Global Warming’s Record Year

It is widely publicized that 2016 will certainly break the record for yearly average global temperature. Again. This will be the third year in a row we’ve set a new record. It’s time we paid attention.

I’ve often emphasized that just because Earth shows an indisputable warming trend, that doesn’t mean every year will be hotter than the one before. In addition to trend, there is also a lot of fluctuation in things like global temperature. So we shouldn’t expect each year to break the temperature record.

But we did in 2014. We did again in 2015, by a substantial margin. We did again in 2016, by a substantial margin. The third year in a row of record-breaking global temperature will probably get the most attention, but it may not be the most important or most worrisome record set last year.


It’s also the second year in a row that we’ve broken the record for temperature over land areas only.


We set a new record low for annual average sea ice extent in the northern hemisphere:


In addition, we set a new record low for annual average sea ice extent over the entire globe:


Despite the importance of the record-breaking levels of important indicators of global warming, more worrisome is that 2016 set a new record for the atmosphere’s concentration of CO2.


Certainly we haven’t really made any progress toward reducing our greenhouse-gas levels. But even more troubling is that fact that we haven’t even made progress toward slowing the growth of atmospheric CO2. The year 2016 brought a new record high amount of increase in CO2:


We’re not just failing to gain ground, we’re losing at a record pace.

Just when we most need to slow our greenhouse gas increase, it’s actually accelerating. And there’s also a new, worrisome record: the U.S. government’s new administration is now in the hands of a record number of the most vile and idiotic climate deniers imaginable.

This blog is made possible by readers like you; join others by donating at Peaseblossom’s Closet.

46 responses to “Global Warming’s Record Year

  1. From your graph the last record _growth_ in CO₂ concentration was in 1998, an El Nino year as well. That can’t be a coincidence can it? I assume the global temperature has an effect on the natural release and uptake of CO₂. The question is, is that a long-term thing or a transient for particularly warm years?

    [Response: I’m no expert on the subject, but I believe it has more to do with the el Nino effect on precipitation patterns (and therefore plant growth). Genuine experts are invited to chime in.]

    • Ed Davies,
      The increase in CO2 levels over the 1997/98 El Nino didn’t drop down to ‘normal’ until May 1999. (See red plot here (usually 2 clicks to ‘download your attachment’). May 1999 is the first month when the red 12-month increase trace dropped below the red smoothed trace.)
      The CO2 increase from the 2015/16 El Nino probably will also take a similar time. The rise in rate-of-CO2-rise during the 2015/16 event peaked a little higher than 1997/98 but given the emissions are greater you would expect perhaps a peak 0.4ppm/yr higher. which is what we see (using monthly data, 4.14ppm/yr compared with 3.73ppm/yr back in 1998).
      It will be nice to get back to normal because the fossil fuel emissions are reportedly plateauing and the impact of that may give us something to be encouraged about, perhaps an end on the acceleration of the CO2 rise. Of course, the non-fossil fuel emissions (ie land use) are a big factor so there may not be much to cheer.

      • Emissions growth is slowing. However forcing and temp is not. Going above a certain temp some natural sinks might become sources …

      • DrTsloul,
        I feel you pull your punch. You say “Going above a certain temp some natural sinks might become sources.” Temperature rise ‘will’ create sources from sinks. (To be pedantic, reducing atmospheric CO2 levels will also turn sinks into sources, but let’s not go there.) And we could add that Greenland will begin an unstoppable melt-down above a certain temperature, raising sea level by several metres.
        Of course, we should be mindful that the “certain temperatures” are not so well known and that there is a significant time period required by those “certain temperatures” to trigger significant new CO2 sources or to set Greenland’s ice cap on a one-way ticket down into the oceans. But the processes are certain, real and dangerous.

