Correlation Between Emissions and Warming in the Central England Temperature Series



This article by Jamal Munshi challenges the whole theory of AGW in the most radical way imaginable.

From the Abstract:

A comprehensive detrended correlation analysis of the daily mean Central England Temperature (CET) series for each calendar month against fossil fuel emissions for the 245-year study period 1772-2016 is presented. Time scales of 10, 20, and 30 years were tried each at four different time spans ranging from 60 to 245 years at all possible locations within the overall study period. The results do not show a relationship between emissions and warming that can be interpreted in terms of the theory of anthropogenic global warming and climate change (AGW). The finding is inconsistent with the proposition that warming in the CET data can be related to emissions.

It requires careful reading by competent people. I look forward to your comments.


  1. Thanks Jamal for your several analyses on this subject. I appreciate the statistical rigor you bring, such as the sliding windows, and the use of monthly trends to keep the data granularity otherwise lost in annual averaging.

    I have done some simpler comparisons that also show weak and inconsistent correlations between temperatures and fossil fuels. I chose to stick with energy consumption statistics because I became aware of the considerable uncertainty that is added when GWP (Global Warming Potential) is estimated on top of the energy statistics.

    For example, this document details the uncertainty in measuring emissions from fuel consumption.

    Click to access Quantitative%20Uncertainty%20Guidance.pdf

    Buried in the notes of the above document:
    “Note that the uncertainty of the global warming potential (GWP) for the six GHG Protocol gasses is assumed to be ± 35% for the 90% confidence interval (see Section 7.2)”.

    A recent effort of mine was:≠-global-warming/

    Liked by 1 person

  2. I look forward to reading the paper, but I fear it may increase the threat to the CET. As far as I know, it has not yet been interfered with for ‘the cause’ in the way that many other inconvenient temperature series have been. But even now some graduate student in the Hadley Centre could have been charged with bringing the English series in line with reality as they imagine it. [XMetMan has shown what can be done with the series depending on what you wish the casual passerby to conclude:

    Just as the pesky Medieval Warm Period & the LIA were such a nuisance that they had to be got rid of, and the general lack of cooperation of modern temperatures has been an embarrassment being actively addressed sometimes by cooling the past, sometimes warming the present, sometimes both, surely somewhere some research effort should be looking at reforming the CET to make it a bit less troublesome. Here are some examples of the nuisances it has enabled:

    1. Remember when Lubos went through the CET, computing heating and cooling rates over various timescales. and finding nothing extraordinary about recent decades? [see ].

    2. Or look at the fun to be had at NotALotOfPeopleLnowThat with that same series? [see

    3. Or when the IPCC relied on it as a surrogate for a chunk of global mean temperature back before the MBH Stick fell into their hands? [See detailed discussion here:

    All in all, it is time the CET was put in its place by the authorities.

    Meanwhile, I look forward to reading that paper tomorrow. Thanks Geoff, it could become a collector’s item.

    Liked by 3 people

  3. Thanks Ron for pointing out that Jamal Munshi’s work on this subject didn’t start with this article. Other references are in his bibliography. His criticism of the use of data for cumulative instead of year by year emissions in previous studies seems to this non-statistician to raise important questions about the statistical competence of the whole climate science community. I’d be very grateful if the mathematically competent here would spell out what this means to the rest of us. My very blurred memory of elementary statistics tells me that trawling cumulative totals for correlations bodges the issue, but I’d appreciate a clearer exposition.

    One attack on Munshi’s paper that’s sure to be made is that CET is just one spot on the immense surface of the globe. It’s tremendously important to defend this data source, with a continuous 275 year history of data collected by scientists, against those who prefer ten thousand (or a thousand, depending on their mood) sources of data collected from time to time by any old bod at some airstrip or firestation in the sticks, as long as they fill a hole in their grid.

    As always, I look at my ancient experience in market research. Being aware that – say – attitudes to chocolate biscuits, unlike gases in the atmosphere, are not well mixed, I’d do some interviews in the south and some in the north, with a bit of urban and rural thrown in. What I wouldn’t do is divide the country into 100km-wide squares and try and find a respondent in each square, using statistical infilling to provide a perfect map of chocolate-biscuit-desire.

    I’d like to hear more on Munshi’s statistical methodology. It looks to this non-scientist hugely … scientific, with confidence levels that the IPCC can only dream about.

    Liked by 2 people

  4. Jamal’s work is great, I’ve been following him for a while. His Ozone studies are worth a look too.
    Regarding CET, he’s right, it’s nowt to do with CO2.

    However, there is a correlation worth looking at between CET and 10Be and 14C
    It’s the sun. Probably.

    Liked by 2 people

  5. It should come as little surprise that CET is not highly correlated to CO2 emissions but is highly correlated to proxies for solar activity (as above). Many scientific studies link solar variability to changes in NAO, which in turn directly affect weather and climate in northern Europe especially, with winters showing a more marked sensitivity. This influence may operate over decadal, multi-decadal and centennial time scales. I fully expect the standardized NAO winter index to start falling fairly rapidly now that the 2014/16 Super El Nino has faded and solar activity is historically very weak. It’s a matter of debate whether NAO is directly forced by solar variability or whether the influence operates via a teleconnection to ENSO (or a combination of both even). There is evidence that during the LIA, a preponderance of El Ninos occurred in the Pacific, which at first would appear to be counter-intuitive (as El Ninos temporarily warm the surface). But El Ninos also dissipate huge amounts of solar energy accumulated in the Pacific Ocean (pushing it towards the Poles), which might suggest a mechanism (among others) which contribute to surface cooling over decades when solar activity is weak. Intriguingly, models are suggesting that 2017 may be another El Nino year. Meanwhile, here in England, it’s unusually cold right now and NAO is forecast to go strongly negative in the first week or so of May, maintaining the very cool feel to Spring. We escaped the ravages of a strongly negative NAO in winter 2016/17. Like in 2009/10, we may not be so lucky in 2017/18. Better get the de-icers ready for the windmills!

    Liked by 2 people

  6. I’ve just been updating my chart of the breakup of the ice in Nenana. It mirrors the warming we think we’ve had and not the story GISS tell. Warming out of the LIA, till the 40s. cooling from then till the end of the 70s, warming in the 80s and 90s but no faster than the past and a plateau from 1998.

    I have wondered if someone was sabotaging the ice because for months someone has been running up and down the same line.

    I can’t say that CO2 has no effect but it can’t be massive. How long before someone influential admits it?

