Climate sensitivity and decadal temperature variability

There are some who argue that natural/internal variability can play a role in driving long-term warming, and – hence – could explain a substantial fraction of recent warming. This, however, creates a bit of a paradox; if the system responds strongly to internally-driven warming, then it should also respond strongly to externally-driven warming. Consequently, we’d expect climate sensitivity to be high which would then make it difficult for a large part of our recent warming to be due to natural/internal variability.

Credit: Figure 1b from Nijsse et al. (2019).

The reason I thought I’d mention this again is that I came across a recent paper by Femke Nijsse and colleagues that consider this issues. The paper is called Decadal global temperature variability increases strongly with climate sensitivity and the title pretty much gives away the punch-line. The paper shows that models that are more sensitive to GHGs emissions (that is, higher equilibrium climate sensitivity (ECS)) also have higher temperature variability on timescales of several years to several decades, which is illustrated in the figure on the right.

The paper also points out that

high-sensitivity climates, as well as having a higher chance of rapid decadal warming, are also more likely to have had historical ‘hiatus’ periods than lower-sensitivity climates.

and consequently, that

the slowdown in global warming during the period 2002–2012 was more likely in a high-ECS world.

So, rather than the supposed slowdown being an indication of a low climate sensitivity, it could well imply the opposite.

The paper then concludes that

[a]chieving a better consensus on the risk that we live in a high-ECS climate is therefore of critical importance to both the climate mitigation challenge and also to inform efforts to build resilience to climate variability.

This entry was posted in Climate sensitivity, Research, The scientific method and tagged , , , , , . Bookmark the permalink.

80 Responses to Climate sensitivity and decadal temperature variability

  1. high or low sensitivity won’t matter if we just continue to accumulate CO2 in atmosphere and oceans. I see no sign that we are turning this around, yet. All the science and spin in the world is a big zero in import unless we stop the rise of gh gases. How are we doing on that?

    June CO2

    June 2019: 413.92 ppm
    June 2018: 410.79 ppm

    Could be worse, right?



  2. but, to be clear, higher sensitivity does not sound like a good thing. I need to check in with Frank Luntz and figure out how to discuss this most effectively.

  3. “models that are more sensitive to GHGs emissions (that is, higher equilibrium climate sensitivity (ECS)) also have higher temperature variability”

    Some models having higher variability than others doesn’t tell you if any of them are correct.

  4. TE,
    That isn’t really the claim, though. The point is simply that higher variability correlates with higher sensitivity.

  5. dpy6629 says:

    If you look at Real Climate’s page for model comparisons with observations, you will see that from 1955 to present the highest decadal temperature increase is about .35-0.4 depending which observational record you use. That has happened twice in about 65 years. Looking at Figure 4 from the paper you post about here, at ECS = 3.5C the probability of a 0.7 C decadal increase is 5%. That increases to 17% at the higher end. There are 55 instances of decades from 1955 to present. I know the trends are not independent but one might think that at 17% probability we would have seen at least one.

    Referring to the pause in this regard as possible evidence for high ECS seems a stretch.

  6. dpy,
    I think you need to consider the upper panel, which is for the historical period. The lower panel is if we were to continue along an RCP8.5 pathway. The lower panel suggests that if the ECS is 3.5K then there is a 5% chance that we’d have experienced a decadal trend of ~0.4K/decade, which does indeed seem to have occurred.

  7. verytallguy says:

    It’s a long standing contradiction of “sceptics”:

    Firstly, the Medieval Warm Period was large and global; the evil Hockey Team airbrushed it out.

    Secondly, climate sensitivity is low.

    These things are flat out incompatible.

  8. JCH says:

    On the other hand, there was no pause in warming could also be congruent with low climate sensitivity. Selling both low climate sensitivity and a warming hiatus and it’s all El Niño warming is a truth hating political hack.

