Global Temperature Update

November marks the end of the “climatological” year, which extends from Dec. through Nov. rather than Jan. through Dec. as the calendar year. Since NASA GISS has updated their global temperature estimate through November, we can compute the global average annual temperature for the climatological year.

And here it is (animated GIF, you may have to click the graph to see the animation):


72 responses to “Global Temperature Update

  1. “November marks the end of the “climatological” year, which extends from Dec. through Nov. rather than Jan. through Dec. as the calendar year.”


    • David B. Benson

      Meteorological winter is Nov, Jan and Feb.
      Check the WMO website.

      [Response: Isn’t that Dec, Jan, and Feb?]

      • ” Isn’t that Dec, Jan, and Feb?”
        Nope, December does not count as it’s climate denial month: the month where there is no climate… No climate, no weather see?

    • I speculate that the conventional climatological seasons (wtr: Dec-Jan-Feb, spr: Mar-Apr-May, sum: Jun-Jul-Aug, aut: Sep-Oct-Nov) and year (Dec-Nov) are because of the thermal inertia of the climate system. (GISS uses this grouping of months for the seasons in their graph (and data) for “Seasonal Mean Temperature Change” at

      So, for example, the NH winter solstice is in the latter half of December when the solar insolation is a minimum, but because of the thermal inertia the center (thermal minimum) of the winter season is not on this solar minimum (Dec), but lags about about a month (Jan). Similarly for the other solstice, and the equinoxes.

    • Timothy (likes zebras)

      I would use the term “meteorological year”, since December to November is the annual period you end up with when you combine the meteorological seasons of winter (DJF), spring (MAM), summer (JJA) and autumn (SON) [in the northern hemisphere].

    • Seasonal weather tends to lag the variations in the orientation of the Earth relative to the Sun so it’s convenient to consider winter as Dec, Jan and Feb, spring as MAM, summer JJA, autumn/fall SON. Closest match to the calendrical year then moves Dec into the following year.

  2. Now, you only need to correct for the Sun and the ENSO to get the corrected value.

  3. What are we seeing there? Why does the charting end in 2000? What is the animation?

  4. I guess the calender year 2012 will rank as the 9th warmest of all years, despite starting with a La Nina and ending with a more neutral ENSO.

  5. Bernard (the other one)

    Obligatory (old) joke: “Look at those last few data points, we’re obviously seeing the beginnings of a cooling trend that will push us into an ice age any year now!”

    On a related note – I’ve finally gotten around to reading Hansen’s “Storms of my Grandchildren”. One rather interesting comment I’ve found so far, is that so long as human civilisation is around, there will be no more ice ages. This is because a single factory producing CFCs and dumping them into the atmosphere would be enough to counteract the subtle forcing that starts ice ages. I think it’s a good point to make when discussing climate with people who doubt that we could have any real impact on the global climate.

    • Rather than CFCs, I believe we should prefer HFCs (no nasty chlorine atoms to deplete stratospheric ozone).

      • Of course, even HFCs will do nothing to forestall ocean acidification as long as the CO2 is in excess.

    • Look, if you consider the mean temp in 2010 and 2012, it has cooled down 0.1 ºC in 2 years. By 2100, we’ll be 5ºC cooler, then – ice age, q.e.d.

      Hey, I’ve got plenty of data to apply for a WUWT post. Certainly good enough for en E&E paper! How much do you think they pay?

    • A few factories doing that would probably be enough to terraform Mars..

      • guthriestewart

        If only. It also needs more atoms full stop, just to bring the pressure up to something liveable. See Kim Stanley Robinson’s Mars trilogy for more information in fiction form. It’s old now but I understand it was properly researched.

      • I second the motion to encourage reading of the Mars trilogy.

      • Snarkrates,
        Not quickly. Mars has 0.38 g at the surface and likely lost some early atmosphere to impact cratering. It could hold on to a terraformed atmosphere for a long time on a human scale. See the equation for Jeans escape.

      • What is more, lacking a planetary magnetic field and being a fraction of the mass of Earth, the atmosphere would quickly revert to the original.

      • OK, yes, there may be the odd issue with a combination of suffocation and depressurization when you tried to walk around on the surface, lack of radiation shielding, and lack of liquid water, bu apart from those teething trouble is would be fine..

        (OK, terraforming was probably not the correct word to use.. and I have read the Mars Trilogy..)

  6. Since 2012 is pretty average for the 21st century and below trend for AGW, the collapse of arctic ice and the crazy weather we’re already getting is more than a bit frightening. The next El Nino we get is going to kick off some wild headlines in the year or two after it…

    [Response: 2012 is hardly “well below the trend for AGW.”]