    Liked by 1 person

  7. Well, after some struggle with the Academia site, I managed to download the paper and give it a quick read just now. First reactions are favourable. The work does involve what is often disparagingly referred to as ‘data dredging’ – going through data sets and computations this way and that in the hope of finding statistically significant results. But the charge is not valid here since, apart from some isolated months showing stat sig correlations which are plausibly rejected as spurious, no such results were found. The rabbit that the author pulls out of the hat after all the work is no rabbit at all. He found nothing convincing to link emission variation with temperature variation.
    1. The study is not comparing ambient CO2 levels with temperatures, merely estimates of CO2 emissions
    2. The author uses Excel for statistical computations. This is not outrageous but seems a bit odd for a professional statistician to do, given the software resources available to him.
    3. A quibble about claiming that rejecting Ho implies that a correlation exists (page 4). I would not put it so strongly. Rejecting Ho merely means, in my view, that some evidence has been found that an underlying correlation relationship may exist. The author does use the magic word ‘may’ later on in this context so I am at peace here.

    Now I’ve just been out in the sun for a few hours before reading through the paper, so I’ll stop this here for the time being (currently on the east coast of the USA). Hoping to get time for another look at the paper tomorrow.


  8. Tallbloke, looks as if you a little “divergence problem” post-1970 in your 2nd graph. And there’s plenty of anti-phase activity and even CET preceding solar. Looks like England controls the Sun at times.

    Generally, I don’t see the point of the study. Is he looking for a link between emissions (global or local?) for a year or period and temperature for England in that period? That would be stupid, so it can’t be that, can it? So what is the aim?


  9. LEN MARTINEZ (30 Apr 17 at 3:24 pm)

    I don’t see the point of the study. Is he looking for a link between emissions (global or local?) for a year or period and temperature for England in that period? That would be stupid…

    I didn’t hear that objection made when scientists claimed to detect a link between emissions (global or local?) and the width of rings in half dead trees on the snowline where no-one lives in Siberia or Colorado.

    The reason for using temperature in England is because it’s the only place on earth where temperature was recorded at that time. Before that we have only proxies like the date of grape harvests in Europe, the first cherry blossom in Japan, etc. They suffer, like the CET, from the fact that they are geographically limited to the sorts of places where civilisation had advanced sufficiently for people to be interested in that sort of thing. They are also places where climate matters, because people live there, specifically people whose civilisation is sufficiently advanced that they can examine scientifically complex questions like global climate change and if necessary do something about it.

    If I were a scientist worried about global warming, realising that the past is the only guide we have to the future, the first thing I’d do is look at the oldest and most scientifically reliable sources for past temperatures and base my science on them. That’s what Professor Munshi has done, as far as I understand his paper. He’s chosen the most favourable conditions for a coorrelation between CO2 emissions and temperature (a 10 or 30 year time lag) and found no correlation, ever, over any time span.

    Not finding something is not the same as finding that it doesn’t exist (viz. unicorns.) So what the scientific establishment will do now is examine Munshi’s paper thoroughly, conduct similar examinations of other long temperature series (Paris, Antrim, Prague..) and come back with a rebuttal.

    Won’t they?

    No they won’t, because the scientific establishment is an establishment erected specifically as a barrage against science, co-opted by politicians in search of the public trust which only scientists (and doctors and hairdressers – see the latest MORI Trust survey) obtain.

    You think it’s stupid to look for a correlation between the longest scientifically controlled temperature series and CO2 emissions. OK. Where would you look for evidence of manmade global warming?

    Liked by 1 person

  10. I agree with your remarks, Geoff, and not least the sad reality that we cannot realistically hope for a reasoned response from the climate science establishment. Their first instinct will more likely be to look for ways to attack the standing of Munshi.

    But I do hope for more studies like his – using actual temperature records rather than the confections of ‘global mean temperature’ currently in the hands of that establishment. That, plus looking at possible links with monthly, seasonal, and annual variations in estimations of ambient CO2 on global and regional scales, and for that matter, with estimated global and regional emissions.

    One of the things I would like to return to in the work of Murray Salby is his analysis of CO2 variations (ambient, as well as emissions, man-made and natural), and global mean (I think) temperature and their integrals or first derivatives wrt time. That applied to estimates of regional mean temperatures (which ought to be less vulnerable to invention/tampering that global means in regions with good climate records) would be good to see. I also recall interesting work by Jonathan Lowe on links between cloud cover and temperature records in Australia – another example using real observations rather than gross global artefacts. As you might expect, cloud cover variation explains a lot of surface temperature variation. How much of the residuals would be explained by CO2 variation?

    But as it is, I struggle to find time to get back to Munshi’s recent paper. But I will, and hopefully some of his earlier work too. But whether I read it or not scarcely matters to anyone but me, but the mere existence of such work is a source of good cheer all by itself!

    Liked by 2 people

  11. My eyes glaze over when I try to read the paper, so it may be that I am the stupid here. But as i gather (I hesitate to say understand), he is looking for a correlation between short term temperature and variations in annual emissions. For reference, a 1% change in CO2 concentration causes a change in forcing of 5.35ln(1.01) = 0.05W/m2 and over the solar cycle, total solar irradiation varies by about 1.3W/m2, or more than 20 times bigger. Moreover, if he is analyzing variations in annual emissions, these will be hugely less than the 1% mentioned above. And of course weather variations from day to day or year to year are huge in comparison with any changes caused by the solar cycle.

    So the question to ask first is whether the solar cycle is visible in the CET records. If it is strikingly present, then looking for something several orders of magnitude smaller might be worth a try but unlikely to succeed. If it isn’t present, then looking for something several orders of magnitude smaller really is just stupid.


  12. I would like to learn more about the detrending used by Munshi. It wasn’t clear to me on my, still too fast, second reading, whether he detrended each time section at a time or did one grand detrending for all the data over the the whole time period. Correlations have been found in the original data, as Munshi notes by giving several references to other studies. Detrending is usually done to prevent correlation caused by other factors e.g. in economic times series, inflation would increase the prices of two items whose time histories were to be compared, and could by itself produce high correlation. Detrending the data would be done in the hope of clarifying a possible relationship between the prices of the two items.

    Now we are confident that manmade CO2 emissions have increased, and although these annual emissions are a small fraction (even today only a few percent) of the total emissions from all sources, the notion is that the manmade emissions have caused or largely caused the overall rise in ambient levels because natural sinks for CO2 have not been sufficient to deal with them. This by itself is an interesting area for study since there are great uncertainties present – in the 1970s as I recall off the top of my head the hunt was on for the ‘missing sink’ since known sources and sinks were in such imbalance that they could not explain why ambient CO2 levels were so low. I don’t know if the ‘missing sink’ has since been found, or whether existing ones have merely been fudged up to show a better balance.

    Len asks about the solar cycle, and of course it produces the obvious seasonal variation in the temperature record, and is also associated with seasonal variation in natural and manmade emissions. It would seem highly desirable to remove this seasonal variation from the data sets before hunting for correlations. I guess this must have been done by someone somewhere. Does anyone reading this know? I think, but ought to check, that Munshi’s analysis that we are looking at here used temperatures averaged over decades at a time, and hence the seasonal variations would be largely obliterated.