  9. TE None of them are correct, they are models – ask GEP Box.

  10. To detect climate change figure 1 is where to start. Climate is average weather over a timespan of 30 years and every 10 years you start a new timespan. You use all the standard statistic for this kind of normal distribution. And then you use standard technics like Student T to find out if and how differences are significant. It is one of the standard ways to detect climate change. If you have time or data you might want to the check all weather related stuff like rain, wind etc.

  11. Willard says:

    > the title pretty much gives away the punch-line.

    You say that like it’s a bad thing, AT.

  12. I seem to remember having a discussion* with climatologists at UEA about this some time ago, but it is great to have a paper with some experiments etc.

    *In the sense that I asked something and then did some listening; if only I could have done more understanding and remembering… ;o)

  13. DKM : “but it is great to have a paper with some experiments etc.”

    The nice thing about model intercomparisons is that they avoid the moral hazard of experiments involving things, which may put narrative arcs at risk by leading to material conclusions.

  14. I have to admit, it is easier than paying Magrathea to produce a few additional Earths to play with, which would be the next step up from using models.

  15. But then again, even models of models can be useful.

  16. izen says:

    By the time that ‘The Paws’ was a thing, the rate of energy accumulation in Joules by the climate system was estimated with some degree of precision.
    One thing overlooked in the climateball confusion of trying to frame the paws as a refutation of AGW was that if internal variation could divorce the rate of surface temperature rise from the rate of Joules accumulated, then it is reasonable to conclude that an equal and opposite internal variation could double the observed rate of warming.

  17. frankclimate says:

    vtg: and the other way around is also valid: Adjusting the GMST for ENSO, solar and volcano makes a very small variability., see . No pause in warming AND high sesitivity… it won’t work together.

  18. frank,
    Well, if you take out some of the variability, you can’t then argue that it’s not variable. I somewhat agree, though, that if we consider the full record, there isn’t much evidence for any real pause in surface warming, so one should be slightly careful of then using this to argue for high climate sensitivity.

  19. The thing to remember about regression-based procedures is they tell you what can be explained by a set of external factors, but not what actually is explained by those factors. For instance there is missing variable bias. Any response of the system to a missing variable that happens to be correlated with a variable that is included will be assigned to that included variable. This can also happen with the noise in the signal. It is risky therefore to use such arguments to rule something out, based on an analysis that is biased to some degree against it a-priori.

  20. frankclimate says:

    ATTP: “Well, if you take out some of the variability, you can’t then argue that it’s not variable. ”
    The ENSO adjusting of the GMST is not the “take out of some variability” in the sense of the paper as they stated:
    “By removing the ENSO signal, on the basis of this index, it is shown that ENSO is not the dominant factor in our relationship.” ( see p.600).

  21. Frank,
    Ahh, yes, I’d forgotten that.

  22. verytallguy says:

    No pause in warming AND high sesitivity… it won’t work together.

    Not sure the point you’re making Frank, care to expand?

  23. David B. Benson says:

    The fires in Siberia have so far burnt 27 million hectares of tiaga and tundra. The smoke has made going outside almost impossible.

    Nonetheless, does this offer an opportunity for climate models to better estimate climate sensitivity?

  24. “The fires in Siberia have so far burnt 27 million hectares of tiaga and tundra. The smoke has made going outside almost impossible.”

    I have a different question about the fires, the permafrost thaws, etc. My question is about scale and whether these events are consequential in terms of CO2 and CO2e levels in the atmosphere? Or are they merely early instances of feedbacks waking up that may take decades or centuries to become consequential in terms of ghg accumulation?

    I would love to see discussion of these events scaled in ways that laypersons could understand. 27 million hectares sounds like a big number, but we need scaling. How many hectares have burned historically? Is the ghg output on the level of the global cruise ship industry? Or similar to the annual output of Lichtenstein?



  25. frankclimate says:

    vtg: When adjusting the observed GMST (C&W) for ENSO ( and solar and volcano forcing) as it did “tamino” in the cited post ( this operation is conform to the methods of the Nijsse et al paper as I showed it above), leaving antropog. forcing and IV allone, one gets the sigma of all decadal trends ( from 1976..1985… to 2007…2017) of only 0.056. See the figure 2b of the paper if it means a low sensitivity or a high one. With the adjustment for ENSO the sigma of the decadal trends goes down to 1/2 in relation to the unadj. observed GMST and the “pause” after 2000 vanishes. No observed pause is not an argument for a high ECS, following the paper.