    • 2012 started pretty cold, but has ended pretty hot: near records for October and November. There have been years where ENSO neutral has lasted 12 months.

      Is it possible to have a hottest year in that circumstance?

    • IMO you’re being too hard on Lamont.

      He didn’t say “well below”, just “below”, and given his comment about the next El Niño, I think it’s clear that he’s implying “since 2012 is pretty average [despite the La Niña conditions early on …] … just wait until El Niño kicks in again [we’ll see a really really warm year and headlines to match]”.

      At least that’s how I read his statement …

  7. Timothy (likes zebras)

    On the animation, I assume that the error bars are those for the trend.

    Do GISS give error estimates for their temperature anomalies, like HadCRUT?

    Showing the error bars on each annual value is, in my opinion, a good way to make a mockery of the claims that the globe has cooled since 1998.

    • Horatio Algeranon

      But if they were for the trend, they’d actually be curved (forming an hourglass shape)

      Unless Tamino has done something different than he did in previous graphs (eg, “You bet”), the upper/lower lines were obtained by simply shifting the trend line (in this case for 1975-2000) up/down by 2-sigma (where sigma= standard deviation of residuals).

      [Response: Right you are.]

  8. As a test of the model, can you use the relationships determined in Foster and Rahmstorf 2011 to “predict” an average temperature for 2012?

  9. Good stuff.
    The lower troposphere temps in the “skeptic” run UAH AMSU (channel 5)dataset is now at record high level. 2013 could be an interesting year, but I would not be surprised if we see a La Nina “triple dip.”
    Remarkable how deniers no longer quote their former favorite UAH dataset, but instead focus on the CRU data that ignore most of the Arctic. The very CRU dataset that they proved “fraudulent”. You can’t make this up.

    • Yeah I like poking them with this pointy stick as well. Haven’t noticed a good response yet. Maybe it’s an unanswerable point.

      I also find some amusement from the observation that the historical surface temperature trend is supposed to be due to the urban heat island effect. The obvious conclusion from the recent CRU trend “plateau” therefore is that the UHI has stopped trending up. Since cities have not stopped growing this would amount to a contradiction. Skeptic land is pretty blind to internal contradictions.

      • Fake skeptic land also appears to assume that scientists are as stupid or uninterested in the truth as they are. They for example claim that the warming since the late 20th Century is all down to the UHI effect and imagine it’s a question that has never been investigated scientifically.That the UHI effect can be removed from the temperature record by looking only at rural locations or only at records taken on windy days when the heat island is dissipated doesn’t seem to cross their minds.

        I think this is also linked to a misunderstanding (intentional or unintentional) of how science works. Fake skeptics appear to imagine that science is a house of cards that rests on one or a few pieces of evidence (such as a single tree on the Yamal Peninsula); if they can pull out one or two pieces of evidence that the whole thing will collapse.

      • Horatio Algeranon

        “Mannstream Science”
        — by Horatio Algeranon

        Science rests
        On a single Mann
        Throughout the Land

        If he shrugs
        The science shakes
        Break his stick
        And science breaks

      • I wonder how much is due to the disconnect between how most science is done and reviewed (minor nitpicking at unresolved questions), versus how it’s taught (skip the boring parts and jump to the major discoveries and paradigm shifts).

      • numerobis,

        One of the problems with denialism in general, not restricted to global warming denialism, is the view that science can be defined as ‘The sum of all statements made by expert scientists’. This view tends to be reinforced by casual media reporting, which tends to focus on a small number of scientists and deals in soundbites.

        This view of science strongly encourages the focus on personalities and rhetoric over substance, and of course disregards internal consistency. As far as the skeptics are concerned, as long as they have ‘Statements from Scientists’, they have evidence.

      • Also, I’ve observed that for some ‘all-purpose denialists’ at least, there’s a clear preference for tin-hat stuff, as opposed to anything vaguely mainstream. I think that’s because one of the emotional ‘draws’ for this type of person is the feeling of being one of the select ‘initiated,’ who have glimpsed the truth veiled from the eyes of mere mortals. IOW, aligning with ‘the few’ better strokes their egos.

      • Kevin McKinney“I think that’s because one of the emotional ‘draws’ for this type of person is the feeling of being one of the select ‘initiated,’ who have glimpsed the truth veiled from the eyes of mere mortals.”

        Agreed. Which is a long-standing tradition – take a look at “The Paranoid Style in American Politics”, by Richard Hofstadter, 1964 ( has a copy), and you will find the parallels with hard-core climate deniers rather disturbing.