    Anyway, what we know is that Munshi found no non-dismissable correlations between manmade CO2 emissions (as estimated by fossil-fuel consumption estimates) using his detrended method and that seems a worthwhile, small piece to add to the might jigsaw of our understanding of the climate system, a jigsaw which one might suppose is still mostly open space on a metaphorical table top. Using climate data that are actual observations, or constructed from a modest number of them (as in the CET) seems an important way to go given the peculiarly artificial nature of ‘global mean temperature’.

    When my current travels, and subsequent jetlag which I know I will suffer quite badly, are over, I hope to have a chance to study Munshi’s work more carefully. Before I’m off again in June! But maybe by then, others will have clarified it all. Might someone persuade Munshi himself him to comment here and sort us all out?

    Liked by 1 person

  13. Clive Best has published a useful couple of plots from the CET data: January and July temperatures:

    As he notes, merely eyeballing the plots leads to this sort of conclusion: ‘Is there really any evidence for anthropogenic global warming in CET? I think you would be hard pushed on the basis of this data to claim that CO2 alone is responsible for recent trends.’

    So, efforts using various statistical methods to try to expose the impact of CO2 in this record are very appealing: the task is not an easy for so weak an effect, as Len reminds us above. Yet what a hoohah this weak effect has wrought in the hands of skilled propagandists and scared scientists! See the Lindzen post and the link to his recent presentation to realise how astonishing this all is:


  14. John, you misunderstood. I was talking about the 11 year solar cycle, not the annual orbital cycle. Annual variations are obviously removed by averaging anomalies, unless the annual cycle itself changes.

    I’m surprised you take this study so seriously, but since you confused the annual and sun spot cycles I suppose I should not be. Once again, just for you, can the sunspot cycle be detected unambiguously in CET? It is relatively large, at 1.3W/m2. I’ve never heard of it being detected, but maybe… What you are looking for, variations in emissions imprinted on CET, would be orders of magnitude smaller. You are wasting your time analyzing the statistics of the study and ignoring the data.

    All of this is quite obvious. It boggles my mind that half a dozen apparently intelligent people who claim to be capable of scientific scepticism don’t see that.


  15. Yes that would be of interest to investigate, Len, and it has been done. And I did misinterpret your reference to the solar cycle, by presuming you were talking of the seasons as the distance between the earth and sun varies through the year. In terms of estimated mean global temperature, for what its worth, that causes nearly a 4C change every year.

    Here is an example of a study linking CET variation with sunspots (in this case, a 22 year cycle): Here is an extract from their summary:

    The peak L1 = 3.36 years seems to empirically correspond to the El Niño-Southern Oscillation (ENSO) signal, which has a period range of within 3 to 6 years. ENSO is arguably the most important global climate pattern and the dominant mode of climate variability13. The effect of ENSO on climate in Europe has been studied intensively using both models and observational or proxy data e.g. refs 14, 15, and a consistent and statistically significant ENSO signal on the European climate has been found e.g. refs 14 and 16.

    The peak L4 = 22.6 years is coincident with the Hale sunspot cycle. This result partly agrees with some of the conclusions of Yang16. Solar variability has been shown to be a major driver of climate in central Europe during the past two millennia using Δ14C records. Furthermore, this result is essentially in good agreement with the findings of Scafetta e.g. refs 17, 18, 19, who found that the climate system was mostly characterized by a specific set of oscillations and these oscillations (61, 115, 130 and 983 years) appeared to be synchronous with major astronomical oscillations (solar system, solar activity and long solar/lunar tidal cycles).

    Liked by 1 person

  16. It doesn’t find an 11 year cycle, John, despite the forcing caused by that 11 year sunspot cycle being two, three or more orders of magnitude bigger than what the study under discussion is looking for. Doesn’t that make you doubt the significance of Munshi’s failure to find a correlation?


  17. Spectral analysis and time series cross correlation analysis are quite distinct techniques, Len. The former can be used to detect periodic effects in time series, the latter to look for apparent relationships between pairs of series. Munshi was not looking for cyclic effects.

    Liked by 1 person

  18. I realize that, John. But Munshi is looking for something perhaps 1000 times smaller than the 11 year cycle. Are you saying that x-correlation is more sensitive than an FFT? Would, for example, cross-correlating an 11 (ish) year period sine wave with CET find a cyclic component where an FFT didn’t?


  19. Well, I look bleary-eyed at my computer and find a new enquiry from Len. Quite had to make sense of it. Where did this claiming of ‘more sensitive’ come from? I think you mean ‘has more power’. As for doing that with a sine wave, it would be somewhat clunky compared with a spectral analysis using the FFT which will span a range of frequencies at the press of a button, or even better for the CET, enhanced by wavelet analysis in case of non-stationarity – the sunspot cycles not being at fixed frequencies throughout the record.


  20. You are dodging the issue, John. You cited papers that can’t find the 11 year sunspot cycle in CET data. Munshi is looking for a signal maybe 1000 times smaller than the sunspot cycle. You and cliscep think there is significance in him not finding it, but that’s like being surprised at failing to find a needle when you can’t even find a haystack.


  21. But there are papers that find sunspot cycles in the CET data – I’ve mentioned one. The major cycle is around 22 years, and the c.11 year is part of that. There’s another by Baliunas et al twenty years ago, and I think there has been at least one more recent one, but maybe you could go check on that.

    I don’t pretend to speak for ‘cliscep’ and nor does anyone else, but I suspect few of us are surprised at the difficulty in finding evidence of an impact on CO2 emissions on the CET, given that that gas, marvelous though it is in other ways, can be expected to have a minor impact if and when emission levels continue to rise over this century, let alone at the smaller levels of emissions earlier in the record. The IPCC folks have taken to calling for a show of hands in order to provide ‘confidence statements’ to the gullible about this and other matters, in lieu of convincing evidence or formal statistical analysis. Can you picture it? ‘.Who here thinks X is a big deal? Most? OK then let’s call X a big deal with high certainty and very high confidence.’ But I digress. The important thing about this Munshi paper, imho, is that he looked in the best way he could think of, and found nothing convincing. A small piece of the jigsaw, nothing more, nothing less. Careful tidying of loose ends is one way for scientific progress to proceed after all. In some ways I prefer it over the ‘here’s our grand scary conclusion, now let’s look for evidence for it, send money’ approach of the IPCC since that is more vulnerable to corruption.

    Liked by 1 person

  22. John, the main cycle should be 11 years – there’s a huge (relative to others) forcing with 11 year period. Anything at 22 years is minor in comparison.

    But the issue with Munshi’s work is that he is NOT looking for the effects of a relatively huge influence like the solar cycle. If it is there, that should stand out like a phallus on a forehead, relative to what he is seeking. Cross correlation of two sets of data is used to detect changes that occur in both datasets. Munshi is looking for the effects of the CHANGE between emissions in one year and the next, or maybe one decade and the next. For that to show up in temperatures there would have to be a step change in global emissions from one year to the next or one decade to the next etc. The only thing close to such a step change is the perhaps 1% annual changes that we have seen in recent decades. Going back before WW2, annual differences were very much smaller. Going back beyond 1900 to 1600, or wherever CET starts, is barmy because there were no sudden large changes in emissions.