  26. verytallguy says:

    No observed pause is not an argument for a high ECS, following the paper.

    Seems reasonable. Has anyone suggested no observed pause is an argument for high ECS? I’m not aware of such.

    [might help if you defined what you mean by “high” here]

  27. vtg,
    I think the point is simply that the paper is suggesting that a world with a high ECS is more likely to have sufficient variability for there to be a “pause”. That doesn’t mean, though, that a “pause” indicates high sensitivity, or that the lack of a “pause” suggests that it isn’t high.

  28. frankclimate says:

    ATTP: from the paper (last paragraph) :”Counter-intuitively, this suggests that the slowdown in global warming during the period 2002–2012 was more likely in a high-ECS world.”
    As tamino ( and others too) showed, there was no pause at all, it was more likely a product of ENSO. When admitting this one could turn this sentence into: “Counter-intuitively, this suggests that the lack of a slowdown in global warming during the oeriod 2002-2012 was more likely in a l o w ECS-world”?

  29. Frank,
    Firstly, there is still debate about whether or not there was a slowdown. I don’t think there is much evidence for anything unusual, but it’s certainly possible that it warmed more slowly than might have been expected. I also don’t think you can invert the argument in the way that you have.

  30. Joshua says:

    > or that the lack of a “pause” suggests that it isn’t high.

    JCH referenced this above…

    So JCH says “could be congruent with” rather than “necessarily means” but I’m curious as to what explains that difference.

  31. verytallguy says:


    I think you’re rather confused here. Your argument seems to be that ENSO should be disregarded when considering the potential relationship between internal variability and climate sensitivity.

    But ENSO is part of internal variability; I would suggest for instance that models where ENSO has a bigger impact on global temperatures will also be more likely to have high climate sensitivity.

    But I do agree with the point that the higher climate sensitivity is, the more likely there is for there to be a “pause” (however defined) in warming – which is I think your starting point?

  32. Joshua says:

    re: my 4:49 – if it is possible to explain that in a relatively non-technical manner 🙂

  33. sounds like one of those problems where P(A|B) is being conflated with P(B|A)?

  34. Joshua,
    I think the point is simply that if climate sensitivity is indeed very high, then we would expect periods of rapid warming, and periods of much slower warming. Hence, if we were to conclude that there wasn’t actually a pause/slowdown, then that might start to indicate that the ECS isn’t high. Of course, here we would probably regard high as > 3C, so it’s not suggesting that the lack of a pause somehow validates the arguments of Lukewarmers. I would add, though, that I suspect the record isn’t really long enough to draw any strong conclusions and, since this variability is somewhat random, even a high sensitivity world could have an extended period that didn’t show much variability around the long-term trend.

  35. frankclimate says:

    ATTP: of course it’s a little bit nitpicking. However in the context of the paper the authors should have been accepting this suggestion of Foster et al (2015)
    “But we suggest that scientists should stop speaking of a “slowdown” in temperature increase as though it were a known fact, when it simply isn’t .

  36. The pulse of 25% of total global emissions in a twenty year span coincided with the real or nonexistent pause, depending on your mood of the moment. It is the two coincident factors–pulse, lack of reaction–that raised my eyebrows when thinking of sensitivity.

  37. verytallguy says:

    Frank, no idea what point you’re trying to make. You may have to expand upon the point to make it explicable.

  38. Joshua says:

    Frank –

    > No observed pause is not an argument for a high ECS, following the paper.

    Did someone argue that it is?

  39. Joshua says:

    Anders –

    >I would add, though, that I suspect the record isn’t really long enough to draw any strong conclusions and, since this variability is somewhat random, even a high sensitivity world could have an extended period that didn’t show much variability around the long-term trend.

