AMO

I regularly get comments claiming that ocean cycles are the cause of global warming. They couldn’t be more wrong.


The oscillation du jour is AMO, the Atlantic Multidecadal Oscillation. In part this is because of a post by Bob Tisdale making a number of ridiculous claims. As the latest such comment I received says,


He concludes, that multiple regression isn’t a way removing the signal. The signal is still clearly present like in Fig. 2 of this post. If you would just plot MEI on on top of that you would see it doesnt do a very good job. I can see it just by looking at it. The “eye balling” method seems to be a much more robust method in removing the signal, as Bob has shown us.

His analysis is much more comprehensive and takes AMO in account, for example. Only roughly 27% of the trend remains in GISS 60N-60S LOTI data after all of the artifacts which are propably due to inner heat transfer processes, have been removed.

Yes, folks. Bob Tisdale actually believes that “eyeballing” the correct lag and scale factor for fitting time series is better than multiple regression. The WUWT crowd laps it up. It’s near impossible to have an intelligent conversation with people who believe such things.

I won’t dwell on the folly of substituting “eyeballing” for analysis. But by all means let’s take a look at the AMO. Wikipedia describes it thus:


The AMO signal is usually defined from the patterns of SST variability in the North Atlantic once any linear trend has been removed. This detrending is intended to remove the influence of greenhouse gas-induced global warming from the analysis. However, if the global warming signal is significantly non-linear in time (i.e. not just a smooth increase), variations in the forced signal will leak into the AMO definition. Consequently, correlations with the AMO index may alias effects of global warming.

Could it be that the global warming signal is significantly nonlinear? Here’s global temperature from GISS:

Here’s a linear fit:

It sure seems that the global warming signal is nonlinear. In fact, here’s a much better characterization of that signal, from a lowess smooth:

Yep. Nonlinear.

Variations in the forced signal do leak into the AMO definition. Correlations with the AMO index do alias effects of global warming.

The way to compute AMO, according to the ERSL AMO page, is as follows:

  • Use the Kaplan SST dataset (5×5).
  • Compute the area weighted average over the N Atlantic, basically 0 to 70N.
  • Detrend that time series.
  • Optionally smooth it with a 121 month smoother.

    So the AMO is actually detrended (but only linearly detrended) North Atlantic Sea Surface Temperature (N.Atl SST). We won’t bother with any smoothing.

    Conveniently, the ERSL AMO page gives a direct link to the SST data averaged over the N. Atlantic. It looks like this:

    Evidently there’s another step we have to perform: we need to translate N.Atl SST to N.Atl SST anomaly. That looks like this:

    I computed these anomalies using the entire data set as a “baseline” period. We can linearly detrend this data, then compare it to the given data for the AMO itself:

    You can’t really see the separate curves, because they’re so close. But we can look at the difference between the two:

    If you look closely (click on the graph for a larger, clearer view) you can see that the only difference is a residual annual cycle — and a very small one at that (it never goes beyond plus or minus 0.035 deg.C). Either they’ve used a different baseline period, or they computed anomalies in some other way (sometimes anomalies are computed by fitting a Fourier series, whereas I simply computed monthly averages and defined anomaly as the deviation from the monthly average). But as I say, the difference is tiny and doesn’t amount to a hill of beans. No doubt about it, the AMO is linearly detrended North Atlantic Sea Surface Temperature.

    What could be the cause of changes in N.Atl SST anomaly? Let’s compare the N.Atl SST anomaly to GISS temperature:

    Obviously they’re strongly correlated. Bob Tisdale (and others) simply can’t wrap their brains around the fact that global warming is the cause, not the effect, of much of the changes in N.Atl SST anomaly. Therefore global warming is the cause, not the effect, of much of the variation in the AMO.

    If we’re really interested in how the north Atlantic is oscillating, apart from the global warming forcing, we can at least get a rough idea by simply taking the difference between N.Atl SST anomaly and GISS temperature:

    Of course there are better ways to do this, but it still illustrates the point: that when you look at AMO correctly, the possibility that it’s much of the cause of global warming vanishes. It’s no more sensible than the notion that “eyeballing” is a better choice than multiple regression, an idea which is — how shall we say? — absurd.

    There’s a host of other serious problems with Tisdale’s “analysis” (is that even the right word?). But I won’t bother with all of them. After all, I only have one lifetime.

    About these ads
  • 70 responses to “AMO

    1. The WUWT crowd are no friends of Tisdale. Friends don’t let friends drink and derive.

    2. yay, the trouble with these longer presumed cycles is they’re not necessarily true cycles. What would be the physical mechanism producing that, is a question that needs to be asked. For ENSO there’s the delayed oscillator model, but I’ve never heard anyone propose a plausible one for AMO.

    3. I notice your final graph has a significant negative trend. I suspect that is because land temperatures are rising faster than sea surface temperatures, so that GISS global temperature anomaly rises faster than sea surface temperatures in general, and those of the North Atlantic in particular. Can you show us a graph of the NA SST anomaly minus the global SST anomaly for comparison?

    4. Another eyeball for the misplaced marble file.

    5. If the forced component of north atlantic sst is correctly removed the AMO is just very slightly positive since mid ’90.
      See Ting et al. 2009:

      http://www.ldeo.columbia.edu/res/div/ocp/pub/ting/Ting_etal_2009.pdf

    6. Vaughan Pratt has a nice simple analysis which uses simple smoothing to remove the effect of the AMO. The result is striking:

      http://tinyurl.com/killAMO

      http://thue.stanford.edu/killamo.pdf

      See also here

      http://judithcurry.com/2010/12/05/confidence-in-radiative-transfer-models/#comment-19481

      and here

      http://boole.stanford.edu/dotsigs.html#ClimateChange

      [Response: Sorry, but I don’t buy it. He hasn’t removed the impact of AMO, he’s just removed the fluctuations on a 65-year timescale.

      You can’t just take the deviations from a 65-year smooth and call them the “effect of the AMO.” First, such fluctuations in AMO are the result of the global warming forcing, not the cause of it. Second, to believe that deviations in global temperature on 65-year time scales are somehow the “effect of AMO,” you have to suppose that they are *not* due to an early-20th-century lull in volcanism, early-20th-century increase in solar activity, and huge mid-century increases in anthropogenic sulfate aerosols. But in fact they *are*.

      Global warming has increased temperature, including in the north Atlantic, and we have a perfectly good explanation not only for the cause of that increase. It works. There are many factors at work, but a mysterious “cycle” isn’t one of them.

      The consensus view — that the cause of the up-then-flat-then-up temperature pattern is the laws of physics — I’ll go with that one.]

    7. I think of the emperors new clothes whenever I hear of the AMO.

      1. Flimsy Non-robust definition of the index. The detrending depends strongly on the data period used. add more pre-industrial data points, the average trend will be reduced, and eventually the whole temperature signal is called AMO.
      2. The “oscillation” is at best an interesting hypothesis. However, There is so few “oscillations” in the short series, that it is impossible to distinguish and other low freq variability.

    8. Tisdale is the equivalent of the chartists playing the stock market. It has no more meaning than looking at the stars to tell your future.

