A “hiatus” in some people’s “skepticism”?

There’s a new paper doing the rounds of the blogosphere, called Varying planetary heat sink led to global-warming slowdown and acceleration. Judith Curry has a post about it. The basic result of the paper is that the “missing heat” is going into the deep parts of the Atlantic because a salinity variation is allowing warm water to sink rapidly. That sounds plausible, but I don’t really know enough to judge. The reason it’s generated some interest seems to be because it suggests that the “hiatus” will last another decade or so, and because it suggests (although this appears to only be in the press release) that only half the warming between 1970-2000 was anthropogenic (or due to global warming).

The paper concludes with

The next El Niño, when it occurs in a year or so, may temporarily interrupt the hiatus, but, because the planetary heat sinks in the Atlantic and the Southern Oceans remain intact, the hiatus should continue on a decadal time scale. When the internal variability that is responsible for the current hiatus switches sign, as it inevitably will, another episode of accelerated global warming should ensue.

Many seem to ignore the very end of that sentence and focus only on the part that says the hiatus should continue on a decadal time scale. Now, the paper seems to have gone through the historical records (including the Central England record) and found some kind of 60-70 year pattern of warm and cool phases, and is then arguing that we’re in some kind of cool phase now and that it will last another decade or so, since it has existed for about 15 years already.

One issue I have is that it is very likely that none of the cool phases in the past were coincident with the planet being significantly out of energy balance – as we are now. Even though we are in this supposed cool phase, we are still warming at something like 0.1 degrees per decade (which is something else the paper rather failed to point out). It’s possible that we could sustain this slowdown for another decade or so, but if we continue increasing our emissions, that would imply that we could remain in a cool phase with slow surface warming even when the energy imbalance is > 1 Wm-2. I find that somewhat implausible, but I may well turn out to be wrong.

Although I find the possibility that the “hiatus” could continue for more than another decade, or longer, a little unlikely, there is a claim in the press release that I find rather strange. It says

Rapid warming in the last two and a half decades of the 20th century, they proposed in an earlier study, was roughly half due to global warming and half to the natural Atlantic Ocean cycle that kept more heat near the surface.

When people say things like this, it makes me think that they don’t really understand anthropogenic global warming (AGW). AGW is simply the fact that we are increasing atmospheric greenhouse gas concentrations (GHGs) which then act to trap outgoing radiation, producing an energy imbalance, and increasing the energy in the climate system. There isn’t really some special anthropogenic mechanism that simply heats the surface. The surface warms because some of this extra energy heats the surface, increasing the surface temperature, which then reduces the energy imbalance.

On the other hand, its very likely that variability means that sometimes the surface will warm faster than at other times. Therefore one could define the anthropogenic (global warming) contribution as the mean long-term trend, and the natural contribution as being variations from this mean. If we consider the period 1970-2000, the Skeptical Science Trend Calculator suggests that we were warming at 0.16 degrees per decade. If half of this is global warming and half is natural, that would suggest that the global warming contribution was 0.08 degrees per decade. If so, that would suggest that over about a 60 year period, the trend should be 0.08 degrees per decade. However, this would imply that the trend since 2000 would have to be about 0 degrees per decade, which it isn’t. In fact, if you go back to the Skeptical Science Trend Calculator, the trend from 1970-2014 is also about 0.16 degrees per decade. So, if half the warming between 1970 and 2000 was natural, and we’ve been in a cooling phase since 2000, how can the 1970-2014 trend be the same as the 1970-2000 trend. It doesn’t really make sense.

So, as far as I can tell, the press release has a claim that isn’t in the paper and that doesn’t really make much sense, and the conclusion about the hiatus continuing for another decade or so is really just based on identifying some kind of 60-70 year cycle in the historical records – none of which contain periods really comparable to what we’re going through today. If I remember correctly, there was a massive outcry when the press release for the Marcott et al. Holocene temperature reconstruction paper rather overplayed the significance of part of their reconstruction. I wonder if the same people will be similarly shocked by the, apparently, unsupported claims in this press release? Don’t bother answering that question. I suspect we all know the answer. And, to be honest, I don’t really care. I do wish people wouldn’t overplay the significance of their papers in press releases, but they do and it’s really a systemic problem, rather than just a few individuals.

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

100 Responses to A “hiatus” in some people’s “skepticism”?

  1. My impression is that the results and conclusions of Chen and Tung deviate from several other papers in a couple of ways:

    – Their analysis tells that the Pacific as whole has a smaller role in the variability of OHC. They seem to tell that the changes in the Pacific are mainly in the upper layers and that the changes of Eastern and Western equatorial Pacific largely cancel. Thus the most important changes take place in the deep Atlantic and Southern oceans.

    – They seem to believe that the role of aerosols was not large in the coolness of 1960s and 1970s as the Atlantic variability played the dominant role in that case as well.

    Their results might have a small effect on the estimate of TCR, but their results are in full agreement with the conclusion that over the last 60 years the observed warming is most likely approximately equal to the influence of AGW (i.e. 100%). As several commenters have written at Climate Etc, concluding that the paper would support Judith Curry’s idea that warming since 1950s could be less that 50% AGW is without merit.

  2. Pekka,
    I haven’t actually read the comments on Judith’s post, so am impressed that people have pointed this out. I agree that their results are broadly consistent with the conclusion that the observed warming over the last 60 years is almost AGW. That’s why I don’t understand why they put that extra statement in their press release. I do think, however, that their argument that the “hiatus” will continue for another decade or so is weak. That seems to be based on a curve fitting exercise using – in some cases – data that only covers a tiny fraction of the globe (central England). It’s possible, but if we continue to increase our emissions, I’d be surprised if it were to continue much longer.

  3. According to email from Tung to Curry the press release is written by a reporter and influenced also by their earlier PNAS paper. Therefore it contains also statements that are not supported by the present paper, but not contradicted either.

  4. Pekka,
    Okay. That’s why I added the final few sentences to the post. I know how press releases are generated and how they can get out of control quite easily. Ideally, those involved should stop anything too outrageous from being said, but it can sometimes be harder said than done. I agree that what is said might not be contradicted by the present paper and, I had, to be fair missed that the bit I quoted included in an earlier study. Tenuous, but okay.

  5. The SkS Trend calculator may not be the best tool to study the influence of the natural variability on the short term trends. It uses a regression method to reduce the influence of the variability caused by of El Nino.

    I do not know by heart how much, but I thought it was known that in 1970-2000 the short term trend was higher as expected due to natural variability. That no one is embarrassed by this continual focus on short term trends (whether this one or the ensuing “hiatus”) is really, really weird. If that is what your ideology needs, …

  6. Victor,
    Sure, I tend to agree, except I didn’t use the version that reduces the ENSO influence. That can be selected, but I didn’t. However, it would seem that if one is going to argue that half the warming in a 30-year period is natural, and hence that the natural contribution in the period after that 30-year period would be cooling, then you should see something if you compare the 30-year trend with the trend for 44 years. We don’t. They’re essentially the same.

  7. Victor,
    In fact, if you use the version that includes the Foster and Rahmstorf ENSO correction algorithm, the 1970-2014 trend (0.17 degrees per decade) exceeds the 1970-2000 trend (0.14 degrees per decade). So, that would seem to agree with the idea that there was a warming contribution in the 1970-2000 period and possibly a cooling contribution in the 2000-2014 period (although not obvious), but they’re small. Maybe, of order, 0.1 degrees per decade.

  8. Victor wrote “It uses a regression method to reduce the influence of the variability caused by of El Nino.”

    The trend calculator linked to the SkS home page gives a reference to the 2011 Foster and Rahmstorf paper, but I think that is just for the details of the trend calculation (and confidence interval) for autocorrelated data. It makes no correction for ENSO, to do that you need to use the version that works on the corrected data, which is linked to the blog post for the trend calculator.

  9. hvw says:

    I guess everyone agrees that the present paper does not support the claims to attribution and to the period of an alleged “cycle”, but rather stems from another paper that presents yet another explorative time-series analysis with mostly educational value.

