from reading the paper as both calculations appear to be estimating the same thing. Could you please explain my error in understanding and/or give a justification for the difference in estimates.

    … wondering if anyone here has some interest/input?

  67. Eli

    “Eli and Tamino are on to something.

    It looks like either a) there is some aging effect in the AMSU receivers or b) the atmospheric/surface composition has shifted in the last 15 years or so (e.g. less ice/ more water vapor) in a way that biases the returns.”

    also the weighting function relies on an assumption of a constant emissivity for earth.
    gimme a few days and I may be able to tell you if gridded delta’s between RSS and BE are
    correlated with changes in emissivity.

  68. Eli Rabett says:

    Cool, that would tie a lot of things together.

  69. -1=e^iπ says:

    I’m surprised the video didn’t mention Cowtan & Way, which uses satellite data to help deal with coverage bias in HadCRUT4 and makes the warming trend significantly different from zero at the 95% confidence level during 1998-2014.

  70. “gimme a few days and I may be able to tell you if gridded delta’s between RSS and BE are
    correlated with changes in emissivity.”

    Isn’t this what Weng et al, Uncertainty of AMSU-A derived temperature trends in relationship with clouds and precipitation over ocean (2014),cited upthread by BBD have already shown?

  71. Willard says:

    > Cool, that would tie a lot of things together.

    If it comes with a rug, I hope it’s cleaner than when it left the room.

  72. Joseph says:

    It looks like the satellite models track the RATPAC measurements fairly well up until about the year 2000.

    I found this to be interesting comparison between the Karl and the RSS datasets here:

    https://www.skepticalscience.com/trend.php

    So from 1979 to 2000 RSS has a trend of .145 C per decade and Karl has .140 per decade. But from 2000 to now RSS has a trend of.025 per decade while Karl has .141 per decade. So something does seem to have happened to cause these datasets to diverge. I was wondering if the fact that the surface datasets have presumably improved in the last fifteen years due to better methods and coverage, suggests that the problem more likely lies with RSS?

  73. -1,
    Has anyone being going around claiming that “Cowtan & Way is the best data we have”?

  74. dikranmarsupial says:

    -1 Satellite data is better than no data. The use of satellite data in C&W does not imply that it is more reliable or better some how than station data. The point is that scientists try to make the best inferences about climate that they can, based on the observations that they actually have, bearing in mind the problems and deficiencies of each source.

  75. MartinM says:

    The whole point of Cowtan and Way is that it combines the satellite and surface data in ways that use the strengths of one to compensate for the weaknesses of the other. C&W exists precisely because of the weaknesses described in the video.

  76. BBD says:

    Thank you for paying attention, oneillsinwisconsin 😉

  77. Martin,
    Indeed.

    Brandon,
    The problem with Josh’s cartoons is that there has to be some kind of truth to them for them to be funnny. Since there rarely is, they typically aren’t. I don’t anyone is claiming that tree rings are the best data we have. In fact, Michael Mann recently tweeted (I would link to it, but Twitter appears to be down worldwide) about PAGES2K being one of the best reconstructions, so I don’t even think that he regards tree rings as the best data for millenial reconstructions, let along millenial reconstructions themselves being the best data. I guess, though, if all you have left is making stuff up about what other people have said, then that’s what you’ve got to do. Says it all really.

  78. -1=e^iπ says:

    @ ATTP – “Has anyone being going around claiming that “Cowtan & Way is the best data we have”?”

    I don’t know.

    @ dikranmarsupial

    I don’t think I said anything that disagrees with your statement.

  79. Brandon Gates says:

    Anders,

    This time Josh’s cartoon came with disclaimers:

    NOTE/update: To be clear, this most influential tree, YAD06, was used in Briffa 2000, not MBH98 and the original hockey stick
    See: http://wattsupwiththat.com/2009/10/01/mirror-posting-yad06-the-most-influential-tree-in-the-world/
    Josh has simply taken artistic license on the claims of accuracy that swirl around dendro reconstructions, and the difficulty in extracting accurate centennial scale reproductions. He writes:

    The reference was to Mann (the pose taken from the recent video) talking about satellites while not being an expert in that field while happily using duff data to pronounce on global temperatures.