    • This link suggests the correlation has to do with plant growth:

    • Although I would not call myself an expert, I do know there are several explanations. The modeling study in the link below gives a possible clue:
      but also should be considered with caution (“reasonably well”…)

      • Marco,
        During El Nino years the atmospheric temperature goes up, but ocean heat content actually is reduced.
        I don’t think this is true, or perhaps just badly phrased. During an El Nino year, global warming continues and so the OHC, form that source, will increase, The effect of the El Nino will be to reduce OHC, so the observed OHC will be the sum of the two. I believe this is what got Watts in such a pickle over Hausfather 17 since he naturally assumes the Earths climate is stuck immobile in the paws(*) of a giant unicorn and therefore the OHC must have decreased in 2016. The fact that Nick Stokes showed that Watts was wrong (see comment thread in above HotWhopper thread) suggests to me that AGW “outstripped” El Nino in this regard in 2016, and is further evidence that “the paws(*)” is an article of faith for Watts and his ilk.

        * This mis-spelling of pause is deliberate and brought to you courtesy of Jim Hunt (used without permission) – thanks Jim

      • I used to say in comments at CargoCult Etc. that the “pause is going paws up.” It’s a reference to common death pose of animals that is often caused by the build up of gas in a decaying carcass. I think I paws in this context before anybody else.

    • I wondered about that too. Another factor that may be acting as a positive feedback is CO2 release from the oceans:

      • Syd, I don’t think that explains a faster growth during El Nino years and slower during La Ninas. During El Nino years the atmospheric temperature goes up, but ocean heat content actually is reduced. Moreover, it would not explain the often-observed lag.

      • Marco, good point, but perhaps SSTs don’t reflect OHC changes in this context. The heat is presumably going to the atmosphere via warmer surface waters (on average), and I’d presume further that the CO2 exchanges (if I may be allowed such a shorthand) are driven by surface temps.

      • Not so much release, if you look at the CO2 balance equation you’ll see that about half the annual emissions are removed, natural systems, the biosphere and the ocean are nett sinks for CO2. The atmosphere contains more than the Henry’s Law equilibrium value, if the ocean surface temperature increases then less CO2 will be absorbed and vice versa. Consequently you’d expect that variation in temperature will modulate the atmospheric pCO2 but the PCO2 will continue to increase overall due to the continued anthropogenic production of CO2.

  2. Didn’t 2015 beat 2016?
    2000 1.61 0.11
    2001 1.61 0.11
    2002 2.50 0.11
    2003 2.27 0.11
    2004 1.60 0.11
    2005 2.54 0.11
    2006 1.68 0.11
    2007 2.27 0.11
    2008 1.57 0.11
    2009 2.02 0.11
    2010 2.32 0.11
    2011 1.92 0.11
    2012 2.60 0.11
    2013 2.06 0.11
    2014 2.15 0.11
    2015 3.03 0.11
    2016 2.77 0.11

    [Response: The way they compute those annual increments isn’t the same as the difference between annual averages. If you difference their own annual averages starting with their data here, you’ll find that 2016 is the largest increase.]

  3. Here’s another record. We’ve had five upward years in a row. that’s the longest run (upward or downward) in the record. The previous longest run was 4 upward years, staring in 1999. There have never been downward runs longer than three years. This is using the NASA data.

  4. Keeling and Revelle show that <atmospheric carbon dioxide continues to increase during El Nino, at a faster rate than at other times. and looks specifically at the then recent El Nino of 1982/83. Changes in the carbon-13/carbon-12 ratio suggest that during the 1982-83 El Nino the deficiency in the flux of carbon dioxide in equatorial waters was more than made up by contributions from the land biota, caused by widespread forest fires and plant deaths due to drought. “”

    This was published in 1985 so it would be interesting to see whether more recent work corroborates the findings, but my understanding is that the explanation is still widely accepted.

  5. CO2 and heat numbers all look very ugly. I think they will “look” better when we hit the trough that occurs with a large EN like 98-99 or the one we have just been in.

  6. Thanks,

    “The way they compute those annual increments isn’t the same as the difference between annual averages.”
    I should have remembered. I explained this same thing to someone last year ;) Anyway, thanks for clearing this up.

  7. Everett F Sargent

    [Response: The way they compute those annual increments isn’t the same as the difference between annual averages. If you difference their own annual averages starting with their data here, you’ll find that 2016 is the largest increase.]