    Apart from press release issues the authors are certainly guilty of fostering a wrong impression, as they could not resist to sex up their abstract with this claim, which doesn’t seem to come out of their present study.

    I’d like to see some forgiveness on the side of the recipients and a shift of focus to what Chen and Tung actually found out: The marked change in oceanic heat transport, which coincides with a flattening of global surface temperature increase. Everybody seems to jump to conclusions about attribution, TCS, and future trends, instead of taking that bit to help working towards an understanding of the mechanisms.

  10. hvw,
    I agree with your last paragraph. That is an interesting issue and is something, I think, that others have also considered. I think the main difference here is that they’re arguing that it’s in the Atlantic and others have considered the Pacific.

    The fundamental point that I would make is that all this work shows/accepts that we still have an energy imbalance and that it is around 0.5 W/m^2. Hence the fundamentals of AGW are sound and the “hiatus” doesn’t really challenge that. The interesting question – as Pekka mentioned in an earlier comment – is what this implies with respect to the Transient Response (TCS). If more energy goes into the deep ocean that we had previously thought, it could reduce it slightly. However, if this results in slower surface warming, that means that our energy imbalance will be larger (for a given emission scenario) than if surface warming is faster. That implies that we will accrue more energy than we would if surface warming were faster and would suggest that the effect may be smaller than one might think.

  11. Catalin C says:

    To add to the insightful observations already made here:

    – it looks to me that the paper in no way shows that the warming that is now seen deep into Atlantic is actually coming from the surface of the Atlantic itself, most notably this graph http://www.carbonbrief.org/media/325897/oceanheatuptake_chentung-2014-.png seems to me to suggest that the surface heat could be mostly created and then sank in Pacific and Indian but (so far) is most evident in the medium depths of Atlantic and Southern Ocean; this also happens to fit very well with the mechanism already described by England 2014 and has the specific advantage of explaining the very peculiar patterns of surface warming in different parts of the Pacific (which this new paper does not explain at all);

    – also to me on this graph http://www.washington.edu/news/files/2014/08/WarmingHiatus.png the two “hiatus” events which the paper tries to compare look very different, the previous one might indeed be more related to salinity (since we see the changes in the salinity graph preceding the changes in warming) but the latest “hiatus” does not look at all similar, I would say that warming precedes any salinity changes and the very steep gradient seen in both graphs (compared to previous “hiatus”) would suggest a cause coming from above and even pushing against and overpowering a salinity barrier (which would be very consistent with our estimates of radiative imbalances in those two intervals);

    – I think that now certain questions regarding depths over 1500-2000m are becoming even more obvious;

    – I happen to be among the people that are a little more worried about the evolution of future trends, IMHO there is a good chance that the warming of the deep oceans could actually accelerate even further, providing extra denial opportunities for the merchants of doubt; in this regard the extended-depth (6000m) ARGO probes plus better sampling can not come soon enough.

  12. Catalin,
    Good points, thanks.

    What do you think of this post. I can see where you’re coming from with your concern about the deep oceans masking some warming, but it seems to me that one could make a hand-waving argument that it is unlikely that we’ll ever build up an energy imbalance much bigger than 1 W/m^2. If so, even if more energy is going into the deep oceans, it may well slow the surface warming, but maybe not as much as a naive calculation would suggest (i.e., the energy imbalance will grow faster than if less energy were going into the deep ocean which will then accelerate surface warming).

  13. I am unimpressed. First off, the press release does indeed built on the earlier PNAS paper as Pekka has rightly pointed out. The very PNAS paper which I found utterly unconvincing given that it dismissed multi-decadal changes in aerosol forcing (more or less) out of hand.

    I’m afraid, the same pattern can be found in the new paper. Just skimmed over it and as before, the premise of the paper is that aerosol forcing didn’t change much. Unfortunately, it is a ridiculous assumption to make if one considers the subpolar (North) Atlantic ocean Basin. We know that clean air SW radiation has changed a lot over the last 60 years. You have to be a statistician w/o physical background to believe that this wouldn’t have any noticable impact on NH-circulation and thus precipitation, salinity, temperature and OHC. Sure, interaction goes both ways, atmosphere-ocean and ocean-atmosphere. But to assume that it only goes one way (ocean-atmosphere) is foolish and ignores a huge body of literature which suggests otherwise. They mention their old nonlinear baroclinic adjustment concept (AMOC increase leads to atmospheric heat transport decrease), but fail to allow for the reverse response. And I have absolutely no doubt that the reverse response (increased atmospheric heat transport due to changes in forcing, strongly changing NH-aerosol forcing in particular) exists.

    Therefore, the paper doesn’t provide any new – let alone convincing – evidence for the Atlantic ocean being the driver of the slowdown. I’d rather go with Huber and Knutti (or Schmidt et al for that matter). Still, it is interesting to see what gets published in Science 😉

  14. little (but embarrassing) burp on my part: clean air = clear sky

  15. If you think press releases in climate science are bad, you should peruse some of them in the field of high temperature superconductivity. These people have no shame whatsoever. They have elevated the art of turning a minor incremental experimental or theoretical advance into a dramatic breakthrough via press release, into a science! When this phenomenon is brought to their attention on their comments section, the post magically disappears! Even criticism from their demonstrably more competent peers does not sway them from their belief that incrementalism in their field equals an imminent Nobel Prize – because some undergraduate in the communications department deems it so. I could name names but it wouldn’t help. They know who they are.

  16. Karsten,
    Thanks for the comment.

    You have to be a statistician w/o physical background to believe that this wouldn’t have any noticable impact on NH-circulation and thus precipitation, salinity, temperature and OHC.

    Could this be why Nic Lewis apparently thinks it’s quite good?

    TLE,
    I think the press release problems are inter-disciplinary.

  17. Joshua says:

    Someone please correct me if I’m wrong – but aren’t the results of this paper derived from Argo data?

    If so, we should be sure that it will be rejected by a large # of “skeptics,” – those who have said in the past that OHC data are insufficient to use for scientific analysis.

  18. Fred Moolten says:

    The error in the 2013 PNAS paper (the source of the press release attribution statement) appears to lie in confusion on the part of the authors between the capacity of natural variability to affect the rate of warming from anthropogenic forcing and the role of natural variability as a source of warming in its own right. It’s entirely likely the AMO and other oscillations (to the extent they can be cites as “natural”) made the late 20th century warm faster and the early 21st century warm slower – the PNAS data are consistent with that. However, based on OHC data in both papers as well as elsewhere, the bulk of the warming must have been a forced response – otherwise OHC uptake would have been negative. What this means is that in the absence of the forcing, the late 20th century would have warmed very slightly (less than 0.1 C) and the early 21st century would have cooled.

    I’ve discussed some of the attribution calculations elsewhere – for example, see Climate-etc-May 19, 3:19PM.

  19. ATTP,
    why am I not surprised that NL likes the paper. If he wishes to continue to ignore all the work Martin Wild and colleagues have done over the years, he is entitled to do so … and so is K.K.Tung. Good for me … I have to read a few papers less.

  20. Similarly with Joshua I have got the impression that the paper is mainly on interpreting the data. It appears to tell that the deep Atlantic and Southern ocean have warmed, when the atmosphere has not and vice versa.

    If the paper is not based on that, or if those conclusions are highly dependent on questionable assumptions, then the paper gives a wrong impression. In the opposite case, I don’t fully understand the criticism of KarSten.

  21. Pekka,
    I consider it a highly questionable assumption that the atmosphere (via forcing response) can’t have impacted the state of the ocean (salinity, overturning strength etc). Essentially, that’s the premise of the paper. IMHO, very unlikely if not implausible. Just my 2 cent …

  22. Fred,

    appears to lie in confusion on the part of the authors between the capacity of natural variability to affect the rate of warming from anthropogenic forcing and the role of natural variability as a source of warming in its own right.