      • Rob Honeycutt

        You know, I had that same thought about that whole crowd of folks. Bastardi, D’Aleo, etc. Rather than trying to dig into the physical processes that move the charts they only look at the charts and trying to, I guess you could say, “over-fit” their preferred conclusion. Same as the stock market chart readers.

    9. I dont understand how do you link me into WUWT, I have nothing to do with it, so calling me “WUWT-crowd” just because of my questions is just a straw man argument.

      [Response: You’re hardly the only one who bought into Tisdale’s nonsense.]

      Secondly, you deleted my complete post, why is that? I had some other questions and points aswell, I guess you just dont have the answer do ya?

      [edit]

      [Response: If you argue with a fool …]

    10. I can understand someone accidentally falling into the “correlation implies causation” trap once, but doing it over and over again like Tisdale does just boggles the mind.

      One quote is a gem, though: “The scaling for the Volcano proxy data appears too large, while the scaling for the ENSO proxy appears too small.” You see, Tisdale tried (Excel) multiple regression, and he decided he didn’t like the results. Not because the results failed any sort of test, but because the results implies something he didn’t like – namely, that he is wrong, has always been wrong, and couldn’t be more wrong if he tried (and he tries hard).

      Tisdale actually *thanks* someone for suggesting he use Excel. This does rather raise the question “what was he using before?”

    11. By the way, I offered a link to Bob’s blog, not to WUWT like you do and then calling me “WUWT-crowd”.
      The evidence:

      [edit]

      [Response: Fine, here’s your link to the link. Now go away and let the adults resume our conversation.]

    12. Richard Telford

      I suspect that following Tisdale’s methods, you could take one of the more severe AR4 GGM runs, show that all the modelled global warming could be accounted for as the sum of various model climate indices, and therefore prove to his satisfaction that even in the models CO2 isn’t important.

    13. Tamino,
      I will have some further comments later but what would be the proper way to detrend the data then. The response is not linear no but the changes over the course of the last 130 years are not assumed to be all GHGs either. Remember at least half of the warming over the last 50 years is the IPCC statement. Therefore could it not work to take an exponential fit out of the data or to correlate it with GHG forcing (modelE forcing for example) and remove the signal which correlates most with the GHG forcing increase? I’m just not sure what the best way to deal with the detrending.

      Also when you correlate North Atlantic SSTs to the Global temperature, could that not be proving that North Atlantic SSTs drive a good part of global temperature, just as much as one could make the opposite inference? I mean we all know that the Northern Hemisphere has a larger effect on the Global average than southern (ocean vs land) and that the NH is affected significantly by changes in the North Atlantic. I’m not argueing that the signal isn’t embedded, but I don’t think that subtracting the NASST and GISS is right considering the NASST play a big role in determining GISS. The NASST signal would be embedded within the variables you’re subtracting wouldn’t it?

      [Response: As I said, there are better ways to remove the global warming signal from N.Atl SST anom, this was just illustrative.

      As for the idea that N.Atl SST anom is driving the global warming TREND, do you really not see that it amounts to nothing more than taking temperature change over some part of the globe and calling it the cause of global warming? Do you really not see the logical inconsistency there?

      If you really want to remove the global warming forcing from NASST, you could try detrending it nonlinearly, or as you suggest, estimate the forcing from model calculations, or you could try fitting or subtracting the global SST, a la Trenberth.

      But please, recognize the foolishness in thinking that temperature change in one region is the cause of global warming. Some ideas really are crap — don’t open your mind so much that your brain falls out.]

    14. Tamino,
      certainly I don’t believe that the AMO is “driving” global warming. But I have always held the belief that the oscillatory nature of the AMO can hide and exaggerate the global warming signal. In the case of the cooling in the 70s ish, I think that aerosols were important, volcanism was too but that the AMO being in its negative phase could also play a role. All I was trying to point out is that IF the AMO is a strong phenomenon, then you would expect to see that its effects would be visible in the global temperature record because of the importance of its geographic location.

      That’s it. I’m not trying to say that long term changes in climate are due to the AMO or that the AMO is the main reason we are in such a warm period right now but I do think it is important to acknowledge that the post-1975 warming *could* have a component which is in part due to the AMO. I feel that the recent literature on the subject supports my conclusion but it does not mean that I am not open to the idea that the AMO is not currently contributing at all. I just think that it is really important to scrutinize these things as much as possible. Once again though, it seems last time you understood my comments in a different way then I meant them so I have to reiterate that I know that the AMO is not causing recent global warming.

      [Response: First, I don’t see any evidence that AMO is hiding or exaggerating the global warming trend, even post-1975, and I don’t see that the literature supports that idea.

      Second, the logical inconsistency of using regional warming as the putative cause of global warming must be acknowledged. Those who wish to support the idea must face this head-on, and provide some powerful evidence for the idea to be taken seriously. Because AMO is nothing but linearly detrended regional temperature, period.

      Third, I have to wonder, where’s the physics? Heat is energy, where is the energy coming from? If even a small part of the modern global warming trend is due to warming of the N. Atlantic, what’s warming the N. Atlantic? I don’t see any support for the idea from physics, in fact it seems to me that physics outright contradicts it.

      But hey, I ran some tests anyway. I regressed global temperature on AMO, and of course I get a strong correlation (nobody disputes they’re correlated, but causation can go either way). I also allowed for lag in the impact of AMO. The best-fit lag for AMO was -3 months. Yes, negative — suggesting that global temperature is cause, not effect.

      I also regressed global temperature on AMO and a time trend. This time the best-fit lag was -2 months. Again negative.

      I even detrended both global temperature and AMO nonlinearly (with a “slow” lowess smooth) and repeated the regression, to see whether AMO might at least account for some of the short-term fluctuations (just as ENSO does). Again the best-fit lag was -2 months. Which again suggests that global temperature is partially the cause, even of the short-term fluctuations in AMO, not the other way around.

      I fully support efforts to research the whole idea. Good luck to those who try to find new ways to remove the global warming signal from N.Atl SST, new ways to explore causation, and yes, investigation of the possibility of causation between AMO and global temperature. But so far, I see zero evidence in favor of that idea, and considerable evidence against it. And I also insist that we don’t ignore the question: what’s the physics behind it?

      All I see is denialists exploiting the correlation between AMO and global temperature to deny greenhouse-gas causation. And in Bob Tisdale’s case at least, botching the job while making a mockery of proper analysis.]

      • Tamino,

        If the AMO is linked to how much heat is being transferred from the ocean to the atmosphere in a given year, then it could be a driver, or at least an index of a driver. It couldn’t have much influence on the trend, but it might well influence year-to-year variation. Just a guess.

        [Response: Indeed it could affect the high-frequency fluctuations but not the trend, and I find a link between AMO and short-term variation plausible *a priori*. But my analysis indicates, even that isn’t the case.]

      • Michael Hauber

        Tamino,

        I would make a guess that a best fit for AMO vs global temps showing a 2 month lag from global temps to AMO ties in closely ENSO, with ENSO driving both temps and AMO and global temps responding faster. Of course the mechanism that ENSO drives AMO may be via global temps….