    Actually the 2016 data is “provisional” AFAIK. 2016 may indeed be 1st but it will take the better part of this year before a 1st set of 2016 numbers are “blessed” by the CO2 Trends website.

    Why do I say 1st? Because previous years have also been ever so slightly modified/adjusted depending either in changes in the CO2 network itself or to ongoing improvements in their algorithms.

    We also know that SST plays a role in CO2 values and rates thereof. So we will always need decadal trends, or some such, to better adjudicate long term trend/rate evolution.

    I fully trust the CO2 Trends website and their data products, whether provisional or final.

  8. Everett F Sargent

    From your own FTP link, and I quote …

    #’NOTE: In general, the data presented for the last year are subject to change,
    # depending on recalibration of the reference gas mixtures used, and other quality
    # control procedures. Occasionally, earlier years may also be changed for the same
    # reasons. Usually these changes are minor.

    For example, SPO usually doesn’t get posted until almost a year later (at either SIO or NOAA’s CO2 Trends).

  9. Everett F Sargent

    Finally, for the sake of clarity I only depend on their GLOBAL estimates of CO2 trends/rates/decadal trends (e. g. the full reported global network).

    Of course, MLO and SPO are still extremely important in and of themselves, given the historical record length of both (1957/8 to the present and ongoing). Both are excellent proxy records for the global trends.

  10. I was going to ask if Antarctic ice also had a record low year, but then I realized I could check myself.
    On the other hand, looking at global sea ice as an example, I started to wonder about variability. One aspect of adaptation (societal and biological) is predictability. With “global weirding,” predictability may be going down while variability goes up. A directional trend may be less damaging than an increase in variance (if the trend is slow enough). So as we see Antarctic sea ice swing wildly, or drought severity increase while deluges become more frequent, I wonder if 2016 might have seen record “variability” as well as record extremes in the other things you listed.

  11. What disturbs me most is that approximately _half_ of all the increase in carbon dioxide has occurred since the early 1990s, when the issue was recognized as one of planetary importance.

  12. As former DOE Secretary and Laureate Stephen Chu recently pointed out, the actual atmospheric burden of CO2 is more like 490 ppm because there are other atmospheric species, like CH4, which will either convert to CO2, or have radiative forcing which are many times equivalent of CO2.

    It’s very non-linear, and, so, the s___ hits the fan.

    • But the negative forcing from man-made aerosols is around -90 ppm CO2-eq, so it (conveniently) pretty much cancels out with non-CO2 gases currently, and CO2 alone provides a reasonable proxy for total change in radiative forcing.

      • Our species should focus on CO2. It is the big dog and the solutions for CO2 will largely clean up most of the other ghg’s. To the extent that we would have to come up with separate solutions, well, ok, let’s do that after we have cracked the CO2 nut. The CO2 accumulation is going to kill us and then the other ghg gas issues are going to electrocute us after we are dead. Probably should solve these threats in the correct order.

        There is nothing wrong with working on both, but as I have said, I think the CO2 solution resolves most of the ghg threats. There really is no indication that our species can address the CO2 issue, so the other ghg are just a distraction imho.

        That would be different if we could actually grab these things with the paws of unicorns, but so far, it looks like it requires human hands. Human hands have so far proven to be as useful as unicorn paws (aren’t they really hooves?).


        Daily CO2

        January 10, 2017: 405.95 ppm
        January 10, 2016: 401.96 ppm

        Warm regards


      • Okay, but that accounting approximation ought to be accompanied by a large footnote explaining it. Concentration and type of aerosols can vary, whether upwards by natural things like a Hurst burst of Pinatubo-scale eruptions, or a rollout of an ill-advised SRM scheme, or downwards by curtailment of tropical forest burning or conversion of coal to more cost effective means of generating electricity. And, after all, radiative forcing is not the only quantity of importance and interest. Aerosols do nothing to impede carbonic acid build-up, and should the day arise when clear air capture is pursued, actual concentrations will determine costs , a bit of the efficiency, and how long capture needs to run to achieve a set level.