    Exactly, this is – in my opinion – precisely the issue. People are confusing the role of internal variability in determining the rate at which we warm back towards equilibrium (which it can do) and the ability of internal variability to warm in its own right (which it can’t really do – or not much).

    Joshua and Pekka,
    Indeed, I think that is right. My reading of the paper is that they do a lot of analysis of OHC data and then discuss the implications of their analysis without really doing anything else in any detail (modelling-wise at least).

  23. KarSten,

    That’s implausible, but how are the observations about OHC dependent on that? Is that assumptions hidden in the methods of analyzing the empirical data (ARGO and/or reanalysis)?

  24. The 50/50 attribution is conceptually simple to understand. I am sure somewhere on the global average Temperature versus CO2 curve you will find a 4C TCR for a restricted span of time. If 50% of this is natural variation, then the TCR=2C — which is more in-line with what Huber and Knutti find in their recent paper, a TCR=1.8C.

    Here is a math example:
    assume
    R=0.1 C/decade underlying warming trend
    and
    N=A*sin(w*t) of oscillating natural variation

    The natural variation compensation resulting in a hiatus occurs periodically whenever

    R+dN/dt=0

    or
    0.1 = -A*w*cos(wt)

    This will result in alternating max warmings of 0.2C/decade and flat regions of 0.0C/decade if the A*w factor equals 0.1.

    Now consider that the skeptics are going to the well TWICE based on not understanding (or faking) what gets “baked in” to the original analysis. They are assuming another 50/50 attribution on top of what is there already and so the Nic Lewis-fans can say that the TCR=1C.

    You really have to understand the denier mind frame, which is not to do science, but to increase FUD.

  25. Pekka,

    OHC observations are fine (within the known limits of reliability/applicability), but I (strongly) disagree with the way they interpreting it. And that’s what the paper is about, an interpretation exercise of observed data.

    My point was, that OHC is changing as a function of internal AND external forcing variability. The latter used to be particularly strong over the North Atlantic. Assuming long delays in the response of the AMOC strength to such forcing changes (say in the order of a decade), any external forcing fingerprint in the OHC data might be hidden, camouflaged (and potentially misinterpreted) as pure internal variability. Unfortunately, the authors didn’t spend much time elaborating on that problem.

    Does that make sense to you? Not sure I got your point completely.

  26. KarSten,

    My question is really about the significance of the experimental observations. If it’s true that the heat content of deep Atlantic and Southern oceans have varied that much, this seems to tell something significant about the ocean dynamics, what exactly is surely not answered by that paper, but the observation is interesting, if correct. The same is true for their other conclusion that the total heat content of the Pacific has varied less in spite of the strong variations in the more shallow layers of both Eastern and Western equatorial Pacific.

    I don’t see anything in their results that would have significant direct effect on estimates of climate sensitivity, but improving understanding of ocean dynamics is likely to improve ultimately also the accuracy of projections of future warming.

    I would judge the real value of the paper on the basis of their analysis of empirical data, and give little weight on the more speculative comments that are not in the focus of their actual work as I see it.

    If this paper includes speculation that’s not of much value, it’s surely not very exceptional in having such a weakness. Such comments are added often both to push ideas that the authors like for some other reason and to help in getting the paper published in journals that insist on high interest value of the conclusions.

  27. Pingback: There’s no such thing as natural warming! | …and Then There's Physics

  28. Rick says:

    Hmm, a 60-or-so year temperature cycle, with, circa 2000, entering a level or cooling period which will last for a few decades or so, followed by another rapid warming trend, with a lower estimated rate of temperature rise as the long term trend. This is a pattern we’ve seen before in the curve-fitting exercises of folks like DocMartyn (see his figure 8 in particular). The question is whether this proposed mechanism will work out to make this more than curve-fitting. I am sympathetic to the idea that there is significant internal variation of the system that needs to be understood. Whether this is a genuine step in that direction remains to be seen.

  29. Rob Painting says:

    Yes, this appears to be yet another example of the authors of a new paper sexing up their claims.

    That the Pacific Ocean hasn’t accumulated much heat in the last 15 years isn’t really surprising. When the Interdecadal Pacific Oscillation (IPO) is negative (as it has been since about 2000) the intensified trade winds and mid-latitude westerlies ramp up the export of warm tropical water poleward at the surface. The strong circulation also pushes more warm near-surface water through the Indonesian Archipelago into the Indian Ocean.

    The intensified trade winds export the tropical surface water to the subtropical gyres where they are subducted into the ocean interior (Ekman pumping) and thus result in the strong warming in the Southern Ocean – as the boundaries in the study have all three Southern Hemisphere subtropical ocean gyres included in the Southern Ocean. The North Pacific subtropical gyre doesn’t reach down as deep as the Southern Hemisphere ones either (for reasons not worth going into here).

    If you look at the warming trends in the individual ocean basins in their paper you’ll see that the Pacific Ocean basin warmed significantly during the positive phase of the IPO (1977-2000) – when the trade winds are weak and poleward export of warm surface water out of the tropics diminishes. Again, this is to be expected.

    I think the most important point neglected in discussion of natural variation – aka the IPO – is that these trade winds in the last decade and a half have been the strongest we’ve ever recorded. We would have seem more surface warming if it were not for this. So even with an anomalously strong wind-driven ocean circulation, and a reduction in solar radiation reaching the Earth’s surface in the recent decade, the Earth’s surface is still warming. This doesn’t bode well for the slight future.

    And when I write slight future, the recent spin down of the North Pacific subtropical gyre and the surface cooling of the North Atlantic polar gyre suggests the circulation is already headed back toward the positive phase of the IPO. We’ll have to wait and see if other indicators start showing the same signal.

  30. I’m tired of seeing talk of a global warming slowdown or hiatus. Coauthor Tung was cited as saying as much:

    http://www.nature.com/news/atlantic-ocean-key-to-global-warming-pause-1.15755
    “Tung says that his results show that global warming hasn’t halted.”

    Global warming – as in global heat accumulation – has probably been accelerating (the underlying trend line is a nonlinear increasing function) ever since 1850. Let me explain.

    In the Judith Curry post that ATTP links to, Curry says this:

    “Anthropogenic warming does not explain why the 21st century hiatus is warmer than the mid 20th century hiatus which is warmer than the turn of the 20th century hiatus.”

    She’s wrong. It does. If we include the heat in the oceans and therefore the accelerating warming in the oceans and include analyses like those done by Foster, then what we probably have very long term since 1850 with respect to global heat accumulation or global warming is not a multi-decadal fluctuation around a linear increasing trend but a multi-decadal fluctuation around a nonlinear increasing trend. One of the properties of such a graph is that the “flatter parts” of the increasing function will have average slopes that are successively greater.

    In this post that ATTP links to, coauthor Tung answers what Curry says above and tacitly communicates essentially the same idea by saying the following:

    “Because of global warming, the current stair step is higher than the previous step.”

    And:

    “Gavin was incorrect to say that Tung and Zhou (2013) assumed that anthropogenic warming response is linear. We did not assume that. The method uses a linear function in the intermediate step. If the actual anthropogenic warming response is not linear, the difference will remain in the residual and it was added back. In fact we repeated the calculation using many different but reasonable nonlinear functions in the intermediate step and obtained approximately the same result.”

    When this new Tung paper was reported in the news on TV, they showed a video of Tung saying that the heat in the oceans should be included and should even be the main metric in measuring global warming.

    This makes sense since the graphs like this one

    show that more than 90% of the accumulated heat goes into heating the oceans and only roughly 2% goes into heating the atmosphere. Only 2%. Including the oceans seems to show that global warming has been accelerating since 1850 because of the greenhouse gas effect being more powerful than previously thought and therefore the long term future of humanity and its environment is worse than previously thought. And contrary to what some out there say, the shape of the surface-temperature graph being staircase-like does not reduce the urgency for change now – Tung’s paper helps to show why.

  31. KaA,
    I saw Judith Curry’s comment about AGW not explaining why the one hiatus was higher than the other. Very surprised as it would seem an obvious explanation.