    Anthony

    A lot of things on WUWT’s pages could be explained as “artistic license” … and dubiously self-proclaimed “expertise”. Your point about Dr. Mann not claiming dendro is the be-all-end-all of climate reconstructions is a good one.

    I got an answer from Dr. Cowtan at SkS re: my confusion about trend uncertainties:

    http://skepticalscience.com/surface_temperature_or_satellite_brightness.html#115631

    The 2sigma for the RSS ensemble is 0.067C/decade, so about 50% higher than the value from the paper. The 95% range would correspond to a 4sigma range, which leads to a similar conclusion.

    Mears (2011) only included data to 2009. The 2sigma spread of the ensemble I plotted on that period is 0.059C/decade, which explains a part of the discrepancy. I am assuming that the remaining difference comes from the ensemble I plotted being from a later revision, but I’ll ask Carl Mears.

    I’m feeling a little sheepish, but at least I have been set right.

  80. -1,
    My question was intended to be rhetorical. Part of the motivation behind the video was clearly to counter claims that the satellite data is the best data we have. Why – given that – would they be expected to included every instance in which satellite data has been used. It’s clear that the goal of the video wasn’t to discredit satellite data, simply put it into an appropriate context.

  81. anoilman says:

    I was decidedly unsupportive of Cowton and Way initially simply because I wanted to see how arguments played out against it. I haven’t seen anything saying its nutty, so I guess its good.

    How the satellite data is used is critical, and in Cowton and Way’s its use can also be verified by other means. The issue with Satellite data is that it may be chancy as a long term absolute measurement. However the temperature difference between locations may be accurate, as its degradation over time isn’t important.

    GHGSAT is performing differential measurements in order to pinpoint sources of emissions with extreme accuracy. (i.e. Parts Per Billion for Methane) Nothing it measures is absolute, so short of sensor failure I don’t think its results will degrade over time.
    http://www.ghgsat.com/

  82. -1 In that case, exactly why should it be a surprise that the video didn’t mention Cowtan and Way?

  83. lerpo says:

    Mentioning Cowtan and Way probably wouldn’t provide any additional clarity in a video that was designed to address the senate hearing, but it is a paper that Lamar Smith ought to be made aware of. Smith is upset with NOAA for failing to incorporate the satellite datasets. (http://initforthegold.blogspot.com/2015/12/lamar-smith-thomas-karl-guilty-of-not.html) Curiously, when the two are combined (as in Cowtan and Way) we still don’t get the answer that Smith is hoping for.

  84. Pingback: A global temperature record for 2015 – Critical Angle

  85. -1=e^iπ says:

    @ dikranmarsupial

    Because if you want to debunk deniers or contrarians that go ‘but the satellite data’ and effective way is to go ‘yes, thank you satellite data for helping to debunk the ‘pause’ by dealing with coverage bias via Cowtan and Way.’

  86. Willard says:

    > Josh has simply taken artistic license on the claims of accuracy […]

    I might borrow that one.

    Thank you, Willard Tony.

  87. anoilman says:

    -1: Cowton and Way is biased low due to introduction of Satellite data. This is acknowledged in the paper as a source of error. For a good time, see figure 2.
    http://onlinelibrary.wiley.com/enhanced/doi/10.1002/qj.2297

    “Note that GISTEMP, UAH and NCEP/NCAR all show faster warming in the Arctic than over the planet as a whole and GISTEMP and NCEP/NCAR also show faster warming in the Antarctic. Both of these regions are largely missing in the HadCRUT4 data. If the other datasets are right, this should lead to a cool bias due to coverage in the HadCRUT4 temperature series.”