    I don’t fault people like Geoff for not knowing that; it is not his area. But that should all be obvious to a statistician, yet you seem not to know.


  23. Munshi’s paper adds to our understanding of just how weak are the effects of our CO2 emissions on temperatures. That’s it. That’s all. There is a lot more that could be done with real or near to real observations, and perhaps one day better observations of CO2 (to improve on crude estimates that we still have to work with, mostly based on estimates of economic activities – such as used by Munshi).


  24. Not sure if this chat has run its course, or indeed is going anywhere. But I am. Off to cross the Atlantic by air later today, and that will put me out of (sensible) action/reaction for at least 48 or maybe 72 hours, being quite a delicate sort when it comes to sleep disruption, and much else besides.


  25. Have a good flight 🙂

    I don’t think the study tells​ us anything new. Forcing from CO2 is already well quantified.


  26. “Forcing from CO2 is already well quantified.”

    No, it is absolutely nothing of the sort, as I suspect you are well aware.

    You’re just clown dancing, aren’t you?


  27. Yes, it is. See for example

    Some people can’t resist trying to show that the greenhouse effect is not real or is not effective, but it just makes them look silly. That seems to be the thrust of this thread. I find it very odd that serious people (you can clearly count yourself out there if you choose, cat) indulge in this way.


  28. Statistics is not my field at all, but doesn’t Len have a point? Shouldn’t the 11 year sunspot cycle show up in the CET record? What’s the link with a 22 year variation which is two 11 year cycles that differ only in polarity (although one 11 year cycle is commonly weaker than the other)?


  29. Len Martinez

    I take your point that changes in emissions pre WW2 were tiny in absolute terms and therefore any effect might be masked by other possible forcings. Measurement of emissions is speculative and of atmospheric concentrations wildly inaccurate before 1957, which is why any graph I’ve seen shows CO2 concentrations as a barely rising straight line up to about 1950 with the Mauna Loa measurements tacked on from 1957, giving a true hockey stick graph. The CET graph doesn’t look like that. OK, that’s only one spot on the globe, so average any longterm series you can find, going back as far as you can, and you still get nothing that looks like the known pattern of CO2 emissions/concentrations.

    The Munshi paper seems to give that simple observation statistical backing. Arguing that the effect is too small to spot, or masked by other effects, doesn’t seem to me to provide any support for the argument for man-made global warming. I’m not saying it doesn’t exist, since the physics of the greenhouse effect suggests that it should, but that the tortuous arguments used to prove its existence don’t work.

    What we’re left with is a possible human warming effect too small to be measured or even detected, and a multi-billion dollar climate science industry which claims to be 95% certain that exists, but can’t explain how it knows, or how big it is within a 300% margin of error.

    Liked by 1 person

  30. There’s a simple argument to be made here. The signal of man-made global warming has been swamped, in the British isles at least, by natural variability of one kind or another. So when, if ever, can we expect the global warming signal to dominate the response of the British climate? 50 years, 100 years, 200, 500, a thousand? The Met office UKCP09 report says of CET:

    “It is likely that there has been a significant influence from human activity on the recent warming of CET.”

    It also predicts an annual mean temp rise in East Anglia of just over 3C and in excess of 4C for medium and high emissions scenarios, by 2080! So something’s amiss somewhere. I suspect the Met office overestimates the effect of CO2 on the British climate and grossly underestimates the influence of natural variability.

    Liked by 1 person

  31. Geoff, “average any longterm series you can find, going back as far as you can, and you still get nothing that looks like the known pattern of CO2 emissions/concentrations.”

    Is that true? The rise in temperature in the 20C and in this century correlate well with the CO2 record, whether you look at global or HADCET.

    Alan, We can perhaps infer from the lack of an 11 year cycle in temperature records that the heat capacity of earth is great enough that a 1.3W/m2 oscillating forcing is smoothed out. In other words, the time constant of the reaction to a change in forcing is many years and earth has yet to respond fully to past emissions. Everyone knows this, of course, except perhaps people who think it is a useful idea to try to correlate yearly variations in emissions with temperature.


  32. Len. To echo Catweazle666: DRIVEL. Climate isn’t just temperature; the 11year cycle occurs all over the place – in situations which are ultimately responding to temperature fluctuations or cloud coverage. I would imagine others more knowledgeable than I would wish to comment.

    Liked by 1 person

  33. Alan, Munshi isn’t trying to correlate CO2 with ‘climate’ but with a local temperature record (which is heavily influenced by weather). The 11 year cycle may well occur in many places, but does it occur in the temperature record he is analysing. If not, why would bother looking for something 1000 times smaller? And does it occur in the global temperature record? If not it is being damped by the heat capacity of earth, exactly like the response to the changes in CO2 levels.


  34. Interesting argument Len but I can use it against your interpretation. If, as you argue the 11 year cycle does not occur in the temperature record because it is being damped by the heat capacity of earth, then what is the 22 year cycle that has been detected? Using your argument this would have to be something that is NOT being damped by the heat capacity of earth.

    Liked by 1 person

  35. Len. How convenient. Do you treat anything that doesn’t fit with your preconceived view as spurious and dismissable?

    Liked by 1 person

  36. “How convenient. Do you treat anything that doesn’t fit with your preconceived view as spurious and dismissable?”

    It’s called ‘Clown Dancing’ Alan.
    Clown Dancers can keep it up indefinitely, you can never truly debate with one, debate is not their intention. They have a practically infinite collection of Alinsky-like pseudo-rhetorical techniques which ensure you will never ever prove any point whatsoever, they are utterly unacquainted with the concept of ‘good faith’.

    Sorry, but any attempt to debate with the likes of Martinez is doomed to continued frustration and is a total waste of time.


  37. If it is from the solar cycle, Alan,, why is it 22.6 and not 22, and what happened to the 11 year peak?


  38. The mean Hale Cycle length would be 22.2 years. So are you saying that because the paper which John linked to above found a cycle of 22.6 years, then this cannot be the Hale Cycle? What alternative hypothesis do you have?

    Liked by 1 person

  39. Len, using your various arguments above, how do you explain the occurrence of approximately 11 year cycles of South American river discharge, and of lake levels in North America and East Africa that correlate, without any time offset, with sun spot cycles? How come net rainfall amounts correlate with something you suggest will be damped out by the heat capacity of the Earth?
    Note that I have already acknowledged that you seem to have a valid point that the c. 11 year cycle appears to be absent from the CET record, but your explanation of this absence seems suspect.