    I will readily admit to no technical capacity here, but it still drives me nuts when people consistently try to justify their views based on short term trends or by truncating longer-term trends.

  40. Joshua says:

    Frank –


    > No observed pause is not an argument for a high ECS, following the paper.

    What is a “high ECS?”

  41. frankclimate says:

    Joshua, I would suggest you to read the paper. All your questions are answerd after this, I hope.

  42. verytallguy says:


    Pulse of emissions??

    I see no pulse

  43. Joshua says:


    I wouldn’t likely understand the paper. Trying to read it would likely be a waste of time. Any particular reason you wouldn’t lend me your understanding to answer my question? Seems like a fairly reasonable request to me.

    Plus, I was asking you a question with respect to your wording (and interpretation). As such, reading the paper probably wouldn’t be very useful even if I could understand it.

  44. Willard says:

    Why read the paper:

    Anyone who has a blender can dig the point.

  45. “Any particular reason you wouldn’t lend me your understanding to answer my question?”

    welcome to my world ;o)

  46. David & SBM;

    If burning 270,000 hectares of Siberia , including peatlands, constitutes a consequential climate forcing, what about 14,000,000?

    That’s the low end of the estimates of total area burned in 1915 . The high end estimates range up to 100,000,0000 hectares.- I’ve linked my Nature paper describing the 1915 burn in this post:

    A caveat though- while that unusually hot and dry summer can be attributed to natural variability the wildfires extent was arguably amplified by an anthropogenic wildcard: thousands of settlers fanning out from the newly completed Trans-Siberian Railroad to clear homesteads in the wilderness, a transmigration that intoduced an unprecedented number of fire sources.

  47. JCH says:

    Well, soon, right around the corner, the Stadium Wave will clobber the nasty alarmists with a truly big DaPaws and that will prove ECS is sky high. Just a matter of time.

  48. No pulse.

    If you add up the numbers in the IEA charts by region-
    you find that if all of North America, South America, and Europe went to zero emissions tomorrow, then global emissions would be 4% less than 1990 global emissions.
    My understanding was that in 1990 we believed that level of emissions would fry the planet.
    Global emissions were 19.89 gton in 1990, 31.08 in 2016.
    The Americas and Europe accounted for 13.24 gton in 1990, 12.05 gton in 2016.

  49. JCH says:

    that raised my eyebrows when thinking of sensitivity.

    Me too. But you went in the wrong direction.

  50. JCH says:

    One often hears about sea ice in the Arctic. Note how it is reported. Does the information refer to sea ice extent across the entire Arctic or for specific regions of the Arctic? In the stadium-wave hypothesis, only one region of the Arctic is critical. This is the Eurasian Arctic Shelf Sea region. According to the stadium wave’s hypothesized hemispheric signal propagation, anomaly trends in Eurasian Arctic sea ice extent begin in the seas furthest to the west (Greenland, Barents, and Kara) and culminate in those to the east (e.g. Laptev and East Siberian seas).

    The message here is to keep in mind the difference between reports on anomalies and reports on anomaly trends! And, pay attention to what specific region those reported anomalies and trends involve. …

  51. izen says:

    Is there a common confusion involved here between Energy (in Joules) – Heat – and surface Temperature?

    The rising CO2 levels ineluctably cause an increase in the amount of Energy that accumulates in the climate system of Oceans, land, and atmosphere. How that changes the surface temperature is mediated by the way the Joules are distributed between oceans, land, and atmosphere.

    IIRC during the ‘Paws’ the OHC and sea level rise did not show a significant change in the rate of energy accumulation. The thermal capacity of oceans means that they can absorb a few percent extra Joules with little impact on the surface temperature. The much lower thermal capacity of the 2m above the land surface where much of the Temperature record of primary social interest is measured has a much lower effective thermal capacity so can show much greater variability in response to any variation into how the Joules are partitioned between the different Energy sinks.