    15. David B. Benson

      Is some portion of the AMO due to causes other than “global warming”? It appears so. From the website for NOAA’s GEophysical Laboratory in Florida, there are proxies extending back several hundreds of years which are, in the historical period, AMO related. The top 1000+ years of a Greenland ice core d18O shows a peak at 57 years using some older power spectral algorithm. These storngly (to me) suggest something other than anthropogenic causation.

      The Geophysical Lab website has a couple of ppts suggesting a relationship between AMO and the changes in MOC rate. I find that physically plausible. There are at least three papers exploring that connection; one of them is
      DelSole, T., M. K. Tippett, and J. Shukla, 2010: A Significant Component of Unforced Multidecadal Variability in the Recent Acceleration of Global Warming. J. Climate, submitted.
      ftp://www.iges.org/pub/delsole/dir_ipcc/dts_science_2010_main.pdf

      In an amateur setting, using a highly simplified cmodel of climate, I obtain a better fit by, in effect, detrending via the AMO:

      http://www.realclimate.org/index.php/archives/2010/10/unforced-variations-3-2/comment-page-5/#comment-189329

      I’m still working on a variation of Tamino’s two box climate model using annualized data. The forcings are the net forcings from GISS. In order of increasing explanation of the variance (1) just forcings, (2) forcings + SOI removal, (3) forcings + SOI & AMO removal.

      • The problem is that AMO is defined as detrended N.Atlantic temperatures. Basically that amounts to saying that everything which is not linear is an oscillation. Tamino shows us that this residual, is a common global signal.
        It may well be that d18o has a spectral peak at the very some frequency. I have no problem accepting that there is some pre-industrial low-freq variability and it is plausible that MOC variations could be the cause. However, that is no proof that all non-linear Anthropocene variability is caused by the same.

        If you want to convince me: 1) present me with a hypothesized physical mechanism causing the oscillations. 2) Show me several homogeneous millenial scale proxies of MOC (from e.g. Atlantic sediment cores) showing correlated multidecadal cycles. 3) Show me that anthropocene MOC variations are not being forced by anthropogenic emissions.

        Finally, I would be careful removing AMO prior to any fits. E.g. You obviously can fit north-atlantic temperatures perfectly with a straight line if you subtract AMO. Global is relatively similar to Atlantic (see Tamino’s post). So, I would not be impressed by a better model fit if you remove the AMO.

        [Response: I’ll add that the presence of a spectral peak (i.e., the implication of genuine periodicity) in AMO is by no means established at all — it’s “period” has been variously quoted from around 40 to 100 years (talk about uncertainty). And the presence of genuine periodicity in d18O signals from Greenland is likewise sketchy at best.

        The whole “periodic” suggestion is dicey, the actual evidence isn’t there. In part its’ misuse or misinterpretation of the word “oscillation.” Sure things fluctuate, but the word oscillation has a connotation of periodicity for which, especially regarding oceanic fluctuations, the evidence for periodicity is either slim or none at all.]

    16. Juakola, you may want to check at Watts’s, someone using the same userid is posting there. Presumably coincidental since you say you have nothing to do with that site. There’s another at Steve Goddard’s page, and a few more; search that username +climate. Perhaps you’re someone else?

      • So what? I follow a number of blogs, including RC, SkS, Tamino, WUWT, Tisdale, Curry, even Goddard. On all sides. Still, that doesn’t make me “WUWT crowd” any less than “Tamino crowd”. That is still just a straw man.

    17. Tamino wrote in the main essay:

      I regularly get comments claiming that ocean cycles are the cause of global warming. They couldn’t be more wrong….

      A while back Atmoz made a somewhat similar argument to yours (I believe particularly with respect to aliasing) regarding the Pacific Decadal Oscillation, another favorite of those that would explain global warming in terms of climate oscillations:

      This implies that the mode of variability known as the PDO has the same spatial and temporal characteristics as the mean global surface temperature anomaly. The PDO doesn’t cause global warming, the PDO is global warming. (Insert all the caveats of PCA; statistical relationship not causal, linear, etc.)

      On the Relationship between the Pacific Decadal Oscillation (PDO) and the Global Average Mean Temperature
      Atmoz, 3 Aug 2008

      http://atmoz.org/blog/2008/08/03/on-the-relationship-between-the-pacific-decadal-oscillation-pdo-and-the-global-average-mean-temperature/

      • David B. Benson

        Same argvument, in effect, as I gave before for the AMO. There are proxies for the PDO extending back over 1000 years which seem to imply it was always there.

        I opine that Pacific MOC rate variation is a principle cause.

        • Seems reasonable. Stray thought — rate variation — could this be primarily a function of the strength of the general hydrological cycle?

        • David B. Benson

          Timothy Chase | January 31, 2011 at 1:36 am — It appears that the big AOGCMs are able to demonstrate some variation in MOC rate. I don’t know enough to assign causes.

    18. In general, doesn’t the fact that ocean heat content and air temperature are both increasing indicate that heat exchange phenomena cannot be a driving factor in warming? How could any oscillation heat both the ocean and the atmosphere at the same time? Does anybody ever try to argue that it has something to do with deep ocean heat content?

    19. You mean anybody claiming something like “Recent massive volcanoes have risen from the ocean floor deep …[” and “… the regions of the Arctic Ocean experiencing thin ice … is the same region that is right over these massive undersea volcanoes just discovered?”
      Of course somebody has. Wackaloons.

    20. <>

      Look I’m not trying to get into a debate over the subject but I do disagree with some aspects of that statement. Firstly from the realclimate website “AMO: A multidecadal pattern of North Atlantic ocean-atmosphere variability… believed to describe some of the observed early 20th century (1920s-1930s) high-latitude Northern Hemisphere warming and some, but not all, of the high-latitude warming observed in the late 20th century.”

      The authors contend that some (but not all) of the high latitude warming can be explained through the AMO. This is further supported by Chylek et al (2009) who state “We consequently propose that the AMO is a major factor affecting inter-decadal variations of Arctic temperature and explaining high value of the Arctic to global temperature trend ratio during the cooling period of 1940–1970.” Chylek et al (2010) make it even more evident with a similar statement, “Recent accelerated warming of the Arctic results from a positive reinforcement of the linear warming trend by the warming phase of the multidecadal climate variability.”

      Using data from Antarctica they identify the bipolar seesaw embedded in the AMO and conclude “The intense Arctic warming since the 1970s arises from an additive combination of the general global warming trend with the warming phase of the multidecadal climate oscillation, while in Antarctica the cooling phase of the multidecadal oscillation opposes the general warming trend leading to essentially no significant Antarctic temperature change since the 1970s.”

      Parker et al (2007) also identify a strong relationship showing some but not all warming being contributed to by the AMO “The regressions for the globe and Northern Hemisphere are 0.05 ± 0.02 C and 0.09 ± 0.02 C respectively, implying peak-to-peak variability of 0.1 and 0.2 C. These variations are far from sufficient to explain the observed magnitude of 0.7 C in global warming since 1901… However, they are sufficient to contribute to the observed variability in the warming rate, particularly of the Northern Hemisphere.”