  13. Hello, I’m interested in knowing the Northern Hemisphere temperature over land areas only. Does anybody have knowledge of that?

  14. In terms of Arctic SIE, 2016 was a record year big time. Over the year two-hundred-and-one days of 2016 saw SIE lower than ever before. But before we have had time to take stock [and change the title on this graph of JAXA Arctic SIE anomalies (usually 2 clicks to ‘download your attachment’)], 2017 has started where 2016 left off. By January 10th, Arctic SIE was sitting 631,000 sq km below the average for 1993-2015. Such anomalies have occurred before but except in recent years (2015 for a bit in March, 2016 from April) an anomaly as low as -630,000 sq km doesn’t appear normally until the end of June when the melt season is in full swing.
    And when I say “normally” that is 2007 and 2012 normal. So hold on to your hats for 2017!!!

  15. @Stercus Feri

    NASA GISTEMP has data for each hemisphere and for land only:

    [Response: Be careful — the NASA land-stations index isn’t for land areas, it’s an estimate for the whole globe *based on* land stations only. Both NOAA and HadCRU do, however, provide global and hemispheric data for land areas only.]

  16. A link to plateauing of emissions is
    Would be interesting, how the discrepancy between those emission data and the positive second derivative of the concentration data can be explained. Emission data not reliable?

    • The emissions report is selective – it’s about emissions from fossil fuels and industry and it’s likely based on self-reporting that will gloss over and minimize stuff like methane leakage from drilling and production. This is the reason that I only follow actual measurement of CO2 and ghg in the atmosphere. In the final analysis, if we cook ourselves with ghg accumulation on the atmosphere and ocean, it really won’t matter if our emission reports show that we are making good progress.

      Last week CO2 numbers from

      Last Week

      January 1 – 7, 2017 405.91 ppm
      January 1 – 7, 2016 402.03 ppm
      3.88 ppm increase in year on year weekly average comparison. Do you see anything in the hard numbers of atmospheric accumulation to support the idea of a meaningful plateauing of ghg accumulation in the atmosphere and oceans? I do not. We are going the other direction and gaining speed. Nature bats last.


    • kinimod,
      the total CO2 emissions are not plateauing. While the increase emission from fossil fuels was smaller in the last years, the land use emissions increase again, and the land sink decreases. According to the data from here

      The data up to 2015 are from the Global_Carbon_Budget_2016_v1.0.xlsx table from the link, the 2016 FF data are estimates from
      C. Le Quéré et al. Global Carbon Budget 2016.
      The LU from 2016 I imply assumed the same increase as in 2015.

      The unit is GtC/year, (for conversion to CO2 multiply by 3.664).
      I hope the format of the table remains, after posting.

      FF fossil fuel + cement emissions
      LU land uses change emissions
      TE total emissions, FF+LU
      AT atmospheric growth
      OS ocean sink
      LS land sink

      Year FF LU TE AG OS LS
      2000 6.79 1.20 7.99 2.63 2.10 3.26
      2001 6.97 0.96 7.94 3.90 2.00 2.04
      2002 7.07 1.07 8.14 5.05 2.26 0.83
      2003 7.47 0.90 8.37 4.83 2.38 1.16
      2004 7.85 1.04 8.89 3.31 2.27 3.31
      2005 8.23 1.03 9.26 5.15 2.29 1.81
      2006 8.53 1.09 9.62 3.75 2.45 3.41
      2007 8.78 0.97 9.74 4.43 2.49 2.82
      2008 8.96 0.67 9.63 3.77 2.47 3.38
      2009 8.87 0.75 9.63 3.43 2.52 3.68
      2010 9.21 0.88 10.09 5.17 2.49 2.43
      2011 9.54 0.91 10.46 3.58 2.65 4.23
      2012 9.69 0.97 10.66 4.98 2.74 2.93
      2013 9.82 0.92 10.74 5.24 2.75 2.75
      2014 9.89 1.10 10.99 4.22 2.88 3.88
      2015 9.90 1.32 11.21 6.30 3.03 1.88
      2016 9.95 (1.54 11.49)

    • “Would be interesting, how the discrepancy between those emission data and the positive second derivative of the concentration data can be explained. Emission data not reliable?”

