SkepticalScience Misrepresents Their Animation “The Escalator”

UPDATE Nov 25th:  The animation disappeared for some reason.  I’ve added it back into the post.


Apparently, one of the proudest achievements of the website SkepticalScience is their “Down the Up Escalator” gif animation. They prominently display it in their right sidebar. The intent of the animation is to show that global temperature anomalies can flatten or cool over decadal or shorter periods while warming over the long term. The first version was created using the Berkley Earth Surface Temperature (BEST) land surface air temperature dataset. That, of course, made SkepticalScience appear two-faced, because the papers associated with the BEST dataset had not yet appeared in any peer-reviewed scientific journals and SkepticalScience downplays any research efforts that haven’t been peer reviewed.


Bogus Escalator

In an effort to distract from their duplicity, SkepticalScience revised and reissued the animation (modified screen cap above), using the average of the GISS LOTI, HADCRUT4, and NCDC land surface air plus sea surface temperature anomaly datasets. If you were to click on the mini “Escalator” animation along the right-hand side of their main page, you’re brought to the updated GISS-, UK Met Office- and NCDC-based Escalator. SkepticalScience describes “The Escalator” animation as (my boldface):

Average of GISS, NCDC, and HadCRUT4 monthly global surface temperature anomalies from January 1970 through August 2012 (green) with linear trends applied to the timeframes Jan ’70 – Oct ’77, May ’77 – Dec ’85, Jan ’86 – June ’94, Nov ’94 – Dec ’00, Jan ’01 – Aug ’12.

You’ll note that they’re now calling it “The Escalator”—no longer calling it the “Down the Up Escalator”. Yet each of the steps in their escalator clearly shows a short-term trend that’s flat or cools slightly.

SkepticalScience misrepresented the trend of the “fourth step”. The time period they selected is November 1994 to December 2000. As it turns out, the only dataset that shows a flat trend during that period is the GISS Land-Ocean Temperature Index (LOTI). Both HADCRUT4 and NCDC have significant warming trends from November 1994 to December 2000 at about 0.08 to 0.09 deg C per decade. The average of the three datasets is approximately 0.06 deg C/decade, and that is a significant warming trend.

Actual Linear Trends During Fourth Step of Escalator

How significant is that 0.06 deg C per decade trend? It’s comparable to the trend in global surface temperatures since 1880.

GISS LOTI Global Surface Temperature Anomalies Since 1880

The following animation will give you an idea what “The Escalator” would look like if SkepticalScience had used the real linear trend for the fourth step. Depending on your browser, you may need to click on the following gif animation.

The Escalator With Actual Linear Trends


As noted in my WUWT-TV presentation “The Natural Warming of the Global Oceans”, “The Escalator” is an exercise in cherry-picked start and end dates. Proponents of anthropogenic global warming will incorrectly cite “The Escalator” during my blog discussions of ENSO-related upward shifts in Atlantic-Indian-West Pacific (90S-90N, 80W-180) sea surface temperature anomalies. (In recent years, I typically present that dataset as the “Rest-of-the-World”, because I usually now start with the East Pacific data, which shows no warming over the entire 30-year term of the satellite era.) When the disciples of SkepticalScience link “The Escalator”, they are simply trying to distract from the process-caused shifts. Those natural processes were described in the WUWT-TV presentation and detailed with numerous datasets in my book Who Turned on the Heat?

Rest of the World Sea Surface Temperature Anomalies With Linear Trends

The above graph was presented in the post Does The Sea Surface Temperature Record Support The Hypothesis Of Anthropogenic Global Warming?

Proponents of anthropogenic global warming have another, related, inconsequential complaint about my research. They claim the decade-long time periods between the 1986/98/88, 1997/98 and 2009/10 El Niño events are too short for the trends to be significant. Curiously, when SkepticalScience is trying to make a point, they have no trouble presenting a series of decadal trends, and when SkepticalScience is trying to mislead their followers, they have no trouble misrepresenting the trend for a shorter 6-year period. Apparently, linear trends over periods as short as 6 years do have significance. Looks like another example of the double standards of the proponents of anthropogenic global warming.