        Elements of the ‘Paranoid Style’ include feelings of dispossession, the idea of an omnipotent enemy (giant conspiracies, control of the press!) emulation of the perceived tactics (any dirty tricks are acceptable), self-identification as “renegades” (various ‘Galileo’ claims), etc.

        Sadly, this particular mental distortion is difficult to treat, as any evidence against their views is taken to be even more proof of the conspiracies they see against them…

    • But….. But….. But, the CRU dataset is all fabricated!!!! It’s all a “trick”. They’re “hiding the decline”!

  10. Lamont — look at the 95-percent-probability range without getting your eyes stuck on that solid line up the middle. I wish more charts would instead use shading from faint at the outside to darker in the middle to indicate the band.

    People get hung up on thinking the dark line is real and dots on either side of it are above or below something.

    Heck, the _dots_ should be fuzzy vertical bars or I-beams, not discrete points, to indicate _their_ uncertainty.

  11. PS for Tamino — when you show the shorter time spans, aren’t the confidence limits different than for the combined total time span?

    • Well if it was a white noise model then no, since the standard deviation of the residuals won’t change appreciably as you increase data points (in fact it should converge to a certain value). The method he used doesn’t show trend certainty, so sample size doesn’t matter. It’s supposed to be illustrative of the fact that the trend and variance in the 2000s matches that which you’d expect from a continuation of the previous decades.

  12. Does anyone have the full GISS rankings for 1998 to 2012 inclusive,
    most lists just give the top ten

    thanks in advance

    • This Wikipedia page has a list of the 20 warmest years (NCDC) and a list of warmest decades (NASA).

      Or you could do your own list from this data, which is easy enough to get into a spreadsheet or a table in your favourite WP to run a sort.

  13. The post-2000 data also fall within the “hasn’t warmed since 2001” band from Tamino’s bet graphic. Alas, the winner of the bet has not yet been determined.

    • B. Buckner, And yet 2011 was the warmest La Nina year ever, 2012 saw the Arctic sea ice melt records smashed, and the Mississippi is closed to navigation. Sorry, Dude, you don’t just have tunnel vision, you’re looking down an optical fiber.

  14. Many thanks P Lewis, purfeck

  15. Minor suggestion: if there’s a way to let the last frame of the animation stay up a few more seconds, that would be nice. I kept wanting to look at it a bit longer.

    • Here’s a work-around for pausing the animation (works in Firefox, possibly in other browsers also with appropriate changes):

      Open the animated GIF in a separate window or tab. Drop down the File menu and hover your cursor over the Print Preview option. When the GIF has reached the desired form, left click on Print Preview.

  16. John Archington

    What we see here is a decreasing slope (although still in the error bars), and a constant slope when corrected from the natural variations a la Foster and Rahmstorf. But even if the data are still compatible with a constant trend, there is however a slight issue : the CO2 forcing has increased by around 30 % since 1990

    [Response: No it hasn’t. Those figures should more properly be called “radiative forcing *anomaly*” because they’re the difference between present and pre-industrial values. Your thinking it’s a “30% increase” reveals a fundamentally flawed understanding.]

    wouldn’t one expect also a 30 % increase in the slope [Response: No. Not even IF forcing had increased by 30% (which it hasn’t). You are operating under a fundamentally mistaken premise.], which should be detectable ?
    The 0.15 C/decade rate would give only 1.4 C in 2100. Without a strong acceleration, it would be impossible to reach much higher values of course.

    [Response: Your entire claim is SO wrong, one wonders whether you even care about getting at the truth of the matter. To put it another way, your understanding of the science is a farce.]

    • If we expected the temperature slope to grow with forcing, that would mean we think temperature is *quadratic* in forcing. What you’ve (sort of) shown is that there’s no obvious statistical support for that theory. Congrats!

      Can you point to anyone who believed the quadratic relationship? All the scientists expect a linear relationship (if they agree the index represents forcing accurately).

      • It is not quadratic, but it is non-linear. However, the expected temperature rise can be approximated as linear if we only look at a few decades. On such timescales, any expected deviation from linear is a) very small, and b) dwarfed by the variation.

  17. > there’s a way to let the last frame of the animation
    > stay up a few more seconds

    With most browsers,
    “esc” key stops any animation
    on whatever frame is then displayed
    Reload to start the animation again.

  18. There is little argument that the rate of warming has temporarily slowed over the past few years (and it will probably speed back up again). Any comments about a relationship to the negative phase of the PDO?

    [Response: Since the “temporarily slowed” is completely accounted for by el Nino and solar variation, there’s really nothing to explain. A relationship to PDO seems highly unlikely.]