    You’re right that broadly speaking, this paper illustrates the importance of considering the oceans. So, it’s rather odd that it is being interpreted in the way that it has (of course, that’s partly because of how the authors have chosen to promote it and their comment in the paper about the hiatus extending another 10 – 15 years).

    One thing I will add is that it’s not obvious that we can talk about previous hiatuses (sp). It’s really only defined as being a period when the warming was slower than we were expecting. I’m not sure that we know that previous periods when warming was slow – or nonexistent – where periods when we would have expected faster warming or not (my feeling is probably not).

  32. John Mashey says:

    There might be a hiatus in skepticism, but like the rise of OHC, there is never a hiatus in pseudoskepticism.

  33. Mal Adapted says:

    dikranmarsupial:

    The trend calculator linked to the SkS home page gives a reference to the 2011 Foster and Rahmstorf paper, but I think that is just for the details of the trend calculation (and confidence interval) for autocorrelated data. It makes no correction for ENSO, to do that you need to use the version that works on the corrected data, which is linked to the blog post for the trend calculator.

    Don’t know if this helps reduce the ambiguity, but the calculator that uses the Foster-and-Rahmstorf-adjusted data has this annotation:

    (For definitions and equations see the methods section of Foster and Rahmstorf, 2011)

    The link is to the paper isn’t pay-walled.

  34. Tom Curtis says:

    Mal Adapted, both the standard trend calculator and Foster and Rahmstorf trend calculator include that annotation, it being a reference for the method of determining statistical signficance. The F&R trend calculator allows you to calculate trends both with, or without an F&R adjustment, but only includes data for the datasets used in Foster and Rahmstorf (2011). The standard trend calculator updates the data sets so that it includes HadCRUT4 instead of HadCRUT3, and also includes BEST (land only), and the Cowtan and Wray HadCRUT4 hybrid.

  35. I got curious on, how the F&R corrections continue after 2011 taking into account ENSO (as MEI) and solar TSI. It just happened that there was a sudden change essentially at the moment data available to them ended in autumn 2011, as can be seen in this graph that uses data from the beginning of 2003 (the start point of
    the TSI data
    used, the curve starts in 2004, because it shows the 12 month moving average.)

    Later I found another source (Fröhlich and Lean file composite_d25_07_0310a.dat) for TSI and extended my comparison (I use this data also to fill a recent gap my the original source of TSI data):

    Here the lack of correction for volcanism shows clearly.

    I looked only at the GISTEMP temperatures, but that should not make any major difference for the results.

  36. My second curve may also explain, why the SkS trend calculator sees little change in trend, when the period is extended, while a visual look at the data tells that there’s a change. The explanation might be in the influence of Pinatubo on the trend calculation (meaning that the trend calculation is not representative and should not be used as it was in the post).

  37. Pekka,
    You’re going to have to explain that, because I’m not understanding what you mean. I’m not seeing some kind of change and am not sure that “it looks different” is a particularly strong argument. Also, your data only goes back to 1983 in the bottom figure, while I was starting at 1970.

  38. Pekka,
    I’ll add, though, that if you do the trend from 1970-2000 you get 0.16 degrees per decade. If you do 2000 – 2014 you get 0.07 degrees per decade (with uncertainty of 0.16 degree per decade). The problem with this is that the end of the 1970-2000 trend line does not join up with the start of the 2000-2014 trend line and so such an analysis ignores a large jump in temperature in the late 1990s and that doesn’t make sense to me. Continuing the trend, rather than starting again seems much more reasonable way to do this.

  39. Pekka,
    Also bear in mind that the trend calculator I was using wasn’t using the Foster and Rahmstorf adjustments.

  40. Pekka,
    Actually your figures don’t look the same as the Skeptical Science Trend Calculator that includes the Foster and Rahmstorf adjustments.

  41. What I meant by my latest comment is that the dip due to Pinatubo lowers the unadjusted trend 1970-2000, because it occurs close to the end of the period, but has little effect for the longer period 1970-2014 because it’s exactly in the middle of the period.

    Warming from 1970 to 2000 is not any less because of a dip in 1991-93, but the calculated trend is. I haven’t tried to check, how large the effect is, but it might be noticeable.

    What I did is to extract the trend and the coefficients of TSI and MEI from F&R paper as accurately as I could. The rest is simply calculation from the data without any fitting. I have no further comments on its similarity or deviation from SkS trend calculator, as I haven’t studied it at all. It’s obvious that the trend agrees well with the adjusted data as far as F&R has data at the time, except that I don’t correct for Pinatubo. Thus I have probably extracted the coefficients correctly.

  42. Tom Curtis says:

    Actually, looking at the SkS F&R trend calculator, it appears not to include any data outside the original F&R analysis period. So, while it can be used at analyze sub-intervals of that period, it cannot be used to compare the data beyond the analysis period as Pekka has done. The standard SkS trend calculator shows a marked difference in trend between the two different start points used by Pekka above (0.024 C per decade for 2004 forward, vs 0.172 C per decade for 1984 forward).

  43. Tom,
    Indeed, but then this runs into the issue that Tamino pointed out here.

    Pekka,
    Okay, I see what you mean. I don’t think that really changes the point though. Pinatubo isn’t going to have changed the trend enough for 50% of the warming in 1970-2000 to have been natural – I think.

  44. ATTP,

    When the statement concerns specifically the average surface temperatures of the latest 2.5 decades of 20th century, I do believe that 50% natural variability is very well within the limits of uncertainty, not even close the edge of the plausible range. There are some questions on, how to define that, but at least many of the possible definitions support that claim, if not all.

  45. jsam says:

    What of the peer-reviewed science on this?

    Huber and Knutti (2011) quantified that human attribution as being 74% and 122% due to humans (with a best estimate of around 100% human attribution). In other words, natural variability is not responsible for the observed warming trend.
    http://www.nature.com/ngeo/journal/v5/n1/abs/ngeo1327.html

    Since then, Gillett et al (2012) also examined the human attribution of the warming trend observed. They found that humans are responsible for 102% of observed warming from 1851 to 2010 and 113% of the observed warming from 1951 to 2000 and 1961 to 2010 (averaged together).
    http://www.agu.org/pubs/crossref/2012/2011GL050226.shtml

  46. Pekka,

    When the statement concerns specifically the average surface temperatures of the latest 2.5 decades of 20th century, I do believe that 50% natural variability is very well within the limits of uncertainty,

    Not sure I agree with that. Certainly doesn’t appear to be consistent with the IPCC attribution figure. Unless we’re talking about something different, I fail to see how internal variability can have accelerated the 1970-2000 warming by a factor of 2 (i.e., I fail to see how the underlying anthropogenic trend can be below 0.1 degrees per decade). Maybe I’m missing something though.

  47. jsam,
    Indeed, and the IPCC’s attribution figure seems to suggest that between 85% and 115% (eyeballing) is anthropogenic for the period since 1951.

  48. ATTP,
    To see, what’s consistent with IPCC you must determine the AGW contribution based on the range given for TCR and compare the warming calculated using that range for the period 1975-2000 to the actual observed warming. There’s some ambiguity in deciding what’s the observed warming from 1975 to 2000, but I’m pretty sure that half of the observed warming falls in the range calculated from TCR for most ways of determining the observed warming.

  49. ATTP,
    The period from 1951 to 2010 is 60 years, the period we are discussing is 25 years, but most of the warming took place over those 25 years, while less than half of AGW falls into that period.

  50. Pekka,
    Okay, so you’re suggesting that given the range of TCR, it is possible that half the warming between 1975-2000 could be natural. However, don’t you then need to understand why the TCR has these low values. As I understand it, the energy budget estimates are based on assuming that all the warming is anthropogenic and the range comes from the uncertainty in the forcing (mainly aerosols). So, it seems that your argument is then not consistent with how TCR is estimated.

  51. Pekka,

    The period from 1951 to 2010 is 60 years, the period we are discussing is 25 years, but most of the warming took place over those 25 years, while less than half of AGW falls into that period.