    ECS at its simplest is a measure of the thermal capacity of the whole system, or how many Joules will raise the temperature by 1C.
    TCR is a measure of how ENSO and other internal variations in addition to the significant difference in the thermal capacity of oceans – land/air can alter the rate of temperature gain as the Joules are shuffled between the different energy sinks.

  52. JCH says:

    In terms of the earth system, the surface is tiny thing, and massive amounts of energy are blowing through it in both directions all of the time. Somewhat amazing that it is habitable. So when there is a DaPaws, and sorry, there was one, just not one that lasted long enough to become an actual one, something unusual has happened. In the first half dozen or so years of the 21st century something unusual did happen, and it points to ECS being higher than some expect.

    The first sign post: Trenberth’s missing heat. Where did he say it went? He said it either went into the oceans – somehow not showing up in surface or OHC measurements – or it was reflected back to space before it could be absorbed by the earth system; therefore, never showing up in the surface measurement and never having gone into the oceans in the first place.

    Final answer is both, but the one that matters is a whole bunch of it was harmlessly and unexpectedly reflected back to space. And then, just as suddenly, it stopped being reflected back to space.

    It’s unusual. It did happen. Will it ever happen again?

    New paper: Distinct patterns of cloud changes associated with decadal variability and their contribution to observed cloud cover trends

  53. Joshua says:

    > So when there is a DaPaws, and sorry, there was one,

    How do you define DaPaws?

  54. anoilman says:

    Joshua: DaPaws was whatever length of time it took to include the El Nino Temperature Spike of 1998 which was huge. Hence… from whenever someone realized that including 1998’s temperature would make it look like a Paws (2007?) it was included. Thereafter DaPaws length was steadily increased annually so as to include 1998.

    The only way for DaPaws to cease would be for there to be another massive (ENSO Driven) temperature spike or for the actual temperature signal to rise above the noise.

    I can only assume that since we’re talking about DaPaws in the past tense, that temperatures have either spiked, or are now drowning out the signal.

    Sorry for being pedantic, but you know… its important to be as scientifically rigorous as possible when talking about DaPaws.

    I suppose I could wake from my lethargy and look the temperatures up. I mean which is it! Could it be the ever increasing signal drowning out the noise? Or is it the dreaded El Nino uncertainty monster again?

    Perhaps we should refer to this point in time as the AntiPaws?

  55. izen says:

    “but the one that matters is a whole bunch of it was harmlessly and unexpectedly reflected back to space. And then, just as suddenly, it stopped being reflected back to space.”

    Would this not show up in the CERES satellite measurements ?
    AFAIK that data is dominated by a decrease in reflected solar radiation because of reduced ice cover in polar regions.

  56. anoilman says:

    According to Berkeley Earth, it looks like a bit of El Nino and increasing temperatures.

    But hey check out 1998!

  57. Joshua says:

    What a funny coincidence that DaPaws began in 1998.

  58. izen says:

    There seem to be pawzes from 1881 – 1934, again from 1945 – 1971, then 1981 – 1997 as well.
    Inter-spaced with jumps of ~0.3C in less than a decade.
    Are the pawzes getting shorter…?

  59. JCH says:

    DaPaws began in 2005, which was quite a bit warmer than 1998.

    Izen, one big difference is, Almost DaPaws has a lot of data, and lots of explanations, and more are coming.

  60. Joshua says:

    JCH –

    > DaPaws began in 2005, which was quite a bit warmer than 1998.

    Why do you say that? Seems to me that usually the starting date is 1998. For example:

  61. anoilman says:

    JCH: You forget that the data set selected always had to show 1998 as higher than whatever point they were trying to make. [For obvious reasons the data set changed annually as well.]

    That’s like a rule or something. You can’t mess with that. However, JCH, using tip jar statistics is clearly a violation of the rules unless someone in the wilds of the internet blogged it to be valid.

    Can anyone tell me if satellite data is still showing a down trend? That was all the rage in some circles a while back.

  62. Ben McMillan says:

    RSS satellite data trends are similar to ground-based data, UAH satellite data (UAH6) shows an upwards trend since 1998 about half that much.