      The cooling during the mid-century is also referred to by Baines and Folland (2007) as being due to the AMO and they provide considerable evidence for this. This is important because they argue that the AMO + Sulphates caused the mid-century cooling which shows that the AMO can have an effect on Global temperatures. “The most likely candidates appear to be a likely reduction in the northward oceanic heat flux associated with the North Atlantic thermohaline circulation in the 1950s to 1970s, which was nearly in phase with a rapid increase in anthropogenic aerosol emissions during the 1950s and 1960s, particularly over Europe and North America.”

      The question of whether the AMO is forced by Global Warming or not has been also analyzed in the literature and I don’t see adequate evidence to sway me into thinking that the majority of variance in NASSTs can be explained through the global warming signal. Knight et al (2005) discuss the issue and conclude that there is not a significant forced component, “Its presence over many centuries in the model supports the suggestion from observations and proxy data that the AMO is a genuine repeating mode of global scale internal climate variability. This is consistent with analyses showing the lack of a forced AMO signal in the ensemble of 1860–2000 HadCM3 simulations used by Stott et al.”

      Chylek et al (2010) also look into this issue “Although Mann and Emanuel (2006) suggested that the origin of the AMO may be anthropogenic, and indeed recent fluctuations of the AMO are likely to have a near in phase anthropogenic component, a long‐term ice core d18O data from the south‐central Greenland indicates that similar quasi‐periodic oscillations have existed for hundreds of years before the age of industrialization.”

      We should also consider that models suggest it is not being significantly forced “The model generates multidecadal variability in Northern Hemisphere mean temperatures similar in phase and magnitude to detrended observations. The results suggest that variability in the Atlantic is a viable explanation for a portion of the multidecadal variability in the Northern Hemisphere mean temperature record.” Zhang et al (2007).

      Finally Desole et al (2010) tries (and I believe succeeds) to show that the AMO is not being significantly forced by the global warming signal. “We use fingerprinting here to distinguish between forced and internal variability…To the extent that the IMP is unforced (see next paragraph), these results suggest that the AMO represents variability that is dominated by internal dynamics, as suggested in previous studies… On 10-year time scales, variability in trend due to the IMP is relatively large and can easily overwhelm the trend due to the forced component…The warming and cooling of the IMP matches that of the Atlantic Multidecadal Oscillation and is of sufficient amplitude to explain the acceleration in warming during 1977-2008 as compared to 1946-1977, in spite of the forced component increasing at the same rate during these two periods.”

    21. How about we discuss it this way. Think of the MOC as a heat pump. When its rate of pumping heat is increased it pumps more heat into the Northern Hemisphere, when its rate is decreased more heat is left in the southern hemisphere. So why would the same amount of heat being redistributed cause a significant change in temperatures on the hemispheric scale? Simple, for two reasons. One is that the Northern Hemisphere has more landmass to be significantly affected by the change in SSTs. Secondly, the enhanced North Atlantic warming can affect the sea ice thereby reducing ice cover and creating other feedbacks that enhance the warmth. You don’t need an increase in the amount of physical energy being transported between the hemispheres to see a change in hemispheric temperature patterns.

      As 2010 can demonstrate, the capacity for warming the Arctic and North Atlantic Basin can vastly overwhelm the capacity for cooling in many other locations. Thereby a year where pronounced cooling in the continents occurs can be overshadowed by extreme warmth in the North Atlantic Basin making 2010 the warmest on record.

      http://www.skepticalscience.com/Ten-temperature-records-in-a-single-graphic.html

      Now I understand like I mentioned before the issues with linear detrending but there have been multiple attempts at quantifying the AMO and detrending linearly is not the only approach. If I understand correctly EOFs have been used extensively and they also show a similar pattern.

      That being said, I have no intention of getting into a big argument on the subject and I really respect that you have given it an honest effort by trying to determine appropriate lags and so forth. In the end you could easily be proven to be 100% right and I respect the work you do here. Cheers.

      • I’d say that the conclusion of Emanuel and Mann about the nonlinear pattern of N.Hem Atlantic tropical temperature variation, also summarized on RC thus:

        they identify this irregular warming pattern with a combination of greenhouse gas warming influences and late 20th century sulphate aerosol cooling influences (which are especially large during the late boreal summer in the tropical Atlantic). It is therefore likely that the non-linear temporal history of anthropogenic tropical Atlantic warming has masquaraded as the ‘AMO’ in some studies.

        would apply to the entire AMO region, not just the N.Hem tropical Atlantic. Frankly I don’t see how it’s possible that the combination of greenhouse-gas forcing and sulfate aerosol cooling could fail to produce such a pattern — these are real physical forcings.

        And much, if not most, of what you’ve noted in the literature is about the redistribution of heat within the climate system (as one region warms another cools), not the addition of heat to the climate system as a whole. I thought we focused exclusively on global warming, not hemispheric or regional, but much of what you refer to is about N.Hemisphere or even high-latitude warming, not the same thing.

        And as I say, my own analysis contradicts any causal link from AMO to global temperature, but suggests the opposite direction.

        You could be right too. So we’ll agree to disagree.

        • Gavin's Pussycat

          And much, if not most, of what you’ve noted in the literature is about the redistribution of heat within the climate system (as one region warms another cools), not the addition of heat to the climate system as a whole. I thought we focused exclusively on global warming, not hemispheric or regional, but much of what you refer to is about N.Hemisphere or even high-latitude warming, not the same thing.

          This is true. What is also true is, that mean global temperatures as measured refer to a thin layer of air some 2 m above the surface, not necessarily the best way to sample the climate system — i.e., atmosphere and oceans — as a whole.

          I would consider it completely possible that there is an internal mode of variability in the ocean which transfers heat back and forth between the ocean surface — where it impacts the measurements — and deeper layers. Part of the AMO could be something like that. A way to figure that out would be to analyse output from general circulation models looking for the depth signature of these temperature variations. I am not aware of anyone having done this, not even DeSole et al.

    22. I don’t understand… what’s the mathematical procedure to tell cause from effect or mere correlation in a case like this?

      I know AGW is not caused by AMO, otherwise we would observe more backradiation (for one thing). But I’d like to learn a bit more of this mathematical analysis.

      [Response: Establishing causation (rather than just correlation) with purely statistical analysis is always uncertain.

      The most reliable indicator (or should I say contra-indicator) is time lag. If “A” happens after “B,” then it can’t really be the *cause* of “B.” This is the basis for tests like Granger Causality. But although failing the test pretty well rules out causation, passing the Granger causality test doesn’t prove causation — it just makes is plausible.

      Related to the “arrow of time” principle is, that if a theory of causation can make successful predictions, that’s pretty strong evidence of its correctness.

      I’d say, the real test of causation is physics, not mathematics.]

      • Thanks Tamino. I’ll try to improve my skills in statistics to play around a bit with Granger Causality.

        Just to correct a mistake I made in my previous post: if GW were caused by AMO, we woud see more Outgoing Longwave Radiation, not backradiation.

    23. Horatio Algeranon

      When a mother pushes her child on a swing, the swing clearly has something to do with the child’s movements, but the real question is What is causing the swing to move?