    Sure, but I’m not really seeing a strong argument for natural variability contributing half the warming between 1970-2000. I can see how it can have accelerated the warming, but doubling doesn’t seem consistent with the evidence. In fact, as I see it the change in anthropogenic forcing during that period was about (eyeballing) 1 W/m^2 and the change in temperature was around 0.4 degrees. Together that would indicate a TCR of around 1.4 degrees per doubling of CO2. If we assume that only half the warming was anthropogenic, then that gives a TCR of around 0.7 degrees per doubling. This seems unrealistic, unless the aerosol forcing is much bigger than we expect. Again, that’s not a definitive argument, but I still don’t see any really strong argument for how natural variability could have doubled the rate of warming between 1970 and 2000.

  52. ATTP,

    I’m not arguing for any specific numbers. I’m arguing only against claims that I see as stronger than IPCC AR5 can support.

    Of course TCR may be higher than the lower part of the range IPCC considers plausible, but I don’t see, how you can justify so much stronger statements than IPCC. The average value proposed by IPCC (i.e. around 1.8 C) explains only about half of the warming between 1975 and 2000 (with the caveats of my earlier posts).

  53. Pekka,

    I’m arguing only against claims that I see as stronger than IPCC AR5 can support.

    But who’s made arguments that are stronger than IPCC AR5 can support? It’s my understand that suggesting that internal variability cannot have produced as much as 50% of the warming between 1970 and 2000 is consistent with what the IPCC AR5 says. I guess they don’t strictly address this time interval, but I certainly don’t think there is anything in AR5 that supports the suggestion that half the warming during a 30 year internal could be natural and most of what I’ve seen suggests that anthropogenic influences dominate after 1951.

    Okay, I’ll make a stronger statement. The IPCC attribution figure seems to suggest that anthropogenic forcings contributed to between 85% and 115% of the warming between 1951 and 2010. The Skeptical Science Trend calculator suggests a trend of around 0.125 degrees per decade. If 85% is anthropogenic, that implies an underlying anthropogenic trend of around 0.11 degrees per decade (I’m rounding a bit here, but this is all a bit approximate). If I apply that to the 1970 – 2000 period, then the anthropogenic component is about 65% of the warming. So, again, I don’t really see a strong argument for anything like 50% being natural. Of course, one can carefully select numbers and we can’t rule out that the anthropogenic component between 1951 and 2010 was less than 85%, it just seems very unlikely.

    So, to be clear, I realise that it could be possible. My argument is simply that it appears extremely unlikely.

  54. Immediately after IPCC AR5 was released I made some calculations and concluded that the warming from 1950 to 2010 was essentially, what a TCR of 1.8C would produce. As the period extends to both sides of 1975-2000, I guess that about 45% of that warming comes from this shorter period. If the warming over the shorter period is 90% of the that over the full period, then we have the result of 50% AGW. The rest is natural variability essentially by definition.

    Whether the 90% I mention is correct, is arguable, but it’s close enough to the observations to be used.

    I have stated before that the shares in the attribution are not essential. You might ask, why I care to argue then. From my point of view the question is, why do you wish to make arguments that are questionable from some fully reasonable point of view, and write posts about them, when the issues are not essential anyway. My impression is that you can only lose in that activity.

    It’s not productive to argue that natural variability must be small, when that’s impossible to prove. It’s better to concentrate on the strength of AGW on which we know more.

  55. Pekka,
    I don’t actually follow your argument, but I’ll give it some thought. I’m really not following how you get 50% AGW over the period 1975-2000.

    It’s not productive to argue that natural variability must be small, when that’s impossible to prove.

    But I’m not arguing that it must be small (you may be arguing against your own construct of my argument). My point is simply that the claim that 50% of the warming between 1970-2000 is natural doesn’t appear to make sense. That doesn’t mean that I’m arguing it has to be small, simply that less than 50% appears more reasonable than 50%.

    You seem to be claiming that it can be 50%, but I’m really not following you. If the TCR is 1.8 degrees, then I can’t see how only half the warming between 1975 and 2000 is anthropogenic. I’ll give it some thought though.

    It’s better to concentrate on the strength of AGW on which we know more.

    I agree, and that was partly what I was trying to suggest in the post (i.e., even if natural variability influences our warming, it really only influences the rate at which we warm as it can’t really warm in it own right). The point of the post, if there was one, was to argue against those who are claiming that half the warming was natural.

    It’s possible that you’re taking my blogging more seriously than I am. If so, you should probably stop doing so 🙂

  56. ATTP,
    It’s likely that have always in mind, how someone significantly more skeptical than myself, but still willing to consider new ideas, would read your posts. I don’t consciously analyze the posts against such background, but may do it automatically. On that basis I feel that such a person is likely to reject many of your arguments outright, while you could have formulated largely the same message also in a way he would consider more positively.

  57. Tom Curtis says:

    Anders, based on pixel counts of Figure 10-5 of AR5 WG1, the lower bound (95%) interval for anthropogenic warming from 1951-2010 is 0.49 C, or 0.08 C per decade. That is 44.76% of the HadCRUT4 trend from 1975-2010 (45.76% of GISS; 47.34% of NCDC). Therefore, assuming anthropogenic trends have been constant over the 1951-2010 interval, it is on the extreme outside of probability that even 50% of recent warming has been anthropogenic. In fact anthropogenic forcings are known to increase sharply from around 1970, but we need to allow for downside error in estimated temperature trends as well so that on this test 50% natural contribution to recent (post 1975) warming is not excluded, but Pekka’s claim that 50% natural is “not even close the edge of the plausible range” goes well beyond what can be supported from the IPCC. Indeed, given that the confidence interval is a two tailed test, and all of the upper anthropogenic range counts against the 50% natural, it is fair to say that is very unlikely that 50% of post 1975 warming has been anthropogenic.

    On the flip side, similar reasoning shows the IPCC position to consider less than 15% natural contribution to the 1975-2010 warming to also be very unlikely.

    For an alternative approach, it is interesting to examine Figure 10.3 d which shows zonal attributions for 1979-2010. In some latitudes it is very clear that natural contributions could be a major, even a dominant player in temperature trends. That particularly applies to regions below 45 degrees South. Other areas, however, far more warming than can be accounted for in the natural 95% confidence range. That is true for all latitudes from 45 S to 70 N. Therefore if we trust models even a little (;)) it is implausible to argue for primarily natural warming over that period, based on IPCC data.

    Pekka is being so careful not to overstate IPCC results that he is distinctly understating them.

  58. Pekka,

    On that basis I feel that such a person is likely to reject many of your arguments outright, while you could have formulated largely the same message also in a way he would consider more positively.

    Quite possibly, but I’ve never really claimed to know how to do this 🙂

    Tom,
    When you said,

    it is fair to say that is very unlikely that 50% of post 1975 warming has been anthropogenic.

    Did you mean that it is very unlikely that only 50% has been anthropogenic, or very unlikely that more than 50% has been anthropogenic?

  59. Fred Moolten says:

    Although I agree with Pekka that expected future warming is of more practical concern than the fractional attribution of past warming, the two are related – e.g., through the use of past data to compute TCR. As best as I can tell, when these data are used, the 1976-2000 warming is consistent with an effect that is exclusively anthropogenic, with no contribution from natural variability – see, e.g, Gregory and Forster 2008. This should not be interpreted to mean that internal variability might not have played a minor role, since the estimates are too uncertain to exclude that possibility. However, the possibility of a substantial role – e.g, one almost as great as the anthropogenic contribution appears to be excluded with high certainty by the OHC data. A link in my comment above from August 22, 3:57 PM provides some additional mathematical detail underlying my reasons for this conclusion..

  60. Tom Curtis says:

    Sorry, typed “anthro” when I meant to type “natural”.