    I highly recommend you read the paper. Its discussions in biases are interesting.. and all point to the fact that they are using very conservative numbers.

    Actually I recommend anyone here give it a read.

  88. -1=e^iπ says:

    @ anoilman –
    “Cowtan and Way is biased low due to introduction of Satellite data.”

    How? They use satellite data as a proxy to deal with coverage bias.

    “I highly recommend you read the paper.”

    I’ve read it.

    “all point to the fact that they are using very conservative numbers.”

    Cowtan and Way might be be conservative due to the methodology of krigging, not due to to using satellite data as a proxy to deal with coverage bias. The satellite data helps make the results less conservative in the hybrid method than in the krigging method alone.

  89. mwgrant says:

    Cowtan and Way might be be conservative due to the methodology of krigging [sic]

    What particular aspect or aspects of the methodology would make kriging estimates conservative?
    Also, ‘might be’???

  90. -1=e^iπ says:

    @mygrant –

    I’ll decline to answer that question because I wish to remain anonymous.

  91. “Isn’t this what Weng et al, Uncertainty of AMSU-A derived temperature trends in relationship with clouds and precipitation over ocean (2014),cited upthread by BBD have already shown?”

    i would look at surface emmissivity. basically I am looking at all the assumptions made at various stages. In the atmospheric profiles there are some assumptions made with regards to to atmosphere and the surface. they are probably sound assumptions.. but it’s just a discipline of checking. so they assume a constant emissivity of .9 .. that is roughly true the question would be have there been any secular changes ( however slight) that might cause issues.

    For example Eli mentions changes in Ice ( possible snow as well)

    the incident EM at the sensor is a combination of EM from the atmosphere ( which you want to measure) and contamination from the surface– which you want to eliminate..

    So look back at what Eli argues.. If there is a ‘drift’ in the measure that drift can come from an change (degradation) of the sensor/receiever.. or it could come from changes or drift in parameters in your retrieval algorithm that are ASSUMED to be static, but which have changed.
    (water vapor and/or land properties)

    Another example would be changes in diurnal cycle. RSS use a diurnal cycle that adjust recordings to local noon.. if the real diurnal cycle has changed over time (1979-present) then that could bias the estimate..

    In any case there are multiple dataset, multiple estimation procedures, and as mears points out the standrd approach is to try to look at all the data and construct a COHERENT view.

    funny.. skeptics tend to be epistemic foundationalists.. and the argument over satellities is a manifestation of that.

  92. “Has anyone being going around claiming that “Cowtan & Way is the best data we have”?”

    hmm.. I’ve come close on ocassion WRT things arctic.

  93. -1,
    Not that you have to answer questions (obviously) but why would answering the question give away who you are?

  94. Steven,
    Well, I guess you haven’t done it in front of a Senate committee.

  95. BBD says:

    -1

    If you are so very strongly associated by publication or blogging with discussions of kriging, how is it that you cannot spell it?

  96. -1=e^iπ says:

    @ATP, BBD –

    I’ll decline to answer those questions because I wish to remain anonymous.

  97. You do realise that you probably mean pseudonymous. True anonymity is pretty hard, as I’ve found out myself.

  98. -1=e^iπ says:

    But thank you for correcting my spelling. I appreciate it.

  99. dikranmarsupial says:

    -1 IMHO that would be deeply misleading because the reason the satellite data suggests a hiatus is because of its greater sensitivity to ENSO (and because it does not measure surface temperatures), but the correction made by Cowtan and Way deals with the Arctic coverage issue. It doesn’t surprise me at all that it wasn’t in the video.

  100. dikranmarsupial says:

    -1 Basically what I am saying is that if you are going to debunk, then the logic underpinning the debunking has to be valid. This is science, not rhetoric (or at least it ought to be); a better approach is to talk about the advantages and disadvantages of the different sources of data.