    Liked by 1 person

  40. Jaime, it’s not that it cannot be the Hale cycle, just that the length doesn’t match. Before claiming that it is the Hale cycle and not just an artefact, that discrepancy must be explained. Maybe there is a good explanation, but I haven’t seen (or looked for) one.

    Alan, I don’t know whether the effects you list are real but it might be surprising if the sunspot cycle had no effect on earth. There is a clear mechanism and a large forcing. The magnetic Hale cycle has a less obvious and more subtle effect on earth. Yet according to the links provided above, it is suggested that the latter is detected in the CET but the former is not. You have perhaps seen evidence to overcome your normal scepticism, but I haven’t.

    Maybe you reason that if a faint signature from the Hale cycle can be detected, this gives strength to the daft proposal of this thread that failing to detect a tiny effect of variations in annual CO2 emissions amongst noise and weather in the CET somehow weakens the case for the greenhouse effect; that finding one unlikely signal means that other unlikely signals should be visible even if much more likely signals are not. I don’t think that is sensible, but I can see why it appeals to those who, for all the evidence and all protestations of being among the 97%, still want to deny the GHE.


  41. The idea of searching for annual variations is clearly exercising you Len. Can you remind us of who has looked, and presumably failed according to you, for an ‘effect of variations in annual CO2 emissions … in the CET’? As I recall, the Munshi paper we were discussing, looks at changes over 10, 20, and 30 years during timespans of between 60 and 245 years.

    And could you also refer us to your source for someone claiming that this failure ‘somehow weakens the case for the greenhouse effect’? The greenhouse effect, although badly misnamed, seems generally accepted around here, and the additional CO2 of the 50 to 100 years or so can be expected to have made a modest contribution to it. It also plausible that our contributions, not just from breathing, have added a lot perhaps even most to that rise in ambient levels. So Munshi set out to see if he could detect an effect on the CET using various techniques he argued would make the most of the data he looked at. Apart from plausibly spurious, because isolated, correlations in (as I recall) a couple of instances, he did not find convincing evidence of an effect. Now this does not mean the effect does not exist, but it is evidence that the effect is relatively weak. You too seem convinced that it is weak, but that seems to upset you inordinately when you hear of someone trying to look for it.

    Liked by 2 people

  42. John, the article begins: “This article by Jamal Munshi challenges the whole theory of AGW in the most radical way imaginable.” So according to Geoff (and the rest of you, since I’ve seen nobody else questioning this premise), the failure to find anything when correlating greenhouse gas emissions with CET – essentially correlating a DC signal (to use an electrical analogy) with and AC signal – challenges AGW theory, i.e. the greenhouse effect.


  43. Len,

    AGW theory ‘around these parts’ is generally taken to be the totality of alarmist climate science mooted on the singular proposition that (via an ENHANCED greenhouse effect – involving water vapour feedbacks + other earth system feedbacks), anthropogenic greenhouse gas emissions MIGHT result in ‘dangerous’, nay catastrophic global warming in the decades and centuries to come. AGW theory is not ‘the greenhouse effect’; it is a house of cards built upon poor science and statistical probability theory, stir-fried with a dash of the Precautionary Principle, popped into the oven on gas mark 7 for a few minutes, to produce a half-baked future Thermageddon. I believe you will find that this is what most commenters here are challenging.

    Liked by 3 people

  44. The Munshi paper is an addition to the stack challenging the AGW campaigners, but it does not challenge the greenhouse effect. So no cigar there Len. Now tell us about the source of these annual effects analyses that so concern you.

    Liked by 1 person

  45. LEN MARTINEZ ’09 May 17 at 1:56 pm)

    Of course it doesn’t challenge the greenhouse effect. Don’t be silly. It challenges the hypothesis of a measurable correlation between CO2 emissions and temperature rise, using a long-term temperature series collected under scientific conditions, by scientists.

    Note what the scientific establishment has done to try and confirm (not invalidate or disprove, as scientific practice demands) a correlation. It’s searched far and wide for a global coverage, taking any old rubbish from ships’ bottoms and airport carparks and averaged it out into an abstraction called mean global surface temperature.

    ACDC has nothing to do with it. Is there something going on? Not that we know.

    Here’s an analogy. Are we getting taller due to better diet? It would seem so. The evidence is long term measurement in an extremely limited portion of the globe (Europe) of the height of adolescents inducted into tha army over a period of about a century. It’s not global, it wasn’t collected with any scientific purpose related to proving anything, it’s just the only data that we have available.

    Imagine the arrival of a UN Commission on Height. They set up an Intergovernmental Panel on Height Change, which divides the planet into thousand kilometer-wide squares and starts examining old photos for evidence of height of the natives, correcting the data for difference in focal length of cameras and estimates of the height of solar topees of their colonial masters. And lo and behold, it turns out that the arrival of the colonialists is correlated to a shortening of the population.

    Well, maybe, and then again, maybe not. Because we’re talking history here, and people who talk history are by nature Sceptical, (see under “About” above, if you see what I mean.) CET is history. So is the industrial revolution and the rise in life expectation in Bangladesh from circa 40 years to circa 60 in about a generation.

    It’s all about what you think is important. Unmeasurable effects due to hypothesised cycles are not. A better life for the world’s poorest is, I think. But that’s just my sceptical opinion.

    Liked by 2 people

  46. John, no it is not yearly but 10,20,30 yearly. But that makes little difference when you have CO2 levels rising by perhaps 40ppm across 80% of the time period under study. It is basically a flat line, DC, zero Hz. Correlating a flat line with a wiggly line is going to give you nothing, whatever the data. For CO2, only the rise in concentrations post 1960 is non-flat, but they de-trended the temperature data (I assume from Geoff’s quoted paragraph) so that post-1960 rise isn’t going to match anything either.

    So according to Munshi and you (a statistician no less), correlating a flat line with a wiggly line tells you enough to “challenges the whole theory of AGW”. It is nothing but GHE denial.


  47. Have you looked at the emissions time series Len? I haven’t, but I would not expect it to be what you are so sure is a flat line. The boom years of for example the railway and steel industries in the the 19th C were over quite short periods, and might show up as discernible. But I also note Ron Clutz’s reservations about fossil fuel data, and so it is appropriate to have reservations about it. Even today, our emissions are estimated by proxy and may have to rely on government reports of fuel consumptions.

    Liked by 1 person

  48. In a similar vein, elimination of deciduous forests in the Americas, mainly by burning, occurred over relatively short time periods and would contribute to deviations from a CO2 flat line.

    Liked by 1 person

  49. John, Alan, I don’t know how flat the emissions data are. But CO2 concentrations rose 40ppm between 1772 and 1960 (much of that in the latter decades of that period) – that is the flat line.

    It is probably too much to expect that you accept that the GHE depends upon the concentration of CO2 rather than the emissions rate. Munshi likely doesn’t, or he wouldn’t have written this study. Just GHE denial, disguised for social acceptability, IMHO.