In the next few days, I’ll present PBS’s sleight of hand about “The Escalator” when they presented it in their Frontline report Climate of Doubt. They were pretty blatant about it.

About Bob Tisdale

Research interest: the long-term aftereffects of El Niño and La Nina events on global sea surface temperature and ocean heat content. Author of the ebook Who Turned on the Heat? and regular contributor at WattsUpWithThat.
This entry was posted in CAGW Proponent Arguments, El Nino-La Nina Processes. Bookmark the permalink.

18 Responses to SkepticalScience Misrepresents Their Animation “The Escalator”

  1. kim2ooo says:

    Reblogged this on Climate Ponderings and commented:

  2. R says:

    What’s the error bounds (when accounting for autocorrelation) on the trendlines you presented above. Technically they are correct if the trends are not statistically significant. Are they? Can you demonstrate that the trendline for any portion is statistically different from zero?

  3. Bob Tisdale says:

    R: The discussion is not if the trends are statistically significant; it’s about the misrepresentation of actual trend lines in their animation. If they had used the actual trend lines, the escalator would appear to have a broken step and the visual doesn’t work:

  4. Glenn Tamblyn says:


    You have completely and utterly misunderstood the point of ‘the Escalator’. And posts such as this simply reinforce that utter lack of understanding – unless of course you totally do understand it, and thus understand that you have to tear it down.

    The Escalator is a graphical way of portraying a psychological observation!

    It isn’t about whether this trend or that was calculated this way or that. It is totally and only about why, when presented with data such as the total temperature series, deniers squirm and wriggle and try to find ways to make it go away. To deny the implications of it, to try and contrive a basis for rejecting it. So they cherry pick a short period in a graph such that they can say: ‘Ahah, SEE, it isn’t really warming’. Then some years latter they pick another period and have another Ahah moment. Without ever for an instant recognising the disconnect between the two ‘moments’

    The Escalator is an illustration of a psychological phenomenon. Not a precise mathematical analysis. What folks will do rather than accept something they can’t accept. How they will misrepresent data to conform to their emotional needs.

    But this then leads to another phenomenon. The rather obsessive need to deconstruct and micro-analyse the tea-leaves that is found among some deniers; presumably they think that in doing this they will find something, anything, of significance; Hockey Stick addictions, some weather stations in the US aren’t perfect so the world isn’t warming, Someone said something in an email so it ‘blows the lid off’, well, whatever.

    The psychology of obsession.

    So here you are Bob, micro-analysing an irrelevent detail of a graph, utterly unrelated to the actual purpose of the graph which is 101% psychological. But then the psychological purpose of the graph can never be admitted because that opens Pandora’s Box doesn’t it.

    Best to segue off to looking at it another way – lets micro-analyse some numbers because of course we all KNOW that is REALLY what it is about. We can’t, of course, ever accept that it’s purpose is to actually get us to question or reconsider our attitudes, obsessions and biases by highlighting them.

    That it’s purpose might be to help us to abandon our obsessions by first recognising their existance through the logical absurdities we are led to in order to prop them up.

    And the hoops we might jump through to avoid confronting any of this.

    So Bob, from the community at SkepticalScience, thanks for providing a real world, observational confirmation of everything we are saying.

    Feel free to return to obsessive micro-analysis of the minutiae at your leisure Bob.

  5. R says:

    Perhaps its important to explain something – if the trend is not statistically significant from no trend then it is virtually meaningless to have a line there at all. Whether its a flat one or a little bit inclined changes nothing – because you can’t prove that your one is more right than the other.

  6. Bob Tisdale says:

    R, perhaps it’s important for you to acknowldege the fact that SkS felt they had to change it without noting the modification, in effect misrepresenting the data for a visual effect.

  7. Bob Tisdale says:

    Glenn Tamblyn says: “The Escalator is a graphical way of portraying a psychological observation!”

    Glenn, you seem to miss the irony of SkS having to misrepresent a trend line in an escalator step in order to portray the effect. The followers of SkS bought it hook, line and sinker. It’s quite entertaining that no one bothered to question, to check. You also haven’t bothered to look to see how easy it would be correct the problem, yet you’ve wasted a tremendous amount of time writing your above comment, supporting the broken escalator.