  101. -1=e^iπ says:

    @ dikranmarsupial
    Fair point. There is also the fact that it has higher error than instrumental data.

  102. Dikran,
    Indeed, there’s also the point that science is meant to be a positive endeavour. We don’t advance knowledge by trying to pick holes in what others have done. We advance knowledge by doing it better. So, we know that there is coverage bias. Cowtan and Way attempts to resolve this. Simply finding reasons to criticise it is not constructive. The ideal way to proceed is to find some improvement on their method, not simply find reasons why it might be flawed.

  103. BBD says:

    ATTP

    We don’t advance knowledge by trying to pick holes in what others have done. We advance knowledge by doing it better.

    But what would become of all the auditors? Surely they would starve, poor things?

  104. dikranmarsupial says:

    ATTP, there is an important place for hole-picking in science, which is picking holes in your own argument, to make sure it is actually valid (e.g. so you can preempt the reviewers of your papers, so they have a better chance of making it through peer-review). Not always easy to do as confirmation bias is human nature.

    I think Carl Mears hits the nail on the head at the end of the video, it isn’t about what dataset is best, we need to look at all the datasets, surface and satellite (and oceans etc.).

  105. there is an important place for hole-picking in science, which is picking holes in your own argument

    True, and I am generalising a bit too much. There may well be scenarios where someone’s work is so poor that it’s essentially invalid. However, even then you’d typically have to demonstrate this in some way by either showing that it’s not possible to do what they claim to have done, or to do it in a way that is valid.

  106. opluso says:

    Think of Pons and Fleischmann claiming to have detected cold fusion. Or Warren and Marshall linking H. pylori to gastric ulcers. Or the OPERA experiment and super-luminal neutrinos. When you first heard each of these news items, did you reject the research or treat them as learning opportunities? Indeed, it sometimes turns out that “crazy” ideas are, in fact, the correct ones after all.

    Science doesn’t dismiss claims as ridiculous or impossible. Rather, the practice of science is the perpetual search for a superior explanation. Even when you “know” something is wrong, it is important to demonstrate why and how the claim is in error.

    That motivation is behind blogs like this one and society benefits from having people who are willing to put out the required effort.

  107. mwgrant says:

    @ -1=e^iπ

    I’ll decline to answer that question because I wish to remain anonymous.

    Well, I won’t break out in warts over the matter. But I am curious–if you can not pursue it why raise it? [Rhetorical–don’t answer–remember, you want to remain anonymous! ;O) ]

  108. Willard says:

    > Science doesn’t dismiss claims as ridiculous or impossible. Rather, the practice of science is the perpetual search for a superior explanation. Even when you “know” something is wrong, it is important to demonstrate why and how the claim is in error.

    There are at least two readings of “demonstrate” at work. The first is the act of providing reasons to dismiss claims, say because they’re ridiculous or impossible. Scientists do that all the time. The second is to repeat these reasons to anyone who’d raise concerns against an established viewpoint — like a teacher would when meeting an incredulous student. Citizen scientists do that all the time.

    Both cases lead to different kinds of “best” explanations because of the target audience. The first is exoteric, while the second is esoteric. The first needs to be convincing, while the second needs to be persuasive.

    In both cases, the demonstration is never completely sealed under deduction, because there’s no such thing as a decisive proof in empirical sciences:

    There are at least two meanings of the word “proof” in usual English usage. I’m going to call them proof1 and proof2.

    proof1: This is the sense used in mathematics: a deductive proof from a set of axioms. In this sense there are no proofs in science (see here as to why: is nothing ever proved in science?)

    proof2: This is the sense in which a scientific theory can be “proved”, by repeatedly surviving harsh tests that were designed to falsify it.

    Most scientists know this. But this makes them even more susceptible to the bait and switch.

    http://initforthegold.blogspot.com/2016/01/spotted-on-quora.html

    The very nature of scientific explanations and demonstrations leads us to conclude that audits never end.