  50. Len. Please explain where you obtained pre 1955 global atmospheric CO2 values. Don’t use ice core data – not global and homogenized over many decades. CO2 values obtained from stomatal data suggests significant temporal variability, at least on a local scale.

    Liked by 1 person

  51. Alan, 280ppm is generally used for pre-1880. But you could choose another value, 270 or even 250 and it would not change the essentially flat line nature of the rise from 1772.


  52. Len. Really? And your EVIDENCE for this? Yes, yes, I know this is a generally accepted value, but as I intimated, some stomatal EVIDENCE suggests excursions, sometimes exceeding 350 ppm.

    Liked by 1 person

  53. CO2 levels passed 350ppm only in 1990 after decades of burning fossil fuels. Are you suggesting that levels might have been similarly high between 1772 and 1950 during the period we have been discussing? From what and where did it go?


  54. Sorry Len, debates don’t work like you would like them to. You made the assertion that CO2 levels pre 1955 were uniform and I asked you for the evidence. This, to date, you have singularly failed to provide.

    I am fully aware that there is no EVIDENCE and that Callender’s figure of 280ppm is highly cherry picked. I did wonder if you might own up to this. You failed.

    Liked by 1 person

  55. Alan​, we’ve been talking about the last 260 years. CO2 levels at the beginning of the Keeling curve were about 315ppm. Decades of gradually rising emissions from the industrial revolution onwards seem very likely to have caused levels to rise towards that 315ppm. If you really think that CO2 levels actually fell in that period you are the one who should present some evidence.


  56. Len, you just don’t seem to get it do you? It was you that first maintained the atmospheric CO2 level was flat and it was I that asked you for evidence of this (knowing that the evidence was both tenuous and disputed). Given that the oceans must have warmed overall in coming out of the LIA, and that the oceans should expel CO2 upon warming, it would seem most unlikely that the record should be flat. Or are you, like the Mann tribe, going to argue that the LIA was not a global phenomenon?

    Liked by 1 person

  57. I’ve taken a first look at the CO2 emissions data ( ), and of course it is not at all flat. For example, it more than doubles during the second half of the 18th century, and there is a fivefold increase during the first half of the 19th century. It is not all monotonic upwards, but is nearly so for most of the record. There are tiny dips for example in 1803 and 1831, bigger dips down in 1914 and 1915, 1918, 1919 and several more larger dips in the 20thC, in the early 1930s and 1980s for example.

    The estimates have been constructed for several sources, beginning with ‘solid’ (coal and wood?), and gaining over the years contributions from gas, liquid, cement-making, and flaring. I imagine it would salutary to go further back into the research to get an idea of the level of guesswork in these numbers. I regard their being reported in terms of tonnes of carbon, rather than of carbon dioxide, as grounds for suspicion since ‘carbon’ seems to be preferred by the CO2 alarm campaigners – presumably because they hope for negative associations in the public mind with such unwanted things as soot and burnt toast.

    Liked by 1 person

  58. Alan, CO2 outgassing from ocean temperature rises is, from memory, about 8ppm per °C. How many degrees cooler do you think average global temperature was in the LIA? For reference, CO2 levels rose about 40ppm between the Industrial Revolution and the 1960s.

    John, yes emissions rise, but we knew that. By 1900 they have risen more than 100-fold; by now, 3000-fold. But correlation works on the variations in that rise, not the rise itself – which is why the data were de-trended; the rise itself is irrelevant. Once de-trended, the curve is flat. I’d take it as read that, as a statistician, you understand that, but for more than a week you have seemed determined not to.

    And above and beyond that, emission rates are irrelevant, atmospheric concentration being the controlling factor. It is difficult to see how you have not understood this in all the years that you have been interested in climate change.


  59. Len, we are making quite good progress, but we’re not quite there yet. It is because correlation can be dominated by an overall rise that it can be of interest to do some kind of detrending to help look for things of interest in the residuals. Now ‘detrending’ is not precisely defined since it can be done in several ways. Perhaps the most common is making a simple straight line fit to the data and then subtracting the fitted values from the original data. This is the technique that Munshi used in another, somewhat similar study on global mean temperatures and decadal fossil fuel emissions ( ), and so it is possible he used the same method in the work we have been discussing (he mentions detrending, but does not state how he did it in that paper – though see page 3, third paragraph for details which suggest he may have used the same method in each time period used in his computations). This technique will not lead to what you call a flat curve in the residuals – the values to be used in the correlation computations.

    With regard to the importance, or at least potential importance, of man-made emissions, the chain of reasoning used by CO2 alarm campaigners is A leads to B which causes C, and C is really really bad, and so to deal with B we must stop, or at least dramatically reduce A. Where A = our emissions, B = higher ambient CO2 levels, and C = global warming. So it would seem they regard our emissions as the control knob we can use to choose our desired global mean temperature. Given that, it is both reasonable, interesting, and (given the political success of the campaigners) important to see what evidence can be found that our emissions are indeed some kind of control knob. Munshi has found none in his researches so far.

    Liked by 1 person

  60. John, getting somewhere? No, temperature changes according to atmospheric concentration, which is related to the integral of emissions. And the trends in emissions and temperature are the signals you should be looking for. Munshi is clearly removing the signal if he is de-trending the data. That he finds nothing is not so surprising, but the chances are he knows no better.

    You guys should know better by now, but for unknown reasons you think that by throwing away the data and looking elsewhere, Munshi “challenges the whole theory of AGW in the most radical way imaginable”. That is not “getting somewhere”.


  61. That should have read, “throwing away the *signal* and looking elsewhere”


  62. I think we’re nearly there now, Len. What you need to realise is that anything with an overall rising trend is likely to show appreciable correlation with anything else with an overall rising (or falling for that matter) trend. This tends to be amplified if cumulative totals are used. So, finding appreciable correlations in these circumstances can be no big deal by itself. For example the temperature rises in the first half of the 20th century are not widely attributed to rising CO2 emissions, but it seems likely that there is positive correlation to be found there since both show an overall rise. So, the objective scientist, as opposed to the demented activist, would seek to assess carefully as much temperature and CO2 data as can be found. The mere overall correlation would not be good enough. So Munshi undertook to see what he could find after detrending. After all, if our emissions, and not Mother Nature’s despite hers being so much larger, are indeed the control knob the campaigners insist we must turn down to get a better, or at least cooler, climate, then it seems worthwhile to look for evidence of this in the variations to be found over and above the overall trend. Munshi has been doing this, and so far has found nothing at all convincing by way of an apparent relationship. Now this does not mean that such a relationship does not exist, but it does challenge the notion that it matters very much. I wonder if he might do the same sort of analysis to compare ambient CO2 variations with our emissions variations, and thereby shed some light on the A to B stage in my precis above. So far, as far as I know, he has just been looking at A to C. There is B to C to be examined as well. I hope all will be in due course, as part of a reasonable due diligence given the detrimental impact the CO2 Alarm campaigners have already had on the world, and the scope for them to do even more harm in the future.