      Perhaps the swing is just moving on its own? And perhaps even causing the mother to move? (other than out of the way when the oscillation gets big enough, that is)

      But of course, that is all beside the point, because there are some who are obviously not interested in root causes, but only that AMO is ammo

    24. Tamino:
      Regarding the Mann and Emanuel (2006) paper.

      ” The AMO has recently been proposed to be a construct of aerosol forcing in the North Atlantic and large-scale warming primarily driven by greenhouse gas increases [Mann and Emanuel, 2006]. However, the ocean-atmosphere variations and climatic effects associated with the AMO are quite diverse, including for instance changes in ocean currents and subsurface temperatures [Bjerknes, 1964; Zhang, 2007, 2008; Zhang and Vallis, 2006] as well as impacts outside the Atlantic [Folland et al., 1986; Gray et al., 1997; Enfield et al., 2001; Sutton and Hodson, 2005; Zhang et al., 2007; Dima and Lohmann, 2007]. Aerosol forcing, as assumed by Mann and Emanuel [2006], would not affect the subsurface as rapidly and distinctly as the ocean surface [Zhang, 2007]. [13] More importantly, while SSTs in the tropical North Atlantic from 0 to 30N and in the far north between 40 and 70N can be used as an indicator of the state of the AMO [Goldenberg et al., 2001], the signal of the AMO was originally discovered as near-global in extent [Folland et al., 1986]. Later studies confirmed that SST anomalies extend outside the North Atlantic [Gray et al., 1997; Enfield and Mestaz-Nun˜ez, 1999]. Decomposing tropical North Atlantic SST into a global trend and a residual component that is assumed to contain the AMO signal, as proposed by Mann and Emanuel [2006], is therefore problematic, as part of the AMO signal will be contained in the global component.”
      Grossman and Klotzbach 2009

      [Response: I still don’t buy it. It’s just circular reasoning — whatever “part of the AMO signal will be contained in the global component” wouldn’t be AMO, it would be global. Or should we call it GMO, for “global multidecadal oscillation,” and say that its influence is most pronounced in the N.Atlantic?

      And please let’s not divert attention away from what’s under discussion. Namely, the suggestion that AMO (as *presently defined*) has a causal influence on global temperature. Such phrases as “variations and climatic effects associated with the AMO are quite diverse” is, in this context, a distraction from the real issue. AMO is defined as N.Atllantic SST anomaly — either that has a causal influence on global temperature or not. That it does, is certainly the argument of Tisdale and others like him, which is the *topic* of this post.

      As for the “real” AMO, it’s hard to accept a plethora of diverse speculations when it’s becoming increasingly clear that the definition of AMO is insufficient. The present “N.Atl SST anom” definition is unsatisfactory, and has researchers seeking alternatives — but no agreement yet even on what it is.]

      Finally regarding the redistribution of heat. My point is that the amplitude of changes invoked in the Arctic through GHG AND the AMO is much greater than the amplitude in other regions. Therefore when the positive phase of the AMO contributes to accelerated warming in the North, it causes global average temperature to be higher. The point of the redistribution information presented is that when heat is pumped into (or taken away) from the NH oceans it results in a greater global warming temperature affect than pumping into or removing from the SH. I don’t think it is terribly controversial to state that the NH drives global temperature more than the SH because of the larger amplitude in changes.

      [Response: I still don’t agree. If the AMO accelerates change in the arctic at the expense of changes elsewhere, you’re just robbing Peter to pay Paul, which doesn’t make the apostles (as a group) one bit richer.]

      Thank you for being open to discussion and I do feel that I have learnt some things. These debates are important but I hope that others don’t misconstrue their implications for their own means.

      [Response: I would hope so too, but I expect certain quarters to misconstrue, since it’s their modus operandi.]

      Beyond what I just said I do have a question. One of the references I cited (Chylek et al 2010) (http://www.agu.org/pubs/crossref/2010/2010GL042793.shtml) seem to calculate correlation coefficients on running means (11-year) between Antarctica and the Arctic. I thought running correlations on running means was inappropriate due to autocorrelation?

      [Response: Absolutely yes (this isn’t the least bit in dispute). You can do it, but you must compensate for the autocorrelation and that’s a bit tricky. If Chylek drew conclusions based on correlations of running means, and didn’t compensate for the autocorrelation, then his result is just plain wrong.]

      • [Response: I still don't agree. If the AMO accelerates change in the arctic at the expense of changes elsewhere, you're just robbing Peter to pay Paul, which doesn't make the apostles (as a group) one bit richer.]
        Is there any good reason to expect the effects to cancel out?
        First of all, heat loss is proportional to the fourth power of the temperature, which means that the derivative of heat loss with respect to temperature is proportional to the third power of temperature. As an example, take heat from a very hot area at 300k and emit it from under the Arctic ice in the dead of winter at 240k. The area-averaged temperature increase will be (300/240)^3= 1.95 times as large as the decrease in the temperature decrease in the original area. Decreased greenhouse warming from water vapor will make up some of the difference. Cloud albedo changes will also have large effects in the Earth’s net average surface temperature. In any case, there’s no obvious reason why all these huge changes in the Earth’s average temperature should sum to exactly zero, or that they must be much smaller than the very small CO2 forcing. Unforced climate models show considerable internal fluctuations of average temperatures as well.

        We have a pretty good idea of what the dominant mode of internal variation should be, and appears to be well correlated to deepwater formation areas. The PDO and AMO may be well correlated to deepwater formation, so Tisdale’s hypothesis isn’t unreasonable. This also means that a the delay you see isn’t a causality problem.

        The strong correlation between AMO and global temperatures doesn’t imply a causality direction. What we really want is to identify what part of the current global temperature is forced and what part is internal variation, but looking just at the global average temperature doesn’t provide us with a good way of doing this.

        One thing we could try is to take the time-dependent greenhouse forcing and the anomalous changes to global temperature (after accounting for a linear response to forcing), and to regress global temperatures linearly against them at each point. This would give us 2-d temperature fields linked to both global warming and to global temperature change due to internal variability. We could then decompose the current temperature field in terms of these fields, which would give results for how much of the current global temperature appears to be due to internal variations and how much appears to be due to global warming.

        Obviously, time-dependent response to greenhouse forcing would be problematic, but I think the results would still be highly informative.

        [Response: First: heat loss by radiation (not by other factors) is proportional to the 4th power of temperature, but heat *content* is not.

        Second: one paper (one opinion) doesn’t constitute “We have a pretty good idea of what the dominant mode of internal variation should be.”

        Third: the known natural variations, averaged over decadal or longer time scales, are *small* compared to the global temperature change. Your suggestion of things like cloud variability is no more valid than the “Atlantic Multidecadal Leprechaun.” This whole argument is just misdirection on your part.

        Fourth: your claim that a negative delay isn’t a causality problem is utter nonsense.

        Fifth: studying the time dependence of the response to *greenhouse gas* forcing is a mistake. Study the time dependence of the response to *forcing* period, instead. I understand that some people want to ignore long-term volcanic and solar trends, and anthropogenic aerosols, because it undermines their intention to blame global warming on “natural variation.”

        Detailed study of the expected response of the climate system to total forcing, and its geographical pattern, is a good idea. In fact it’s been done. See the IPCC reports.]

    25. I’ve created a few animations of the NCDC gridded temperature anomaly data, which gives an idea of spatio-temporal temperature changes.