    To make it clear, it is very unlikely that as much as 50% has been natural. Taking HadCRUT4 as the temperature series, then 50% natural contribution represents 1.99 sigma from the mean distribution for anthropogenic contribution. That means there is a greater than 95% chance that “natural forcing” plus “natural variability” contributed less than 50% of the trend. That assumes normal distributions, ignores error margins on the trend estimates, and ignores the fact that based on forcing history, the anthropogenic warming rate would have been greater in 1970-2010 than in 1951-1969. If we properly account for all of that, the chance of 50% natural may creep just below 95%, but there is no way it can be interpreted as “not even close the edge of the plausible range”.

    As noted, similar things can be said about the probability of anthro contributing greater than 85% over the 1975-2010 interval. On the other hand, 1975 is a strong La Nina year, and 2010 is an El Nino which was followed by arguably the strongest La Nina on record so that there is even less chance that 50% of the 1975-2014 trend was natural, and it is possible that 100% of it was anthropogenic.

  61. > On that basis I feel that such a person is likely to reject many of your arguments outright […]

    Again, Pekka, what’s this “likely”: 67%?

  62. I just noticed that Judith Curry keeps stubbornly the line that even over the 60 year period from 1950 less than 50% AGW is as likely as more than 50%. I have argued against that so many times there that I have to figure out something new to say before I comment there.

  63. Pekka,

    I have argued against that so many times there that I have to figure out something new to say before I comment there.

    Good luck. I think Gavin Schmidt has also been trying to argue against this too – also without success.

  64. FYI, my own tentative was this:

    It is be very easy to condemn any insertion of confidence levels post hoc. It is so easy that we can surmise that Mr. T is a very, very sharp shooter.

    http://judithcurry.com/2014/08/24/the-50-50-argument/#comment-620839

    I concede that this comment may not meet Pekka’s likeliness.

  65. My argument against Judith Curry’s position is basically the same I have had in this thread:

    The best approach is to estimate TCR and it’s uncertainty using all evidence available. The anthropogenic contribution (excluding aerosols that must be handled separately) has been essentially, what can be calculated from TCR, because the time scales are close enough to those used in defining TCR. If we wish to give limits for the contribution of natural variability, which may involve both changes in OHC and variability in TOA balance, they are obtained as difference between the observed warming and calculated AGW.

  66. Pekka,
    I think that sounds sensible. A question though, if you calculate TCR using an energy budget approach, what time interval would regard as sufficient to assume that the influence of internal variability is zero, or small?

  67. ATTP,
    That’s a difficult question (at least for me, and presently I think that also for everybody else).

    My understanding is that most long term variability is controlled by the state of ocean circulation, but understanding oceans is far from satisfactory. It must be possible to estimate some limits for the time scales and amplitudes even without full understanding, but I’m not aware of such estimates.

    The state of oceans affects also the state of the atmosphere and calculating the dynamics of oceans requires that the coupling of oceans with atmosphere is included.

    I can comment only on this qualitative level, giving any numbers seems impossible for me.

    Even the historical data tells too little as interpreting it properly requires that all major factors are known at some satisfactory level. If one important piece is missing that may distort the understanding of the whole. Looking at the data, it’s plausible that there’s significant variability on the time scale of roughly 60 years, but that conclusion may be also erroneous as there’s not sufficiently detailed data to tell anything definitive. What takes place over somewhat longer periods is even less bound by anything. My understanding is that oceans might have persistent enough modes to support even such variability.

    All that makes it very difficult in my view to make strong statements about natural variability, but that uncertainty doesn’t prevent estimation of the strength of AGW at the level it has been estimated so far. Getting more precise estimates seems, however, to be difficult as indicated by the fact that the error range given for ECS has been essentially unchanged for decades in spite of all the progress in science. (In hindsight the early error ranges must have been narrower than the knowledge of those times could support, the present situation is probably more justified.)

  68. FWIW, Pekka, you and Judy, may be in violent agreement:

    Well that whole thread made me want to pop more ibuprofen, but earlier in the comments Pekka says this:

    ‘When the statement concerns specifically the average surface temperatures of the latest 2.5 decades of 20th century, I do believe that 50% natural variability is very well within the limits of uncertainty, not even close the edge of the plausible range. There are some questions on, how to define that, but at least many of the possible definitions support that claim, if not all.’

    WHich is essentially what I am saying.

    http://judithcurry.com/2014/08/24/the-50-50-argument/#comment-620885

    Her emphasis.

    ***

    Speaking for myself, I’d observe that it’s the lukewarm gambit all over again. This time, instead of betting under the mainstream estimates of climate sensitivity, Judy’s betting on the lowest bound of the attribution problem. Because she can, I guess.

  69. One very general comment about the uncertainties of natural variability.

    In my view the uncertainties are large enough to avoid making strong claims about natural variability. That does, however, not imply that it’s correct to use such uncertainty as an excuse against accepting evidence on AGW and it’s strength.

    It’s a logical fallacy to think that the uncertainties about the natural variability make it impossible to draw strong conclusions about AGW and about understanding of AGW should affect policy decisions. That it’s a logical fallacy becomes clear, when the best approaches to address AGW are deployed in practical scientific work and policy analysis.

  70. > It’s a logical fallacy to think that the uncertainties about the natural variability make it impossible to draw strong conclusions about AGW and about understanding of AGW should affect policy decisions. That it’s a logical fallacy becomes clear, when the best approaches to address AGW are deployed in practical scientific work and policy analysis.

    I agree, Pekka, and I surmise it’s also the case for climate sensitivity, and just about any so-called “technical discussions”.

    ***

    The most important trick to notice here is the implication that “50-50” cut as something that cuts any ice whatsoever. There are at least three problems with this trick.

    First, it’s presented as an “is”, i.e. “it is 50-50”, when it’s a could, i.e. “could be as low as 50-50”.

    Second, selling out the “half-and-half” dog whistles “almost less than half” and “we’re not that responsible for GW”. The same applies to those who insist in finding “more than 50%” meaningful. The only science behind such claims is marketing psychology.

    Three, Mr. T would hate such a crisp cut. If we have big uncertainties, it’s not “as low as 50%” that needs to get promoted, but the complete uncertainty span, e.g. “between 51% and 99%”. Only thus we could see that how the quantification the IPCC, BartV and NG are selling, which goes way above 100%, rests on another interpretation of percentages.

    ***

    Oh, and AT, your ears might be ringing:

    I read ATTP on salby, his response is pretty thin gruel

    http://judithcurry.com/2014/08/24/the-50-50-argument/#comment-620944

    An assertion that does not even deserve a dot, it seems.

  71. Tom Curtis says:

    Anders, I have updated my analysis to include the trend uncertainty for HadCRUT4, and to allow for the different mean rates of forcing increase from 1951-2010 and that from 1975-2010. To do the former, I combined the uncertainty of the temperature trend with that for attribution (summed in quadrature), and for the former, I took the ratio in rates of increase in anthropogenic forcing between 1951-2011 and 1970-2011 from figure 8.19.

    Overall, the result shows the mean estimate of anthropogenic forcing is 73.45%, with a 95% confidence interval of 43.58-103.32%. That means the mean estimate lies 1.57 standard deviations from 50%. Consequently there is a 91.91% probability that anthropogenic contribution to the 1975-2010 warming was 50% or greater. Conversely, there is a 98.5% probability that the anthropogenic contribution was 100% or less. There is a 90.41% that it lies in the 50-100% range. All this assumes normal distributions, of course. Extending the period of interest back to 1971 (and hence closer to the actual inflection point in anthropogenic forcings) both decreases the temperature trend and uncertainty, thereby significantly decreasing the probability that anthropogenic forcings contribute less than 50%. Pushing the end of the period forward to the current date has the same effect.