    Meanwhile it seems to me that the climate system continues to behave just as it might if our CO2 emissions were of little importance to it. But some further investigation of this nevertheless does seem worthwhile to some modest extent. So I look forward to more work like Munshi’s in the future, or to becoming acquainted with more such work already done by him or others.

    Liked by 1 person

  63. John, I think we are all well aware that correlation does not prove causation. To interpret a correlation one must consider also the physical mechanisms involved. That is lacking in Munshi’s work – he’s just playing with numbers. They don’t mean anything to him, there’s no understanding of whether what he is doing relates to real things​. It is odd that retired folks like you, Geoff and Alan​ don’t see that, but if enriches your declining years, I’m happy for you.


  64. Your good wishes are much appreciated, Len, but you have just blundered again in suggesting Munshi has ‘no understanding of whether what is doing relates to real things’. A general overall rise in ambient CO2 levels seems reasonably ‘real’, don’t you think? And even for that can of worms, estimated global mean temperature, an associated overall rise over the past 150 years or so seems at least plausible, and is held to be real enough by lots of folks. Emissions of CO2 from the burning of fossil fuels seems to be out there in the set of real things too. Here is Munshi linking his work to all of them:

    ‘At the core of this line of reasoning are the assumed linkages that exist in climate models between fossil fuel emissions and atmospheric CO2 and between atmospheric CO2 and surface temperature, with the implication of these linkages leading to a direct causal relationship between anthropogenic emissions and surface temperature (Taylor, 2012) (Good, 2013) (Geoffroy, 2012) (Caldeira, 2013). The first causal connection – that between fossil fuel emissions and atmospheric CO2 – is the weak link in this chain because it is confounded by large uncertainties in natural flows of CO2 between the surface and the atmosphere (Ballantyne, 2015) (Huntingford, 2009) (Falkowski, 2000) (Roe, 2007) (Guan, 2012). In previous works we presented empirical evidence of uncertainties in the IPCC carbon budget (Munshi, Uncertain flow accounting, 2015) and of the absence of a correlation between annual changes in atmospheric CO2 and annual rates of human emissions in the detrended series (Munshi, Responsiveness of atmospheric CO2, 2015). In this short note we look for empirical evidence for the implied causal connection that links fossil fuel emissions to changes in surface temperature. It is a pivotal relationship. The worldwide effort to lower fossil fuel emissions rests entirely on the ability of changes in fossil fuel emissions to attenuate surface temperature.’

    I happened to come across that earlier today when I went back to that paper ( to see what might have been done on the A to B to C front which I talked about in an earlier comment. Quite a bit it would seem, as you can see from the extract above. The promise of further enlightenment beckons.

    Liked by 1 person

  65. When I said M had no understanding I was being polite, giving him the benefit of the doubt. Just a look at the Keeling curve and at emissions curves is enough to make one think that the two might be causally connected. Any doubt is removed when one realises that emissions exceed the amount of CO2 needed to cause the Keeling curve. So someone who decides that the presence or absence of a correlation between yearly changes in the two quantities challenges the causal connection is at least ignorant. Similarly for the link between short term emissions and temperature fluctuations. Only someone who doesn’t appreciate his own level of ignorance would take M’s path. I think this is referred to as Dunning-Kruger-ism.

    Liked by 1 person

  66. Len what utter tripe:
    “Just a look at the Keeling curve and at emissions curves is enough to make one think that the two might be causally connected.”

    Just a look at the Keeling curve and at a curve for the use of microwave ovens is enough to make one think that the two might be causally connected.

    Liked by 2 people

  67. Comment: AGW does not suggest that there should be a correlation between emissions and temperature of some really tiny region (CET) of the surface of the planet. Not finding such a correlation (which is not expected) does not, therefore, somehow challenge AGW.


  68. Attp, yes clearly. Geoff did point that out early on I think, but he and everyone else continue to think the work is groundbreaking. The study even de-trended the data to remove the obvious correlation…

    I was talking in the context of the climate debate. Of course one needs some background information. By the way, did you figure out how many degrees cooler the LIA must have been for your proposed outgassing to disturb the flatness of the CO2 curve? 8ppm per degree, I think it was….


  69. By the way, did you figure out how many degrees cooler the LIA must have been for your proposed outgassing to disturb the flatness of the CO2 curve? 8ppm per degree, I think it was….

    If you consider only the oceans, then I think it is more like 12ppm per K. However, I think there is also some coming from vegetation changes, so that the total effect is about 18ppm per K.


  70. Well, this thread may be turning into a bit of a clinic to help victims of the CO2 Scare. That’s a worthy cause, and although we can but help a few, perhaps our compassion will inspire others to join in.

    Here’s what that high temple of AGW, the UK Met Office (, had to say in UK Climate Projections:

    ‘Central England Temperature has risen by about a degree Celsius since the 1970s, with 2006 being the warmest on record. It is likely (>66% probability, IPCC) that there has been a significant influence from human activity on the recent warming.’


    Now I suspect that by a ‘significant influence from human activity’, the authors, and the IPCC, are not referring to badly sited weather stations, nor to the adjustments to climate records by such as NOAA. I think they mostly mean manmade CO2 emissions. Indeed the above report has a helpful spreadsheet with a column headed ‘Emissions’ wherein one can see that ‘High’ is associated with higher temperatures to come than ‘Medium’ which in turn has higher temperatures than ‘Low’.

    In the light of such as that, the toils of Munshi are most welcome by way of adding to investigations into the possible impact of our emissions might have on temperatures. Other recent work is of interest. Here are a couple of geographers playing with an ‘Earth Systems Model’ which shows how temperatures might rise following a sudden surge in CO2 emissions – they find the biggest rise is within about 10 years, but that temperatures may continue to rise a bit for much longer: . Now I am wary and weary of climate models, but they are part of what might be called the AGW culture and so it is good to keep in touch with them.

    And there are some handy numbers in this blog post ( looking at CO2 and temperature variations in these five time periods: 1850 – 1910, 1910 – 1940, 1940 – 1977, 1977 – 2001, and 2001 – 2014. Only in one of them is there any fuel to stoke the anxieties of those alarmed by CO2, and the warming then was similar to an earlier period which had a much smaller rise in ambient CO2.

    Liked by 2 people

  71. Len, ATHL, With an important proviso that those of your ilk commonly ignore AOTBE (all other things being equal).

    Liked by 1 person

  72. Attp, thanks for the correction. The graph means shows CO2 rising 12ppm for a 1°C temperature rise, but does that refer to the bulk temperature or only the surface? Alan’s line was that a big dip in temperature during the LIA meant CO2 levels could not be considered stable at 280ppm pre-industrial revolution (with no mention of all other things being equal). It seems unlikely that the ocean bulk cooled by a degree or anything close during the LIA.