  109. anoilman says:

    -1=e^iπ says:
    January 21, 2016 at 12:39 am

    @ anoilman –
    “Cowtan and Way is biased low due to introduction of Satellite data.”

    How? They use satellite data as a proxy to deal with coverage bias.

    ’cause satellites are measuring lower temperatures. Not hotter… colder. See figure 2. That’s their bias. Actual surface measurements are hotter.

  110. -1=e^iπ says:

    @anoilman –
    … I’m not sure if you understand the meaning of as a proxy…

  111. John Hartz says:

    Recommended supplementary reading:

    Thorough, not thoroughly fabricated: The truth about global temperature data by Scott K Johnson, Ars Technica, Jan 21, 2016

  112. Michael 2 says:

    I followed your link to http://initforthegold.blogspot.com/2016/01/spotted-on-quora.html on the topic of mathematical (deductive) proofs versus inductive proofs. That is a nice presentation.

  113. Eli Rabett says:

    Think of Pons and Fleischmann claiming to have detected cold fusion. Or Warren and Marshall linking H. pylori to gastric ulcers. Or the OPERA experiment and super-luminal neutrinos. When you first heard each of these news items, did you reject the research or treat them as learning opportunities?

    Eli’s initial response on P&F and OPERA was bullshit, because a) the Bunny could think of a whole list of theory and experiments that shouted bullshit and b) Eli has too much experience with complex experiments.

    On Warren and Marshall, well if you are too far out of the field you simply have not valid opinion.

  114. anoilman says:

    -1: I know what a proxy is. The Proxy Cowton and Way used is colder than surface temperatures.

  115. -1=e^iπ says:

    @anoilman – that doesn’t matter if it’s just being used to deal with coverage bias.

  116. opluso says:

    Eli:

    Eli’s initial response on P&F and OPERA was bullshit…

    More recent work on “cold fusion” by Rossi, Lundin and Lidgren at least leaves open the possibility that something very interesting is going on in these experiments. They may turn out to involve equipment or measurement error (like OPERA). But I would suggest that scientific investigation is a more reliable test than deploying a personal BS detector.

  117. Willard says:

    > I would suggest that scientific investigation is a more reliable test than deploying a personal BS detector.

    And I would suggest that the former does not preclude the latter. Informed incredulity is one step away from (both contrarian and orthodox) incredibilism, however:

    http://planet3.org/2012/08/24/incredibilism/

  118. Joshua says:

    ==> ” I would suggest that scientific investigation is a more reliable test than deploying a personal BS detector.”

    Hmmm.

    Have you ever had that convo with Judith and the Denizens?

    http://judithcurry.com/2012/09/15/bs-detectors/

  119. Phil says:

    opluso> I would suggest that scientific investigation is a more reliable test than deploying a personal BS detector.

    WIllard> And I would suggest that the former does not preclude the latter.

    The former is induced by the latter, and there is nothing wrong with that as long as you acknowledge that the latter can be wrong.

  120. Willard says:

    > The former is induced by the latter, and there is nothing wrong with that as long as you acknowledge that the latter can be wrong.

    Indeed, and more importantly, one must be ready to acknowledge that there’s a cultural gap between your viewpoint and the one about which you feel incredulous, e.g.:

    https://www.xkcd.com/808/

    I could have posted 793 again, but it’s time for a change.

    ClimateBall ™ can be modelled as a Rock-Paper-Scissors game between Engineer, Scientist, and Philosopher.

  121. “ClimateBall ™ can be modelled as a Rock-Paper-Scissors game between Engineer, Scientist, and Philosopher.”

    that would make 가위 바위 보 a winnable game

    http://www.telegraph.co.uk/men/thinking-man/11051704/How-to-always-win-at-rock-paper-scissors.html

    then again some players always act like they won

  122. Willard says:

    You call that a RPS machine? That’s an RPS machine:

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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