      These are rather large files, therefore I used sendspace :
      NCDC 1880-2010, 12 month moving average, plotted every month:
      File name
      (NCDC_Surface_Land+Ocean_12m_movavg_1880-2010_1m_incr.XviD.mov)
      Download page

      http://www.sendspace.com/file/dytb4u

      and one with more smoothing using 120 month moving average, plotted every 12 months:
      File name
      NCDC_Surface_Land+Ocean_120m_movavg_1880-2010_12m_incr.mov
      Download page

      http://www.sendspace.com/file/jeicm7

      The links to the animations can be found at the bottom of the pages.

      Does anyone have an idea how coverage may affect the global temperature anomaly series? Notably during WOI and WOII coverage is poor for Antarctica and parts of the Pacific Ocean. Part of the temperature increase for the decades around 1940 seems to be related to higher temperatures recorded in the North Atlantic and North Pacific temperatures.

    26. Excellent work tamino. I am with you that the AMO has been given carte blanche to explain many things, for which the case is not strong. This is mostly because of the poorly defined nature of what causes AMO variations and how it is really defined. This is the reason that a large international program is examining the issue, that U.S. CLIVAR: Climate
      Variability and Predictability is a participant in . We are a long way from pinning this one down in terms of mechanisms or impacts. Science is examining the details now and it will be interesting to see the outcome.

    27. I found good correlation between the de-trended Arctic’s temperature and the AMO.

      http://www.vukcevic.talktalk.net/NFC1.htm

      Odd thing is that some of the monthly CETs (Central England temperature) show also high degree of agreement with the AMO, while others do not.

      http://www.vukcevic.talktalk.net/anomalies.htm

    28. I don’t see how such a statistical “analysis” as Tamino’s has truly separated natural variations from AGW. And certainly has not shown cause versus effect. False regressions happen all the time.

      Almost every model that has looked at natural oscillations has detected periodicities very similar to what are observed. The 1990 Nature paper by Wigley et all “Natural variability of climate systems in the detection of the greenhouse effect” wrote fluctuations in global mean temperature of up to 0.4C temperature changes over a period of thirty years or more” He just concluded it didn’t account for all the heating that at the time was about 0.65C/century

      Hansen’s 1988 “Global Climate Changes as Forecast by GISS Three-Dimensional Mode” may have approximated the average temperature change with reasonable approximation with scenario B, but the model showed warming over all the continents and oceans. His model was forced by CO2 and failed to show any oscillations that have now been well observed as well as regional ocean coolings that have been well observed. In that respect the CO2 model failed, suggesting the warming trend was possibly a lucky coincidence. And the main reason it failed to detect regional ocean cooling was because Hansen held ocean heat advection constant, which in real life never happens.

      I would suggest that Tamino should spend less time with ad hominem attacks vs Tisdale and discuss the mechanisms that would lead him to suggest AGW causes oscillations. Evidence of oscillations like the NAO, PDO, etc have been detected centuries before the forcing of CO2 can be suggested.

      [Response: Please get your facts right before commenting. I did not suggest that AGW causes oscillations. I suggested that most of the change in N. Atlantic SST over the last century was a global-warming signal, not an internal oscillation. And there’s a helluva lot more support for that idea in the literature than the meager references you’ve given for oscillations causing global warming. What’s most clear is that the current definition of AMO is inadequate.

      Not to mention — oh what the hell, let’s mention it — the idea of oscillations causing global warming (rather than transport of heat from one region to another) needs a serious dose of reality therapy regarding the laws of physics.

      You’re the one indulging in ad hominem by implying my criticism of Tisdale fell into that category. Ridiculing his claim that “eyeballing” is a better idea than numerical analysis, and calling that idea absurd, is plainly true. Perhaps you agree with him.]

      • Jimmy, no offense, but you’re kind of lost. What do oscillations do? They go up and down, up and down. What is the temparature doin? I’s going up and up. See the difference?

    29. Ray Ladbury | February 4, 2011 at 2:18 am | Reply
      Jimmy, no offense, but you’re kind of lost. What do oscillations do? They go up and down, up and down. What is the temparature doin? I’s going up and up. See the difference?

      Hey Raymond,

      Always a pleasure to receive your insights. Before we discuss this further I suggest we create a time frame. You can see temperatures go up and down by reading Taminos glacial interglacial. You can download the GISP2 data and see the trend is down but punctuated with warmings. I can create a century long temperature trend that matches both the observed temperature trends and the dips by simply adding sunspots trends with PDO trends (PDO values are scaled) and then using a ~22 year running average to simulate ocean lag times. No CO2 and no sulfates required. I suspect Raymond you are either too narrowly focused when you suggest temps only go up and up or you are projecting too much into the future.

      [Response: You forgot to include the Atlantic Multidecadal Leprechaun. B-bye.]

      • Jim,
        I’m relying on physics and looking at the evidence. There is more to climate change than matching the temperature curve. Why not try science?

    30. Tamino why did you remove my reply to your reply? Why shy away from a good debate?

      [Response: I don’t think you want to debate at all, you just want to argue. And judging from your response, you’ve bought into some real crackpot theories. You’re welcome to your opinions, don’t expect me to waste my time arguing with you about ‘em.]

    31. Wouldn’t a better method to remove the climate signal from AMO, be to do a quadratic (ax^2+bx+c) regression against the raw SST anomolies and subtract the result from the anomolies? When I do this the resulting series shows that the current peak is less than the mid 20th century peak. The series also has a flat trend.

      The rationale for this approach is that the ln(accumulated co2) forcing curve can be well fitted quadratically, meaning that the rate of change is linear (i.e. it’s derivative). If this forcing is being realized at a rate that can be reasonably estimated as linearly increasing, then a quadratic fit (i.e. it’s integral) might be a more appropriate method for removing the climate signal from the SST.

      I guess this all depends on how a given forcing is realized. Perhaps that would be a good topic for another post. You could give an example of a climate system that was in equilibrium and then there is an immediate cut in co2 concentration that will ultimately result in a 1C decrease in temperature. What is the equation for the curve representing the realization of that 1C decrease over time?

      [Response: But the CO2 signal isn’t the entire global warming signal. That includes volcanic forcing (including, in fact in this context *especially*, the long-term effect due to thermal inertia of the oceans), anthropogenic sulfate emissions, solar changes, etc. All those factors can be expected to influence N.Atl SST (just as they do global temperature), none of them are the AMO. Truly to remove the external forcing from N.Atl SST, it seems to me, requires more than just a quadratic fit.]

      • [edit]

        [Response: This is not your blog, and you are not welcome to post a 2400-word essay masquerading as a comment. I’ve said I don’t want to argue with you about it, now let’s be even clearer: based on your statements (including, perhaps especially, your ramblings in this “comment”) I think you have absolutely nothing of value to add to the discussion. Furthermore, your comments about “period” and “oscillation” confirm my opinion that discussing it with you is a waste of time.

        Feel free to post on your own blog how brilliant is your orgy of indulgent curve-fitting, and how mean, stupid, closed-minded, and unscientific I am. I’ll respond directly to your ideas when they pass peer review and appear in GRL. I won’t be holding my breath.