      Lag time?

      • No, incorrect expectations. See my comment above.

      • I think the lower emission data are based on some hard numbers, like sales, export/import of fossil fuels, national conversion of energy systems from coal to natural gas and other less ghg heavy fuels, so I think the emission data is somewhat accurate, it’s just misleading because it glosses over the predicament we are in with ghg accumulation.

        I think carbon cycle is jammed up now by the shocking amount of ghg that our species has slammed into the atmosphere. The science is starting to catch up and support that with studies of changes in natural carbon sinks to carbon sources due to warming and related changes. I think it’s a fool’s errand to spend much time on this. From a practical pov, we now have to change our impact to make our species a carbon sink instead of a carbon source. That really is the bottom line. The hard number to follow is CO2 in the atmosphere. There it is, no explanation or pr gloss can change that number and what it means.

        Last Week
        January 1 – 7, 2017 405.91 ppm
        January 1 – 7, 2016 402.03 ppm
        3.88 ppm increase in a noisy number.

        Dr. Mann said in 2014 that we should stay under 405. I think he was right. To stay under 405, or now, to get back below 405 would require massive changes in the way we live. That is what we have to do. I don’t see much sign that we are willing to do that.

        Warm regards,


      • As I indicated above, there are a lot of processes who “have their hands in the game.” For instance, there’s the respiration of the northern hemisphere forests.which create the sawtooth on the Keeling. So, between that and mixing, I would not expect atmospheric concentrations to track year over year changes in emissions. There’s the whole oceanic and soils equilibration process going on, too. Any attempt to infer a connection or absence of one by seeking evidence of such perturbations suffers from lack of statistical power.

      • agree with what you say here hypergeometric

  17. @kinimod, I would not expect much of the small scale signal in human emissions to survive in concentration figures. It’s being heavily mixed and smoothed by other processes, including trying to equilibrate with oceans and soils, variability in northern forest growth due to varying weather patterns (those causing the sawtooth pattern), and mixing with decaying methane and emissions from concrete and burning. Recall the NASA animation from OCO-2:

    You probably know this, but just to remind the wider readership … CO2 concentrations will continue to go up even if emissions are flat (zero second derivative) or declining (negative first derivative). The only way they won’t go up is if emissions are zero. That’s because there’s nothing removing CO2 at any rate comparable to introduction. Sure, a bunch is going into oceans and soils, but that isn’t sequestered. That’s what Phil wrote about above. So we’re filling a big bathtub, and the drain is badly clogged. Henry’s Law doesn’t really apply because the system’s not in equilibrium.

  18. Larry Edwards

    Regarding the “NSIDC N. Hem” and “NSIDC Global” charts, I wish to put forward some thoughts for discussion:

    1. While the red fit lines are interesting and troubling, the occasional major valleys are of great popular interest and as well they set a base condition from which a rebound to the next peak must proceed. In addition, positive feedbacks caused by these deep minimums and/or other climate system processes lead to a downward trend throughout all elements of the chart (the red trend line, the series of major valleys, and (relative to the trend line, and (except for 2013 and 2014 in the global chart) the series of major peaks.

    2. Feedbacks and other factors manifest themselves differently in creating the major valleys and the major peaks. I therefore think it would be interesting to analyze them separately, say fitting a blue trend line for each series. E.g., in global chart (the more dramatic one) I would note 8 major valleys and 8 or 9 major peaks.

    3. If Tamino were to do this analysis, in the N. hemisphere chart I suspect the right end of the Major Vallies Trend Line (MVTL) would end up just below the end of the red line. However, if 2017 is another low record year (as some speculate above), that could be quite interesting. But in the Global chart, the this kind of analysis would already have a dramatic MVTL, and a screwy M-Peak-TL as well.

    4. It seems that the trends in the extremes (especially the trend for the low ones) may be more important than the red trend line. Such an analysis could lead to new insights. Does this approach make sense to you, and is it worth a look?