  19. Question: It seems that there has been a slow-down in the global temp warming for the past decade or so. I have read of 2 major speculations about why this may be the case : aerosals and/or deeper ocean warming absorbing much of the heat from our out-of-balance heat budget (due to GHGs). I am wondering about an obvious ‘visual’ analogy- when you look at the warming from 1940-1970 it clearly also slowed down. That period is ascribed mostly, if I am not mistaken to the massive aerosals emitted by the exponentially industrializing West. Does it not also follow, then, that the equally massive aerosal out-put from India and China has had a similar temporary ‘blocking’ effect over the last 10-12 years? We have continued warming, albeit at a slowed rate, ‘fighting against’ lower insolation, frequent La Ninas, etc- does it not follow that the warming would likely have remained as robust (if not more so) as the 1975-2000 period if not for the new aerosal output?

    [Response: Since the “temporarily slowed” is completely accounted for by el Nino and solar variation, there’s really nothing to explain. Look at the graph again. The departures from the trend line are what would have been expected. If such fluctuations did not occur, that would require explanation.]

    • Tamino, how is what you are saying different from what Tsonis and Swanson are saying? They have found a mechanism for the change in direction of the GMT temperature trend. They conclude the directional changes in the 20th Century caused by this mechanism were entirely natural, but that natural variability rhode a continuously rising AGW signal: sometimes enhancing it and sometimes depressing it.

      They think such a change occurred ~2000, and that the new climate regime has a flattish GMT for several decades at the start of 21st Century and a persistent rise in OHC in the deep ocean. So for these first few decades, they see natural variability depressing GMT.

      [Response: I think that their claims of “regime shifts” are fundamentally flawed. They’re too enamored of their nonlinear dynamics to realize this. I admire enthusiasm, and believe that their research path is both interesting and important, but I think it’s still in its infancy.]

      • I’ve found no serious criticisms of their work.

        In one paper they say in the 20th Century aerosols are consistently overcome by the CO2 signal, and seem to conclude they could have nothing to do with mid-20th C cooling/flattening or late ~70s warming. I may be misinterpreting what they mean, but that seems to me to be what they are saying.

      • Horatio Algeranon

        “Putting the neutrino before the photon”
        — by Horatio Algeranon

        Faster than light
        Neutrinos might
        Result from wormholes
        Yes, that’s right.

    • I may be wrong–or at least out of date–but I don’t think the question of attribution of the cooling episode you mention is considered to be settled. And the effects of aerosols are not simple, and in fact can be either cooling or warming, depending upon the physical characteristics of the specific particulates. So, no simple answers to your question.

      As a side comment inspired by Tamino’s response, I must say that I find it rather odd that some of the same folks who in 2009 were predicting an imminent Little Ice Age based upon the ‘quiet sun’ which we have in fact been experiencing, are now the same ones who are crowing about ’16 years of non-warming!’ as if they’d never even heard that the sun plays a role in climate… this, despite the fact that Dr. Hansen estimated back then that the (expected) dip in solar output was unlikely to exceed the radiative equivalent of 7 years’ worth of greenhouse emissions.

  20. The recent furore on Watts re some leaked IPCC pages which refer to the GCR explanation (Svensmark’s stuff) left me wondering what the odds are of a GCR change (that did not occur) causing warming at EXACTLY the time that we released all this lovely CO2 (that can’t possibly be a problem) so as to fool all the scientists on the planet. ;-)

    Actually thats sort of a serious question which I am not certain how to approach. What WOULD be the odds of that sort of coincidence? Clearly some assumptions have to be made about how often the GCR thing actually might happen…. not clear… but I figure that if anyone can help with this it’d be the folks who hang out here.

    • It could also be that we have an (unknown) natural cooling effect that is offsetting the calculated-from-basic-physics greenhouse effect, plus another (unknown) natural warming effect which is by coincidence behaving just like greenhouse-driven global warming would look like, if it wasn’t being offset by the unknown natural cooling effect.

      Anyway, using my Wattsian Post-Sane-Science Probability Calculator, the probability of this being true is 1, because of taxes. So there you go. Surprise result.

      • He would dock you points for using weasel words like “it could be”. Then he’d dock you more points when you express certainty. But you get some of them back for mentioning taxes, although you’d have gotten more back if socialism entered the picture somehow.

      • Anthony’s Razor: That hypothesis is best which leads to the most pleasing conclusion.