    From the perspective of 2019, it is indeed hard to see DaPaws. Now you have to cherry-pick an end date as well as a start date. It really confirms the point made at the time that if you want to talk about changes in trend, you should actually look for statistical significance (ie changepoint analysis).

  63. Dave_Geologist says:

    thousands of settlers fanning out from the newly completed Trans-Siberian Railroad to clear homesteads in the wilderness, a transmigration that introduced an unprecedented number of fire sources.

    Fireman: What are the three main causes of fires?

    Audience: Smoking, electrical faults and cooking.

    Fireman: No. Men, women and children.

  64. JCH says:

    The reason climate scientists are continuing, and it’s in pretty large numbers, to write papers trying to explain DaPaws is because something unusual happened. You can deny it it until the cows come home, they are going to continue writing papers about it. Why? Because it, studying DaPaws, may lead to a far better understanding of how the earth system works. One indication, studying DaPaws already has led to far better understanding of how the earth system works. Or, you can just stay with what was not known when the DaPaws started.

  65. JCH says:

    Joshua – from 1998 to the end of 2005, the rate of warming was essentially .019 ℃ per year, .2 ℃ “decadalized”. That is robust warming, and it makes no sense whatsoever to come along later and suggest it never happened. It’s akin to claiming summer never happened because the temp in September was just as low as the temp in the prior April; therefore, the six straight weeks of temps over 100 degrees in Dallas (summer) never happened..

    If there had been no warming after 1998, all years thru 2012 essentially the same as 1998, then you could correctly argue there was no warming since 1998. You could also legitimately argue climate sensitivity is pretty low. Lots of new ACO2; no action.

    But that is not what happened. There was a lot of warming until 2006, and then there were some very powerful La Niña events and a lot of cooling. This suggests the system is very sensitive to radiative changes. Because, starting around 2006, something unusual started happening. For instance, like this:

  66. lerpo says:

    Any explanation for a pause between 2005 and 2014 probably needs to explain why it was so durned warm in 2005 as much as it needs to explain why it was so durned cool in 2014.

  67. JCH says:

    Changes in Earth’s Energy Budget during and after the “Pause” in Global Warming: An Observational Perspective

    Net downward TOA flux anomalies after 2014 are generally positive due to the large decrease in reflected SW, which overwhelms the increase in outgoing LW radiation. Global mean SW TOA flux decreases by 0.83 ± 0.41 Wm−2 for July 2014–June 2017 (post-hiatus period) relative to July 2000–June 2014 (hiatus period), whereas for outgoing LW and net downward radiation the differences are 0.47 ± 0.33 Wm−2 and 0.39 ± 0.43 Wm−2, respectively. Following other El Niño events observed during the CERES record, increases in outgoing LW flux generally dominate over decreases in reflected SW flux, resulting in negative net TOA flux anomalies. TOA flux anomalies during the post-hiatus are thus highly unusual. To test the robustness of the results, we compared TOA flux anomalies from EBAF Ed4.0 with those produced separately for CERES instruments aboard Terra, Aqua and S-NPP. Anomalies from the different records were very close to one another with monthly RMS differences <0.2 Wm−2 and no apparent instrument drifts.
    The decrease in global mean all-sky SW TOA flux between the post-hiatus and hiatus periods is primarily associated with areas over the eastern Pacific Ocean off North and South America, as well as over the west tropical Pacific and the Southern Pacific Convergence Zone. A partial radiative perturbation analysis reveals that decreases in low cloud cover are the primary driver of the SW TOA flux decreases. Furthermore, the regional distribution of decreases in SW TOA flux associated with low cloud cover changes closely matches that of SST warming, which in turn shows a pattern typical of the positive phase of the PDO over the eastern Pacific. In contrast to the decreases in SW TOA flux over the Pacific, increases occur over the north Atlantic associated with the North Atlantic Cold Blob, which partly compensates for the SW TOA flux decreases over the Pacific. …

    Almost DaPaws

  68. izen says:


    I do not have enough familiarity with the issues and factors in this field to make an informed judgement, perhaps other can.
    But this study you cite does not seem to agree with the research I linked to earlier on the variations found in the CERES data of reflected solar radiation and outgoing long wave radiation.
    That does find a instrumental drift affecting the data.
    And while it does describe the same shift in the ratio of reflected solar to outgoing LW over the same time periods, (see fig7) it attributes them to changes in see ice cover altering the albedo at the poles with little apparent affect from pacific cloud cover. (fig 4)

    The impact of changing ice cover up to 2012 is also found in –

    Neither mentions significant effects from cloud cover and ENSO changes.
    Is anyone sufficiently familiar with this aspect of the subject to disentangle the apparent contradictions?

    The changes in albedo from reduced polar ice cover are a well predicted effect from models, and presumably have an impact on TCR, while they do not directly alter climate sensitivity, they do alter the amount of Joules added to the system.

  69. “Is anyone sufficiently familiar with this aspect of the subject to disentangle the apparent contradictions?”


  70. JCH says:

    When it comes to the GMST, the eastern Pacific is a big dog. It cooled from around 1985 until around 2013-14.

    The La Niña tongue is a big dog. From 2006 to 2013, those two big dogs were biting the GMST on the butt.

    Impact of decadal cloud variations on the Earth’s energy budget

    From Hartmann and Ceppi 14:

    The spatial structure of the trend in annual-mean reflected shortwave in Fig. 3 strongly suggests a trend toward La Niña conditions over this period, as indicated by the decreasing reflection near the dateline and the increased reflection west of the dateline and from an analysis of sea surface temperature (not shown). The La Niña trend also accounts for the upward trend in zonal-mean SW between the equator and 10°N in Fig. 1, since as equatorial east Pacific SST cools during La Niña the ITCZ is more concentrated north of the equator there. A northward movement of the ITCZ would be consistent with the greater absorption of energy in the Northern Hemisphere accompanying the polar ice melt following the reasoning of Chiang and Bitz (2005) and Frierson and Hwang (2012), but the integrated effect of this increased Arctic energy intake seems too small to be important compared to the influence of ENSO during the period of the CERES data.

  71. Ben McMillan says:

    No doubt lots of interesting work done on internal variability sparked by the putative hiatus, but is there any statistical evidence that something ‘unusual’ happened in the temperature trend itself over some early-2000s period, taking into account recent data?

    I think that this question came up before and the answer was ‘no, but fluctuations are interesting anyway’, which I guess is fair enough, but odd when the hiatus is treated as worthy of special note. There are some tamino threads on it.

    I mean, something like the ‘anomaly from trend’ exceeding 2 sigma or changepoint analysis finding a break in the trend?

  72. John Hartz says:

    It’s definitely time to buckle up……

    One common metric used to investigate the effects of global warming is known as “equilibrium climate sensitivity”, defined as the full amount of global surface warming that will eventually occur in response to a doubling of atmospheric CO2 concentrations compared to pre-industrial times. It’s sometimes referred to as the holy grail of climate science because it helps quantify the specific risks posed to human society as the planet continues to warm.

    We know that CO2 concentrations have risen from pre-industrial levels of 280 parts per million (ppm) to approximately 410 ppm today, the highest recorded in at least three million years. Without major mitigation efforts, we are likely to reach 560 ppm by around 2060.

    When the IPCC’s fifth assessment report was published in 2013, it estimated that such a doubling of CO2 was likely to produce warming within the range of 1.5 to 4.5°C as the Earth reaches a new equilibrium. However, preliminary estimates calculated from the latest global climate models (being used in the current IPCC assessment, due out in 2021) are far higher than with the previous generation of models. Early reports are predicting that a doubling of CO2 may in fact produce between 2.8 and 5.8°C of warming. Incredibly, at least eight of the latest models produced by leading research centres in the United States, the United Kingdom, Canada and France are showing climate sensitivity of 5°C or warmer.