    SkS has two choices: They can argue about it, as you are doing, which is a waste of time. Or they can acknowledge the mistake, correct the problem and move on, which you aren’t doing. Make sure you give me credit for finding the error, when the mistake is eventualy corrected. You can also give me credit for showing you how to correct the problem, since the answer lies in one of the graphs in the above post.


  8. R says:

    “Bob Tisdale says:
    November 26, 2012 at 4:23 pm
    R, perhaps it’s important for you to acknowldege the fact that SkS felt they had to change it without noting the modification, in effect misrepresenting the data for a visual effect.”

    I don’t really care what SKS did or didn’t do. What I’m saying is you can’t use bad statistics to disprove bad statistics. Saying your excel trendline is better than theirs is meaningless when they’re statistically indistinguishable (which they are). There is no trend in the data for any of those periods that is distinguishable from statistical noise. It is essentially handwaving to say your meaningless trends matters more than their one.

    But then in addition you bring up misrepresenting data for visual effect when you present no error bounds in any analysis above and none of the graphs you show have any statistical meaning (with the exception of the GISS global one which is surely significant).

  9. Bob Tisdale says:

    R: Here’s the bottom line. How data is presented is up to the person doing the presenting. If you’d like to see data from this post presented in some other way, feel free to do so. Create your own blog and present it.

    Now, let me ask you a couple of questions: Have you asked SkepticalScience to present the statistical significance of each of their trend lines in their Escalator? Have you complained to the authors of posts at RealClimate when they present trends without your desired info?
    If not, why are you belaboring the point here?

  10. JohnB says:

    B.T. – If they had used the actual trend lines, the escalator would appear to have a broken step and the visual doesn’t work:

    As I noted on WUWT, SkS had a bigger problem with Escalator 1.0 (not just because of using BEST). For some reason known only to SkS they used six steps with the last/top three steps overlapping each other in a very un-Escalator-like manner.

  11. Kristian says:


    Have you checked out the ERA Interim Reanalysis on Climate Explorer? I calculated their absolute power values for the four surface energy fluxes (DSR (pos.), OLR, latent and sensible heat (neg.)) from 1979 to 2012. Got some very interesting results. It turns out that of the three negative fluxes (surface heat loss mechanisms) only the change in latent heat transfer really matters. Also, they’ve all grown more negative (more efficient, that is) globally during the modern warming. As one would expect. The net IR flux UP from the global surface has increased with approx. 1 W/m^2 between 1979 and 2012. No trace of an increased GHG insulating effect that is. It seems to be a function of the increasing difference in temperature between the surface and the layer of air right above it over the same period, which points to a surface-controlled atmospheric warming rather that the opposite (the AGW ‘mechanism’).

    The global energy imbalance since 1979 results from more coming in from the sun than is being released. The gap, though, is gradually closing – we’re moving towards zero. Mostly because of the increased transfer of latent heat from the surface to the atmosphere.

    ERA 40 and NCAR/NCEP reanalyses indicate that global latent heat transfer for some reason was much more effective during the period before 1970-75. Hence a gradual net loss of global energy content back then. We’ve seen the opposite during the last 30-35 years, though at a slowing rate.

    The key is in the latent heat transfer. Not (at all) in surface radiation.

    I have figures and graphs, but I can’t access them right now. Later.

  12. Bob Tisdale says:

    Thanks, Kristian. I’ll await the graphs and your further discussion.

    One of these days you’ll have to open your own blog.

  13. Kristian says:

    Here is net global surface solar radiation (SSR) from 1979 to 2012 (ERA Interim Reanalysis of the ECMWF – data downloaded from KNMI Climate Explorer):

    Here are the other net global surface energy fluxes – sensible heat (green), thermal radiation (STR) (red) and latent heat (blue):

    Subtracting the sum of the three outgoing net fluxes from the incoming net solar flux gives this net surface energy balance curve for the Earth as a whole from 1979 to 2012:

    Robustly positive all along, yet still trending unmistakably downward and now finally getting pretty close to perfect balance – maybe withing 3-5 years we’re there, crossing the line … The mean imbalance between incoming and outgoing (1979-2012) is +7,22 W/m^2 (which sounds like a lot).