  21. The animation looks a lot like mine, though I haven’t updated it a couple of months:

  22. I used the temperature slope tools at SkS for GISS, HadCru, RSS, and UAH and saw very different results from what you are describing here (to use a round number, I ran it for the past 15 years).
    I’m hoping to better understand why RSS and UAH had negative slopes, HadCru had an ever so slight positive slope, and GISS had a higher positive slope.
    Taken together, they don’t paint as clear a picture as you are describing here.

    • Horatio Algeranon

      What Tamino is showing is different than what the SkS trend calculator graphs..

      In short, Tamino’s graph shows that recent temperatures (eg, since 2000) are consistent with the “extended long term trend line (from 1975-2000) . The “scatter” of recent temperatures about that extended long term trend line is consistent with what one expects to see based on temperature variability due to short term ‘weather” (El nino, la nina, etc)

      The SkS graphs show the ‘apparent” trend (central line) for the specified period (eg, last 15 years), along with the “uncertainty” (2-sigma error bounds) for that trend. DON”T ignore the uncertainty (+-). The “actual” trend most probably lies somewhere between those bounds. You can draw lots of lines that lie within the hourglass envelope and the ‘actual trend” is most probably represented by one of those lines (though you can’t say which one).

      In a little more detail:
      What Tamino’s graph shows in red is an extension of the long term trend line (from 1975-2000) It is important to point out that the center red dotted line is NOT the central “trend” line for the past 15 years — or even past 12 years – shown by the SkS trend calculator)

      And the “boundary lines” (dotted) on Tamino’s graph are obtained in a different way — and represent something different — than the curved (hour glass shaped) lines on the SkS graph. On Tamino’s graph, these lines are meant to bracket the expected “scatter” of the year to year temperature about the central trend line . On the SkS graph, the bounds bracket the uncertainty in the trend line itself over the specified period.

      Finally, it is absolutely CRITICAL to pay attention to the (2-sigma) error (+-) on “the trend” given by the SkS trend calculator (eg, for the past 15 years). The central line is only the “apparent trend”. The “actual” trend most probably lies between “Central value + error ” and “central value – error” but one can not say what the actual value is with certainty.

      So, for example, the GISTEMP trend (for the last 15 years) obtained with the SkS tool comes out 0.098 +- 0.143 which means the ‘actual” trend is likely (at 95% confidence) to lie between 0.241 and -0.045 deg C per decade.

      Note that this is a pretty big range AND that the range of values for the “probable trend” encompasses the long term trend calculated with the SkS trend grapher for 1975-2000 , which comes out 0.165 +- 0.077. (also shown with Tamino’s central blue line)

      In other words, even though short trend ‘appears” to be different than the long term trend, appearances can be deceiving – and, given the uncertainties, the two are actually consistent with one another.

      So, at the core, there is really no inconsistency between Tamino and SkS

    • Jack,
      Tamino found in the course of the paper he did with Rahmstorf that the ENSO index affected the satellite based measures more than the land-based measures. It appears that that ENSO has a bigger effect on lower troposphere readings than on near-ground readings. The past few years have had more La Nina than El Nino.

  23. My apologies:
    GISS is mildly positive, UAH is slightly positive, HadCru3 is very slightly negative, and RSS is slightly negative. Mixed them up in my head (yes, my head is spinning trying to keep up with this).

    • Why, at this point, would you use HadCRU3? It’s known not to capture Arctic warming, which is a big reason why there is now HadCRU4.

      Why RSS has been trending lower has not to my knowledge been publicly explained. Dr. Spencer of UAH suspects an uncorrected bias of some sort, perhaps stemming from orbital changes (though I’d thought that all that was well under control by now.)

      woodfortrees doesn’t quite reproduce your result:

      But why are you futzing around with 15-year trends anyway? They are not going to be statistically robust anyhow. What’s the point?

      • Philippe Chantreau

        Well Kevin, perhaps the point is that one can start in 1998 if doing a 15 yeat “trend.” Since that was the year of the giant El-Nino, such trends could be quite attractive if one prefers to see things in a certain light. Back at SkS I asked Jack O’Fall what were the “trends” starting in other years, even just 1996; I guess he hasn’t run these numbers yet.

      • But if you want a debate, you have to stick to 15 year trends. And its going to be a real bugger when we have the next big El Nino. Imagine how unconvincing “no statistically significant warming in the last year” is going to sound…

      • The denialist argument morphed from no warming since XXXX to no statististically significant warming since XXXX where they leave out the “statistically significant” wherever they like. I predict they’ll keep doing this for a few years yet. It’s possible that in October 2017 they be loudly trumpeting that there’s been no (statistically significant) warming (in Hadcrut3) for 20 years! Cherry-picking and quotation out-of-context extra-ordinaire.