        I stated in response to Ron Broberg that this post was not about him. It’s about you.]

        • This is not your blog, and you are not welcome to post a 2400-word essay masquerading as a comment.

          My apologies for whatever it was that you deemed inappropriate.

          Feel free to post on your own blog

          Ok.

    32. Tisdale has a response with an eyeballed ( and wrong) North Atlantic area and a wrong formula to compute the global trend without the North Atlantic.
      Using oiv2 sst the computed area 0/70n 80w/0 is 11.24% of the global sst(not 15%), also assuming G is the global trend A the North Atlantic trend and GA the global trend without the North Atlantic then:
      G=0.1124*A+0.8876*GA—>GA=(G-0.1124*A)/0.8876 while he compute the trend as GA=G-0.15*A

      Both those corrections reduce the difference between G and GA, the North Atlantic has warmed faster than the global average since 1982 but not that much….

    33. “But the CO2 signal isn’t the entire global warming signal.”

      Sure, but isn’t there a dimishing return when including the other forcings. Isn’t co2 the primary forcing and isn’t it’s rate of change the best known? Would modelling the other forcings materially effect the results of this excercise? If you are looking for a method that is a little less dirty, but just as quick, it doesn’t seem unreasonable.

      As for thermal inertia, I agree it is important. I was speculating that since the rate of change of co2 forcing is increasing linearly, that the rate of change in temperature might also be reasonably estimated as increasing linearly. I’m not suggesting that these two lines are the same. I don’t know if they are converging, diverging, or parallel. I think that would make a great topic for another post.

    34. Apropos wiggles, you might enjoy the puzzle presented at

      http://johncarlosbaez.wordpress.com/2011/02/04/carbon-dioxide-puzzles/

      “… The puzzle is to explain this graph. Why is the red curve roughly 40% lower than the blue one? Why is the red curve so jagged?

      Of course, a lot of research has already been done on these issues. There are a lot of subtleties! So if you like, think of our puzzle as an invitation to read the existing literature and tell us how well it does at explaining this graph. You might start here [link in original], and then read the references, and then keep digging.

      But first, let me explain exactly how Renato Iturriaga created this graph! If he’s making a mistake, maybe you can catch it….”

      — definitely not the place for instant answers.

      [Response: There’s some speculation here.]

    35. Thanks Tamino, I’d forgotten that one!
      Scholar finds that 2007 paper cited only once:
      2010 Environ. Res. Lett. 5 034007
      (http://iopscience.iop.org/1748-9326/5/3/034007)
      Climate control of terrestrial carbon exchange across biomes and continents

      (Those interested can find full text with a bit of hunting)

    36. @ David B. Benson | February 4, 2011 at 11:29 pm | Reply

      Dave… I believe you misunderstood me. The accumulation of ln(co2) is certainly not linear. But the rate of change (i.e. the derivative) is.

      What I find interesting is… What is the relationship between the rate of change of the initial forcing and the ultimate forcing? Can these two lines (curves) can be approximated as linear equations? Do these lines converge or diverge?

      • David B. Benson

        AJ | February 5, 2011 at 7:56 am — Always David, never Dave; that’s someone else around here.

        You obviously didn’t bother to study the provided link; the first differences are all over the place.

    37. AJ, be careful here. Since all mathematics is ultimately addition or subtraction–looking at derivatives involves subtracting nearly equal quantities and then dividing (repeated subtraction) by another small quantity. Doing so with coarse measurements, you can wind up with very large errors. Know your error bars.

      • Yes Ray, error bars are important. The correlation, however, is very strong. IIRC, r^2=0.999, which is near perfect. Will the coefficients will hold up in 40 years or so? My guess is that they won’t. Uncertainty seems to be the rule of the day.

    38. Tamino (and others) – I know you do not agree with Vaughan Pratt on the AMO. I think Vaughan is an exceedingly interesting person, but I do not agree with him either and find your argument on the AMO compelling.

      This Vaughan Pratt graph was not linked in the earlier comment.

      I think I actually contributed “WW2 aerosols” to his graph. I got him thinking about it. Later Hank Roberts linked to a site that listed a paper about atmospheric nuclear testing having had a cooling effect.

      The purple line is some sort of marriage of the growth in anthropogenic atmospheric CO2 and the AMO.

      Could it actually be a marriage between brightening and dimming and the AMO?

    39. There seems to be a new paper on AMO in Nature communications: “Tracking the Atlantic Multidecadal Oscillation through the last 8,000 years” http://www.nature.com/ncomms/journal/v2/n2/full/ncomms1186.html . Thery try to find preindustrial proxies for AMO.

      Any comments?

      PS: Sorry for posting in this thread. It is not at “denialist” paper.

    40. Willard drew my attention to the comments in this thread about some of my graphs, in particular killamo. I’m particularly mystified by JCH’s Tamino (and others) – I know you do not agree with Vaughan Pratt on the AMO. I think Vaughan is an exceedingly interesting person, but I do not agree with him either and find your argument on the AMO compelling.

      The graphs have no supporting assertions by me, they are simply the result of computations, the details of which I can provide if they’re not obvious. So I’d be interested to know what it is that you disagree with, so I can tell whether I disagree with it too. (Note that by “AMO” in the context of global temperature I mean “AMO correlate,” not the AMO itself which of course is local to the Atlantic. This is made explicit here for example.)

      These additional graphs likewise have no substantive supporting assertions, deliberately so in that case since it seems likely they would only draw flak like the above. I would be very interested in any reactions people might have to the fifth graph on that page.

      Regarding the following,

      Sorry, but I don’t buy it. He hasn’t removed the impact of AMO, he’s just removed the fluctuations on a 65-year timescale. You can’t just take the deviations from a 65-year smooth and call them the “effect of the AMO.” First, such fluctuations in AMO are the result of the global warming forcing, not the cause of it. Second, to believe that deviations in global temperature on 65-year time scales are somehow the “effect of AMO,” you have to suppose that they are *not* due to an early-20th-century lull in volcanism, early-20th-century increase in solar activity, and huge mid-century increases in anthropogenic sulfate aerosols. But in fact they *are*. Global warming has increased temperature, including in the north Atlantic, and we have a perfectly good explanation not only for the cause of that increase. It works. There are many factors at work, but a mysterious “cycle” isn’t one of them. The consensus view — that the cause of the up-then-flat-then-up temperature pattern is the laws of physics — I’ll go with that one.

      Were Tamino using terms the same way I do, I would take strong exception to a number of things. However context suggests that he must be using some terms differently from me. We would therefore be talking at cross purposes were I to start raising objections before we had agreed on terminology. For example he must mean something very different by “fluctuations in AMO” for me to make any sense of the statement that the fluctuations “are the result of the global warming forcing” — my definition of “AMO correlate” makes this analytically impossible. Also I don’t understand the phrase “mysterious cycle” in the context of a temperature curve that contains 16 radians worth of a very obvious cycle. And we evidently have a different definition of “the laws of physics,” unless I’ve misunderstood the consensus explanation of the flatness which infers the extent of aerosol cooling from the temperature rather than vice versa.