    So, accounting for the factors I fudged in the previous calculation shows I over estimated the impact of the inflection in anthropogenic forcings relative to the uncertainty in temperature trends. There is in fact a 3.54% probability that anthropogenic forcings contributed less than 50% of the warming, but that the contribution still lies within the 95% confidence interval. Whether or not you think that justified Pekka’s claim that “I do believe that 50% natural variability is very well within the limits of uncertainty, not even close the edge of the plausible range” you will have to decide for yourself 😉

  72. Tom,
    Thanks, very useful.

    Willard,

    I read ATTP on salby, his response is pretty thin gruel

    Hmmm, does the link suggest that Judith is going to write something in which she finds Salby’s work “interesting”?

  73. anoilman says:

    Anders… If she gets a bigger weather contract with the oil companies she might. 🙂

  74. Tom,
    My statements have always been about 1975-2000, not 1975-2010. Adding 10 years to the end makes a major difference, because AGW contribution has probably been accelerating in spite of the hiatus in the observed temperatures.

    Arguably essentially 100% of the observed change in surface temperatures has taken place in 25 years, and nothing during the first 25 years or latest 10 years (arguably, because determining the turning points around 1975 and 2000 is open to interpretation).

    Determining anthropogenic forcing is problematic due to the poor understanding of the aerosol effects. Another important issue is the nonlinearity of the AGW warming trend over a period as long as 60 years. Both of these factors affect the interpretation. The shortness (25 years) of the phase of rapid observed warming allows, however, making fairly solid statements on the relative share of AGW trend over that period, and defining AGW quantitatively as anything else than the (somewhat nonlinear) trend seems impossible.

    How long is the “hiatus” likely to continue? My own guess is that not for many more years, but that’s just a guess. It’s hardly possible to say that more than 5 years or even 10 years would be virtually impossible. I think that it’s wise to acknowledge that: Longer hiatus is possible, but not expected. Being too certain about near term future tends to bite credibility as even the best near term predictions fail so often.

  75. Pekka,
    Broadly speaking I agree. As Tom has pointed out, it is possible for 50% to be natural for time periods of 25 years or so. However, it does appear that it is unlikely. Similarly I agree with you about the hiatus. It seems unlikely that it will continue for many more years. Again, it is possible that it could, but it seems unlikely. So, I have no issue with the suggestion that 50% of the warming over 25 years or so could be natural, or the suggestion that the hiatus could continue for much more than another decade. My issue is more that some are making these things sound much more likely than I think the evidence allows.

  76. ATTP,
    When we see a wrong argument, where some real observations are used incorrectly, I think that you have more a tendency to say No, whereas I say Yes, but ...

  77. Pekka,
    Possibly, but I’m not sure I’ve ever really said, “no, that’s impossible”.

  78. Tom Curtis says:

    Pekka, first, warming did not stop in 1998 – not even for surface temperatures. Granted you can get flatter trends in the GMST from 1998, but the that is only because of strong cooling in 2008 and again in 2011-12 so that the real inflection point in GMST is 2007 (if you are looking for one). Therefore there is no reason to pick 1975-2000 as a significant interval except to cherry pick for maximum natural variability. Of course, a short (26 years) interval which starts with one of the two strongest La Nina’s on record, and ends with one of the two strongest El Ninos of the last 70 years is ideal if that is your purpose. (Note, I am not saying it is your purpose, as you may only be discussing that interval because it was introduced by another.)

    Second, as anthropogenic forcing increases approximately linearly over time from 1970 to 2010, calculations of anthropogenic contributions will vary inversely with temperature trend over that interval. Therefore estimated anthropogenic contribution to the 1975 – 2000 trend will be less than that for the 1975 -2003 trend (for example), and only marginally greater than that for the 1975-2010 trend. (For comparison, here are the 1975- terminal years trends on HadCRUT4 for terminal years –
    2000: 0.187; 2001: 0.180; 2002: 0.184 2003: 0.190; 2004: 0.195; 2005: 0.193; 2006: 0.197; 2007: 0.196; 2008: 0.193; 2009: 0.186; and 2010: 0.183) Consequently any increased probability of less than 50% anthro contribution for the shorter period is largely a factor of increased uncertainty in trend rates over the shorter interval, which is statistically and scientifically uninteresting.

    For both of the above reasons I consider the 1975-2000 period uninteresting for assessing relative anthropogenic contribution. It is certainly not justified by the claim that:

    “Arguably essentially 100% of the observed change in surface temperatures has taken place in 25 years, and nothing during the first 25 years or latest 10 years (arguably, because determining the turning points around 1975 and 2000 is open to interpretation).”

    The temperature increase from 1951-2010 is 0.65 C according to AR5 (0.64C estimated from the HadCRUT4 trend). The HadCRUT4 trend from 1975 to 2000 yields an increase of only 0.49 C over that interval, or only 74.8% of the increase. The trend from 1975 to 2010 yields an increase of 0.66 C, or just over 100% of the increase. Again, if you wanted a shorter interval for 100%, the best choices are 1975-2006 (0.63 C, 97%) and 1975-2007 (0.65 C, 99.5%).

    The reason deniers focus on periods starting (for discussions of the “hiatus”) or ending (for discussions of natural variability) around 1998 is not because such periods are interesting. If they were genuinely interested in a purported hiatus, or slow down in warming, their focus would be on the apparent inflection point in the temperature record in 2007. Such a focus, however, would draw attention to the fact that the “hiatus” consists almost entirely of two unusually strong La Nina events. Therefore they make use of the fact that running linear trends shift the position of apparent inflection points to look at the spurious inflection point in 1998 . Granted many of them may not recognize that almost entirely as a statistical artifact, but you should. (“Almost entirely”, of course, because of the exceptionally strong 97/98 El Nino does also play a role.)

    Having said all that, I am happy to grant that for the shorter interval you focus on we cannot establish that natural variability plus forcings contributed less than 50% of the warming – but that is almost entirely due to the large error margins from the short time interval.

  79. As I wrote, defining the temperatures around 2000 is ambiguous. The strong short variability at time allows for many alternatives.

    I agree fully that the shorter the interval considered is and the more it’s fixed based on cherry picking end and start points, the less significant the behavior over the period is.

    When I wrote that I favor often the answer “Yes, but ..” I meant arguing along the line:

    Yes, it’s possible to interpret the time series (I looked most recently at HadCRUT4) since 1950 as consisting of three sub periods:
    – essentially flat 1950-75
    – rapid rise 1975-2000
    – essentially flat 2000-2014
    but the correctness of that description does not tell nearly as much about what has really been going on and what’s to be expected. Shorter term variability is a factor that has contributed to that interpretation as it’s dependent on giving weight for the 1998 El Niño as part of the end point. Shorter term variability allows, however, also for other interpretations, which do not agree with the one presented.

    Then I might continue on other caveats in drawing conclusions from the temperature time series.

    The main idea is that I recognize the fact that the time series has the look. I don’t argue against people, who look at the curve and tell that they see it like that, but I argue against giving erroneously strong significance for the observations.

  80. clivebest says:

    Pekka’s first comment.

    Their results might have a small effect on the estimate of TCR, but their results are in full agreement with the conclusion that over the last 60 years the observed warming is most likely approximately equal to the influence of AGW (i.e. 100%).

    That is only true if you take the observed warming to be that from 1942 to 2008. In other words if you calculate the warming from one peak to another peak of the apparent 60 year oscillation. Only then does any hypothetical AMO oscillation average out to zero. I claim that the only reliable conclusion from the temperature data is that the world has warmed by about 0.45 C since 1940. Before 1942 the geographic distribution of weather stations is too sparse to measure global temperatures reliably.

  81. Clive,

    I claim that the only reliable conclusion from the temperature data is that the world has warmed by about 0.45 C since 1940. Before 1942 the geographic distribution of weather stations is too sparse to measure global temperatures reliably.

    You can claim whatever you like. To be honest, given your rhetoric about climate scientists not saying anything unless they have a solution to the problem (which just seems so ridiculous that I can’t actually believe that an intelligent could express such a view) and your view that nuclear is the only possible solution and that anyone who thinks otherwise is an idiot, I rather take what you claim with a pinch of salt.

  82. Joshua says:

    Anders –

    ==> “Hmmm, does the link suggest that Judith is going to write something in which she finds Salby’s work “interesting”?”