    This is still according to the ECMWF of course.

    Just out of curiosity I made a running total on the data behind the plot above. It came out like this:

    So the funny thing is, even though all of Earth’s net surface heat loss fluxes has steadily increased in strength/efficiency (becoming more negative) since 1979 (sensible heat by ~0.8 W/m^2, STR by 0.8-1 W/m^2 and latent heat by ~6 W/m^2, to a total of 7.6 – 7.8 W/m^2) and with the mean net solar input to the Earth system today pretty much equal to what it was in 1979, Earth has been accumulating a LOT of energy/heat. The global solar input has simply been larger the last 33-34 years than the output from Earth’s surface, the heat loss processes working hard to catch up. And that’s the funny bit. According to AGW theory, what would cause the energy imbalance is a DEcreasing of the total net upward heat flux from the surface. For instance, in a theoretical steady state, with solar IN (considered static) exactly balanced by IR+latent+sensible OUT, more GHGs would indirectly lessen the total heat flux from the surface, making it less negative (more positive) which would then create the observed positive imbalance. But this theoretical course of events is quite the opposite of what apparently actually happens in the real world. Here the IR flux, the sensible heat flux and the latent heat flux are all increasing as a function of surface temperature. Or should we say, as a function of the increasing difference/divergence between the surface temperature and that of the air layer directly above it. If the standard AGW hypothesis were right, the lapse rate should lift the mean temperature level off the ground with increasing concentration of GHGs in the atmosphere. That is to say, the incremental same-temperature levels would be situated gradually higher from the tropospheric mean emission height on down to the surface. This means that in the end, the layer of air just above the ground/sea would warm independently of the surface (a tiny bit) and would thereby reduce the temperature difference between the air and the surface, reducing the total net heat flux from the ground/sea. For this to be the case, though, we would have to observe just that, that there is a gradual decrease in the temperature gap between the surface and the air layer adjacent to it. If this gap were instead to INcrease, then the surface cannot be the follower. Then the surface is the driver. Which is what all common sense is telling us is the case. Look at these two graphs:

    This is ICOADS SST vs. ICOADS Tair. The first graph covers a large chunk of the Pacific Ocean (30N-40S, 150E-100W). The second a significant part of the North Atlantic (62N-0, 60-15W). Watch how the SST trends are distinctively steeper than the Tair trends in both diagrams. How would an air layer colder than the surface and at the same time with a lower warming rate force the warming of the surface?

    Finally, I did the same operation for the tropical Pacific fluxes as I did for the global ones. Here is the result, directly compared to the global (tropical Pacific (24N-24S, 120E-80W) (black), global (red)). From top to bottom – solar, sensible, IR and latent. Watch how much more positive the solar is in the Pacific and accordingly how much more negative the latent heat flux is. For the other two fluxes the difference seems inconsequential:

  14. Kristian says:

    I have to correct myself on one point. I said: “(…)we would have to observe just that, that there is a gradual decrease in the temperature gap between the surface and the air layer adjacent to it.” We wouldn’t necessarily have to observe a gradual decrease in the temperature gap. They could just both steadily rise in step (the increase in GHGs/emission height is gradual/incremental, after all). BUT, we would still NOT expect to see an INcreasing temperature difference.

  15. Bob Tisdale says:

    Kristian: I haven’t had the chance to look at your two recent comments yet. Hopefully, early next week I’ll be able to do so.


  16. Kristian says:

    I’ll refer to Dee et al. 2011, “The ERA-Interim reanalysis: configuration and performance of the data assimilation system” - – for a thorough discussion on the ERA Interim project. What comes plainly out when reading the document, is that the reanalysis model has overestimated the mean surface solar input:

    “Due to a programming error in the calculation of incident solar radiation as a function of solar zenith angle, the global solar radiation in ERA-Interim is overestimated by about 2 W/m^2.”