      [Response: When you use the phrase “a temperature curve that contains 16 radians worth of a very obvious cycle,” then I have very good reason to believe you don’t understand the statistics of period analysis, you’ve just fallen into the all-too-common trap of thinking that it’s “obvious.” When you say things like “the consensus explanation of the flatness which infers the extent of aerosol cooling from the temperature rather than vice versa” then I realize you don’t understand the evidence for aerosol cooling or for its quantitative estimation, which is from the impact of volcanic eruptions, not “inferred” from the mid-century temperature stalling.

      I think your theories have no merit. I respect your civility. But no, I don’t wish to argue with you about it.]

      • David B. Benson

        Vaughan — The usually definied AMO contains a trap; I’ve fallen for it. The problem with

        http://www.esrl.noaa.gov/psd/data/correlation/amon.us.long.data

        is that only a linear trend is removed; what remains contains both forced and unforced variations. The paper cited and linked in the just prior comment, MS | February 13, 2011 at 1:12 pm , attempts to produce a ‘modern AMO’ which only includes (hopefully) the unforced component. In some work with this definition, very little of the variance over the instrumental record is explained by the ‘modern AMO’, so little that via the Akaike Information Criterion a model with it is strongly equivalent to that model leaving it out.

        Also in earlier comments (including mine) there are other papers which attempt to isolate the unforced variation component.

        • David: good to run into you again. On the off chance that Tamino will let me continue to comment on his blog, let me say that I agree with you. I wish I could say more than that, let’s pursue this offline.

          [Response: No, I’m not interested in hosting your comments. Those who wish to communicate may visit Vaughan Pratt’s web site. Those curious about his lengthy comment will find it here.]

      • Vaughan,
        OK, I’ll play. I apologize to the regulars who’ve seen this example before. Consider the following series:
        1-2
        2-7
        3-1
        4-8
        5-2
        6-8
        7-1
        8-8
        9-2
        10-8

        It looks periodic, doesn’t it. You’ve got what looks like a periodic oscillation and pretty consistent amplitudes. And yet, I know that it is not truly periodic, and I will predict that the 11th mumber in the series will not be 1 or 2, but 4. How do I know this? Because the y-values of the pairs are the digits of the base of napierian logarithms, e, in order.

        Please realize that Tamino is busy. He run this blog AND he has a day job. He can’t take time with every poster to show them where they’ve gone wrong. In your case, though the issue is that you can’t eyeball periodicity with just a wee bit of data. It doesn’t work.

        • OK, I’ll play

          Unfortunately it appears you’re allowed to play but not me. I’ll be happy to address your point elsewhere.

          [Response: As I said, those who wish to communicate with Dr. Pratt may contact him through his website.]

    41. I came to this website when it was drawn to my attention that people were criticizing me behind my back for reasons I could not discern. I tried to start up a dialog aimed at understanding what the complaints were, but instead of getting a straight answer I get the following.

      I have very good reason to believe you don’t understand the statistics of period analysis … you’ve just fallen into the all-too-common trap of thinking that it’s “obvious” … you don’t understand the evidence for aerosol cooling … I think your theories have no merit.

      Whoa, where to begin?

      On other blogs, one way to identify the climate skeptics is that they’re the ones who talk in that dismissively pseudoscientific way. I guess it doesn’t work on this blog.

      Maybe someone else on this blog can explain to me what I said that got Tamino’s back up in this way.

      I don’t wish to argue with you about it.

      I’ve heard that many times from climate skeptics, but even then never after just a single exchange. This is the first time I’ve heard it from a non-skeptic. Usually it’s the non-skeptics that are the patient ones.

      I am unable to take seriously any attack on my presumed views when the attackers refuse to even tell me what they believe those views to be. The alternative is to be cast in the role of Kafka’s Joseph K., who was never told what he was charged with. Without knowing the basis for the original criticisms I can’t even defend myself from them on some other blog.

      I have no idea what I said to deserve such an uncivil welcome on this blog, but it’s not my idea of how science is normally conducted.

      [Response: I said that I have good reason to believe that you don’t understand the statistics of period analysis. That’s because you argued that the so-called cycle is “obvious” when there’s no analytical justification for such a claim — you just think it “looks” obvious. And in fact I’ve analyzed it in depth (period analysis is my area of expertise), and the claim cannot be justified on the basis of the data. This isn’t a criticism of you personally, but of the claim that you made.

      I said you don’t understand the evidence for aerosol cooling because you attribute it to “infers the extent of aerosol cooling from the temperature rather than vice versa.” This reveals an ignorance of the literature, otherwise you’d know that the extent of aerosol cooling is estimated from the measured aerosol optical depth due to volcanic eruptions and their consequent impact on global temperature, and estimates of aerosol emissions during the 20th century. Again, I made no criticism of you personally, but of the claim you made.

      And now you are the one who decides to “get your back up.” You start by accusing me of criticizing you “behind your back” as though I were whispering in the hallway to some school chum, trying to conceal what I was saying. This is wrong, offensive, and frankly, rather childish. Then you play the part of the wounded, when I only criticized your ideas (and with very good reason) but you are the one who’s decided to make it personal. Your silly remark about “uncivil welcome” and “not how science is normally conducted” completes your delusion of martyrdom.

      All you’ve done is an exercise in contortional curve-fitting, while ignoring (and apparently denying) the physics (e.g., aerosol cooling which is very real). And when I won’t buy into your theory, you get all huffy and apparently can’t find anything more “scientific” than to accuse me of exactly what you yourself have done — make it personal rather than scientific.

      And that is why I don’t care to argue with you about it. If you think your model is so good, publish it in the peer-reviewed literature. That’s how science is normally conducted.]

    42. Tamino,
      was just sent this paper by a colleague and thought you might find it interesting:

      “Here, we show that distinct, ~55- to 70-year oscillations characterized the North Atlantic ocean-atmosphere variability over the past 8,000 years. We test and reject the hypothesis that this climate oscillation was directly forced by periodic changes in solar activity. We therefore conjecture that a quasi-persistent ~55- to 70-year AMO, linked to internal ocean-atmosphere variability, existed during large parts of the Holocene. Our analyses further suggest that the coupling from the AMO to regional climate conditions was modulated by orbitally induced shifts in large-scale ocean-atmosphere circulation.”

      “Moreover, the AMO shifted into its warm phase in the 1990s (Fig. 1), which may have accentuated global warming in this period. A return from a warm to a cold AMO phase could temporarily mask the effects of anthropogenic global warming47, and thus lead to possible underestimation of future warming if the variability of the AMO is not taken into account.”

      Tracking the Atlantic Multidecadal Oscillation through the last 8,000 years

      Kundsen et al 2011
      Nature Communications

      [Response: Fascinating paper. I do have some misgivings about their statistics, but I need to do a lot more analysis before forming a solid opinion. If my initial suspicions turn out to be true, I’ll contact the authors to inquire further.]

    43. Ray Ladbury | February 27, 2011 at 10:31 pm | Reply

      “Robert, And 2010 was the warmest year on record.
      Next!”

      I agree it was. Hence why my post on skeptical science comes to the same conclusion when all truly global datasets (NCEP, GISS, ERA) are looked at

      http://www.skepticalscience.com/Monckton-Myth-2-Temperature-records-trends-El-Nino.html