    The terrible truth of climate change, Commentary by Joëlle Gergis, The Monthly. Aug 2019

    Joëlle Gergis is an award-winning climate scientist and writer based at the Australian National University. She is the author of Sunburnt Country: The History and Future of Climate Change in Australia.

  73. JCH says:

    Recent strengthening of the tropical Pacific zonal SST gradient is a dynamically consistent response to rising greenhouse gases

    Why has the subsurface Equatorial Pacific cooled in a warming world?

    Attributing Historical and Future Evolution of Radiative Feedbacks to Regional Warming Patterns using a Green’s Function Approach: The Preeminence of the Western Pacific

    A key question is thus whether the western Pacific will continue to warm quickly relative to the rest of the world oceans, as we have seen in observations to date, or whether it will warm by relatively less in the future, as GCMs predict. These results suggest that only in the case that the western Pacific warming keeps warming at a greater place than the rest of the global oceans can we expect ICS to remain as low as that derived from recent energy budget constraints (e.g., Otto et al. 2013; Lewis and Curry 2015; 2018; Armour 2017; Knutti et al. 2017). If GCMs are accurate in their projection that the western Pacific warming will not keep pace with the eastern Pacific and high latitude warming, then we can expect a less negative feedback, and a higher value of ICS, in the future. …

  74. John Hartz says:

    An now this…

    Ocean heat waves, which can push out fish, plankton and other aquatic life, are happening far more frequently than previously thought, according to a study published today in the Proceedings of the National Academy of Sciences.

    “Based on historical and lived experience, people expect certain conditions to prevail in the ecosystems they depend upon. Climate change is now introducing strong trends that push conditions beyond historic levels,” the authors wrote.

    Led by the Gulf of Maine Research Institute, researchers looked at 65 large marine ecosystems around the world over the past 164 years to determine how frequently “surprising” ocean temperatures occur, with surprising defined as an event expected to occur about two times in 100 years, lead author and chief scientific officer at the Gulf of Maine Research Institute Andrew Pershing told EHN via email.

    Pershing and colleagues reported that over the past seven years, the planet averaged 12 ecosystems each year experiencing the kind of unusually warm temperatures that someone in the given region would expect to see only a couple times in a century. In 2016 alone there were 23 such events.

    Surprise! Unexpected ocean heat waves are becoming the norm by Brian Bienkowski, Environmental Health News, Aug 6, 2019

    The sub-headline for the above article:

    “We are entering a world where history is an unreliable guide for decision-making”

  75. John,
    They mention that unexpected ocean heat waves “also occurred following El Niño conditions” which is not surprising to most scientists that follow ENSO. So is the unexpectedness of this due to the inability to accurately predict the occurrence of El Ninos?

  76. John Hartz says:


    Here’s the “Significance” statement of the PNAS article. I believe the bolded sentence answers your question.

    Based on historical and lived experience, people expect certain conditions to prevail in the ecosystems they depend upon. Climate change is now introducing strong trends that push conditions beyond historic levels. Using ocean ecosystems as a case study, we show that the frequency of surprising temperatures is increasing faster than expected. We then use these events as motivation to develop a theory for how temperature trends and events will impact natural and human communities. The theory suggests that strong trends will decrease the abundance and productivity of natural communities. Increasing trends will also challenge how people make decisions, and our theory identifies the conditions under which there is a significant payoff for people to bet on the trend.

  77. OK, so it looks as if it may be mainly in the arctic?

  78. John Hartz says:


    The text immediately preceding the set of graphs you have shown…

    The past variability and the recent increase in surprises was not uniform across the globe. In most regions, the difference between the number of surprises and the number expected based on the local trends and variances increased suddenly in 1998 (Fig. 2A). The number of these “surprising surprises” continued increasing in the Arctic and North Atlantic. The Indian Ocean and Pacific regions leveled off then began increasing in 2010. Between 1999 and 2018, there were many more regions with more surprising temperature events than expected (48 of 65; Fig. 2B).

Leave a Reply

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

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

Twitter picture

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

Facebook photo

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

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.