    “For solar irradiance, ERA-Interim uses a constant value of 1370 W/m^2 throughout, i.e. no account is taken of the solar cycle. Variations due to the varying distance between the Earth and the Sun are incorporated as described in Paltridge and Platt (1976).”

    According to the newest satellite estimates, the mean solar irradiance is ~1361.7 W/m^2 – . PMOD’s average since 1978/79 is ~1366 W/m^2. ACRIM’s is ~1361 W/m^2. The range in total irradiance between high and low within each cycle is ~1 W/m^2 with PMOD and ~1.5 (1-2) W/m^2 with ACRIM.

    Disregarding the solar cycle amplitudes won’t affect the long term average (over several cycles). It will, however, affect the decadal variation. That means the graphs I’ve presented (based on ERA Interim) show less variation than reality. The 8.3 W/m^2 (1370-1361.7) difference between model assumption and real-world measurements is significant. It will probably overestimate the average energy input from the Sun at Earth’s global surface by ~1 W/m^2 (1361.7/8,3 = 164 W/m^2).

    Dee et al. continue:

    “The energy balance at the top of the atmosphere in ERA Interim has improved, with an estimated energy loss of 1.2 W/m^2 (7.4 W/m^2 for ERA 40). However, the energy balance at the surface boundary is poor in ERA Interim, with a global value of 6.9 W/m^2 (3.8 W/m^2 for ERA 40). This degradation occurs primarily over oceans and is associated with an increase in net solar radiation there. Over land the surface energy balance actually improves in ERA Interim, to 0.5 W/m^2 (1.3 W/m^2 for ERA 40).

    Källberg (2011) suggests that the model clouds are the major contributor to the imbalance in surface energy, based on a correspondence between spin-up/spin-down of cloudiness and of the net energy fluxes.”

    My own calculated mean value for the global energy balance (1979-2012) turned out to be +7.22 W/m^2. Dee et al. finds a +6.9 W/m^2 imbalance (1979-2010).

    Based on the quotes above it seems justified adjusting the ERA Interim solar input down by 2+1= 3 W/m^2. This would reduce the global net energy imbalance 1979-2012 to 4,22 W/m^2 (second graph below), which actually sounds AND looks much more plausible than the original +7,22 (first graph below):

    Note how in the lower graph (the ‘new and improved’ +4,22 one) we’re already very close to perfect balance and have been so for a few years, quite on the verge of crossing the line into negative territory.

    Here are the running totals (accumulated energy) for the +7,22 and the +4,22 scenarios:

    We’re obviously at the summit plateau.

    What’s very interesting to observe, is how the evolution in Earth’s energy balance seems to follow the same pattern as ENSO East (NINO3.4) does. One might imagine an oceanic equilibrium line, across which the Earth system fluctuates in giant cycles. Below the equilibrium line the ocean’s heat loss is on average greater than the input from the Sun. There is a net loss of energy content. Above the equilibrium line the situation is reversed. There is a net builup of energy content. The main regulating mechanism seems to be the rate of evaporation from the ocean surface.

    From the 70s to the 80s this equilibrium line was somehow crossed. The Earth system shifted from a negative to a positive balance. And here’s the take-home message: After the shift is completed, the trend starts falling at once, on its way back towards the equilibrium line. The initial divergence is gradually and steadily reduced. But the positive energy imbalance is still there all along. Energy is accumulating in the system, only at a slowing rate until it finally reaches zero. We’re very close now to that point.

    Compare this to the MEI curve. What do we see?

    A mighty upward shift in 1976/77. Before this, the curve is generally running below the zero line. After, the curve is generally above. But what about the trend? It starts falling directly from 1977 onwards. It’s basically negative all the way ’till today. Yet the ENSO phenomenon has made the world warmer and warmer during the same period. Since a few years back now the MEI/NINO3.4 curve fluctuates around the zero line, straddling the border between El Niño and La Niña dominance.


  17. Kristian says:

    Sorry, more confusion. Here is the ERA Interim +4,22 W/m^2 global energy balance curve:

    And here is the running total for the +4,22 W/m^2 imbalance scenario:

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 )

Google photo

You are commenting using your Google 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 )

Connecting to %s