    She has promised a post on Salby. “Till then you must eat the thin gruel of her thin gruel critique.

  83. Joshua says:

    Which, of course, will taste better with Clive’s pinch of salt!

  84. clivebest says:

    Anders,

    First you argue that any internal natural variation cannot warm the planet. I agree with that statement over the long term. The new paper describes evidence of a regular 60 year oceanic heat sink cycle which is also evident in the temperature data. Do you reject this ? If no, then how can all the increase in temperature since 1940 be anthropogenic ?

    To clarify my “rhetoric”: I only refer to those Climate Scientists who enter the political lobbying arena and call for immediate action to drastically curb carbon emissions. That is a perfectly reasonable thing to do but then they should also consider exactly how governments could curb emissions while maintaining living standards. If there were to be another VI IPCC assessment report then perhaps it should contain a Volume 4 : Low Carbon Energy scenarios for 2050 – where the sums actually add up globally.

    I agree with James Hansen that a large part of the solution is nuclear energy.

    Save the pinch of salt for thorium reactors !

    🙂

  85. Clive,
    You start to move in the direction of ATTP’s thought experiment.

    1) Can all increase in temperature since 1940 be anthropogenic?

    2) Can 100% of the increase in temperature since 1940 be anthropogenic?

    3) Can anthropogenic contribution be 100% of the increase in temperature since 1940?

    Do you have the same answer to all these three questions?

  86. Pekka,
    I think the answer can be the same to all 3 (yes) but maybe I’m missing something. I’m not seeing a difference between the 3 questions.

  87. ATTP,
    Some people seem to see a difference. They seem to add up all positive and negative changes separately, thus “all increase” refers to more than the difference between the final and initial state, if the change has not been monotonic. That can also be applied separately for AGW and for the natural variability.

    That’s not the way I like to think, but that kind of thinking appears to exist. Thus (1) and (3) might have different meanings and (2) might be interpreted in either of the two ways.

  88. Clive,

    The new paper describes evidence of a regular 60 year oceanic heat sink cycle which is also evident in the temperature data. Do you reject this ?

    As I understand it they went back to the 1700s using the Central England Record, so am not that convinced that it’s a robust result. You’ve just said you don’t trust the temperatures before 1942, so I’m hoping you reject it as doing otherwise would seem inconsistent.

    If no, then how can all the increase in temperature since 1940 be anthropogenic ?

    I think there is a difference between asking why we are warmer today than we were in 1940 (almost all anthropogenic, with a small solar component) and what influenced the rate at which we warmed during the last 74 years. The rate can be influenced by many factors including internal variability, but the reason we’re warmer today is entirely anthropogenic (okay, small solar effect). Of course, if there is some underlying cycle, and if we haven’t completed a full number of cycles, we could be slightly warmer today than we’d be without the cycle, or slightly colder. However, the influence of internal variability is small (0.1 – 0.2 degrees) so the longer the time interval, the smaller the relative effect is.

  89. Pekka,
    As I see it a reasonable answer to all 3 is yes. That, of course, doesn’t mean that all of the warming since 1940 was anthropogenic, simply that it could have been.

  90. clivebest says:

    Pekka,

    To answer your questions:
    1) Yes all increase in temperature since 1940 could be anthropogenic. But not all the increase from 1970 can be anthropogenic. This must include also a natural warming cycle.

    2) Can 100% of the increase in temperature since 1940 be anthropogenic?
    – Yes I think it is.

    3) Can anthropogenic contribution be 100% of the increase in temperature since 1940?
    – Yes because 1940 was a peak in the 60y cycle and so too is 2000. Therefore the net rise between them is due only to man because the net natural component is zero.

  91. Clive,

    But not all the increase from 1970 can be anthropogenic. This must include also a natural warming cycle.

    I think that’s a very strong statement, but I would agree that it is probable that the warming since 1970 was accelerated by internal variability. I would argue that it is highly unlikely to be as high as 50% though.

  92. clivebest says:

    The hiatus in warming is due to the natural cooling cycle kicking in again post 2000. Unlike 1950-1970, temperatures will not fall because anthropogenic forcing has increased since then. By 2030 a rapid warming phase will begin as the natural warming cycle begins again reinforcing AGW.

    I made a plot of this about 2 years ago – available here

  93. clivebest,
    Sure, broadly speaking I agree, but I don’t see how we can continue warming slowly till 2030 if we continue to increase our emissions. If there is a 60 year cycle, remember that it is a cycle that – in the past – existed when the planet was broadly in energy balance. We don’t know how this cycle will evolve as we move further and further out of energy balance. It seems unlikely that the cooling phase can be as long as 30 years if we continue increasing our emissions.

  94. anoilman says:

    clivebest: I find it funny that you focus so much on UK data for a global problem. I also note that your ‘blog’ promotes many many myths about power grids or even how they work. I think its unfortunate that you don’t know more about modern power generation.

    I thought you’d like this video on why we don’t need batteries for reliable power generation using renewables. They help, but aren’t necessary. This is after all why we use so many large scale utility grade batteries today.
    http://www.engineering.com/ElectronicsDesign/ElectronicsDesignArticles/ArticleID/8272/Is-Storage-Necessary-for-Renewable-Energy.aspx

    That article is brought to you buy Amory Lovins, the man who brought you the BMW i3. You can drive it today Clive!

    This is his talk on the energy grid.

  95. clivebest says:

    So how come Germany is building 19 new coal fired power stations ?

  96. Marco says:

    Clive, Germany’s decision to build new coal fired stations is largely based on the replacement of its old plants by newer, more efficient plants. However, according to DENA, the German Energy Agency, by 2020 there will be around 50% more decomissioned than the new planned plants produce (18.5 vs 11.3 GW). With energy consumption expected to increase, this shows an planned *reduced* proportion of energy produced by coal.

  97. clivebest says:

    Amory Lovins certainly is a charismatic speaker! He is right about energy efficiency, future transport etc. but wrong about electricity generation. There are intrinsic limits to wind power generation not just its load capacity of 20-30%. You need to cover vast areas of land and even then there are becalmed days across the US and Europe. Same for solar. Try this scenario instead for future energy source with low environmental impact (hopefully it embeds itself).

  98. anoilman says:

    Marco: That is very similar to China’s De-carbonizing plans. Most of its new plants were replacing plants being decommissioned. Coal consumption is expected to peak from 2020 to 2030.

    Most of Germany’s decarbonization has been to removal of old inefficient Soviet era industry in East Germany.

    Germany pays the most of energy in the world, yet its still the economic powerhouse of Europe. That shows that energy prices aren’t going to hurt.

    Clive Best: You are wrong on all points. How do you explain the areas that are already utilizing greater than 30-40% renewables? They are doing it without covering huge swaths of land too. I think you know that too.

    Perhaps the best place to put solar is to use NO land. Put it on the roofs that already cover lots of land. Most wind is placed where it is actually effective, in that way it never covers huge swaths of land either. (Round here its called Windermere, so you can probably guess where it got its name.)

    Anyways, this report will give you a beginners grounder on renewables. You will learn a lot from reading about this stuff for the first time Clive.
    https://www.purdue.edu/discoverypark/energy/assets/pdfs/SUFG/publications/SUFG%20Energy%20Storage%20Report.pdf

    This will also explain to you all the GW of batteries currently in play, as well as the fact that coal/fusion, whatever are unreliable, and why you need many many many more power plants. This is why we use renewables and batteries now, today.

    But don’t take my word for it, take a conservative think tank’s word for it;
    “Solar power and other distributed renewable energy technologies could lay waste to U.S. power utilities and burn the utility business model, which has remained virtually unchanged for a century, to the ground.”
    http://grist.org/climate-energy/solar-panels-could-destroy-u-s-utilities-according-to-u-s-utilities/

    (This links to the original think tank report, which is predicting a disruptive effect of solar alone on the power grid.)

Leave a Reply

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

WordPress.com Logo

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

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s