    John, at its most basic, scepticism seems to be a rejection of the GHE, but dressed up to appear more sophisticated, less stupid. Just for clarification, do you actually accept the forcing formula 5.35ln(C1/C0) mentioned earlier? If not, why?


  73. Len,
    That’s a calculation I did based on equilibrium carbonate chemistry and that included how it responds to changes is sea temperature, which is probably representing more the sea surface (or the well-mixed layer) than the bulk ocean. My understanding is that the LIA is associated with a change in global temperature of less than 0.5K, so would probably be associated with a change in atmospheric CO2 of less than 10ppm.


  74. “Just for clarification, do you actually accept the forcing formula 5.35ln(C1/C0) mentioned earlier? If not, why?”

    It really is religion, isn’t it? Is there any physics behind the formula? Does it link into other laws of physics or is it just a statistical fit over a small band of observed data?

    As a hypothesis it can be accepted as long as it gives reasonable results, just like the Phillips curve in economics that said that inflation and unemployment had an inverse relationship. All went well until the 1970s where unemployment and inflation both rose massively.

    Liked by 1 person

  75. Does that mean that you reject the formula, that you accept it, or that you think the physical characteristics of greenhouse gases are as fickle as the human behaviour underlying economic rules?


  76. ATTp’s graph lacks context of course. There is no credible measure of temperature that shows a 4 degree C increase since pre-industrial times, including SST and of course total ocean temperature has barely gone up at all.

    Liked by 1 person

  77. Since we are showing graphs here is a more credible one from Steve McIntyre that is more relevant.


    I would like to hear some context for the Holocene conundrum. During most of the Holocene temperature have been falling significantly while GHG’s increased significantly and AOGCM’s show a significant increase in temperature.


  78. The whole ‘forcing’ notion seems to me to have been invented to give computer programmers an easy way to include everything but the kitchen sink in their models as and when they wish. All you need to do is come up with a ‘forcing’ at the top of the model atmosphere, and there you have it! Easy peasy, very appealing. Maybe one day we’ll have models good enough to include more physics, and see these ‘forcings’ produced as effects – rather than skipping all that, and just putting the presumed effect in as a cause. In the meantime, the ‘forcings’ wheeze provide a playground for the modellers to ply their trade with.

    As for the formula, as far as I know it is another reflection of weaknesses in atmospheric theories, being a semi-empirical fit to observations using a relationship possibly first conjectured by Arrhenius way back in the day, and deployed in a formula by Callendar in a 1938 paper. Steve McIntyre has an amusing and engaging post ( in which he shows how Callendar’s 1938 work outperforms most GCMs tested for comparison when it comes to fitting estimated global mean temperatures. It is interesting to note, as M does, that Callendar’s approach points to a relatively low climate sensitivity. This means it is of little political, and therefore fund-raising, interest, and the lists of ‘forcings’ in the GCMs give more scope for heating things up quite a lot more.

    Liked by 1 person

  79. Just to be clear (or, for those who are actually interested) the figure I posted in this comment was simply illustrating the relationship between sea surface temperature and atmospheric CO2 for a fixed CO2 reservoir size. In other words, from the oceans only, atmospheric CO2 will vary by just over 10 ppm for every 1K change in surface temperature. It’s, of course, not actually linear, but for the temperatures of interest, this is a reasonable approximation. This was also a response to a query about how the change in global temperture during the LIA would probably have influenced atmospheric CO2 (the answer being not very much, probably less than 10ppm).


  80. Met office climate scientist Ed Hawkins, and others, believe that the fingerprint of (man-made) global warming is clearly visible in the Central England temperature record.

    “It is perhaps surprising that the fingerprint of a warming climate is clearly emerging on very small spatial scales and our new paper examines this topic.

    Consider central England – an area which occupies just 0.005% of the planet’s surface. For this region, the year-to-year variability in annual temperatures is much larger than for a global average (Figure 1) – this is obviously because such small areas are sensitive to fluctuations in the weather, and these fluctuations are spatially averaged in a global mean.

    But, note that the familiar features of the global temperature record are clearly visible by eye in the central England record too, just overlaid by the large annual fluctuations. When averaging over multi-decadal timescales, the correlation between the global and local timeseries is r = 0.96.”

    So, AGW advocates can’t have it both ways. Either global warming is, or isn’t a significant influence on regional warming in central England, in which case, one might expect to find a correlation between emissions and rising temperatures, at least over longer time scales.

    Liked by 1 person

  81. Jaime,
    You appear to have switched from correlating emissions with temperature, to a long-term trend emerging from a noisy dataset. These are not entirely consistent, even if our emissions are what is driving the long-term temperature trend.


  82. Ken,
    If there is a discernible long term trend in temp in the CET, it appears relatively unimportant compared to decadal fluctuations in temperature. If global GHG emissions were driving temperatures and if we accept the premise that, ignoring the year to year ‘noise’, 21 or 31 year smoothed CET is a reasonable proxy for GMST, then we would expect the smoothed CET plot to at least resemble the plot of global emissions. It doesn’t, however. Emissions climbed sharply and uninterruptedly from 1950 onwards. Smoothed CET still betrays the signal of global cooling from the mid 40s to the mid 70s, only then showing a sharp warming trend (not unprecedented in the record) from 1980 to 2005/6, whereafter an equally sharp cooling trend sets in, interrupted in 2014, which just happens to coincide with the beginning of the El Nino/Pacific Warm Blob years. Thus, it appears that smoothed CET (and global mean surface temperature) do not correlate well with the supposed dominant global climate forcing (GHG emissions), even over multi-decadal timescales. What this in turn suggests is that multidecadal variability has dominated thus far regional and global climatic variability, which in turn suggests, given the exponential increase in emissions post 1950, that anthro GHG forcing of climate is, at present, very modest. That’s it. Munshi’s work would appear to confirm this conclusion in a statistically rather more robust manner.


  83. John, the formula is indeed derived from observations – from raw data. Its validity depends on the quality of those data. Do you think the data are wrong, if so on what basis? How would you prefer this characteristic of CO2 to be determined?

    It surprises me that people here should engage in such clear GHE denial, when the avenue of challenging climate sensitivity offers a much richer seam to tap, one that doesn’t open you to derision.

    Jaime, there is of course a correlation between long term t rends in cumulative emissions and temperature and a clear cause- the GHE for those who believe in it. That doesn’t mean it’s sensible to correlate de-trended short term emissions and temperature.


  84. Why do you think I am questioning what you call the validity of that formula, Len? After all, I gave a link to where it was demonstrated that it could perform better than GCMs in fitting historical data (and that confection called global mean temperature). Now beating GCMs is faint praise, but it is something.


  85. Len, as far as I’m aware there is no really convincing evidence that GHG emissions have been the dominant driver of climate change since 1950 – as one might expect if the transient climate response is large and significant – not even in terms of a simple strong correlation.


  86. “Does that mean that you reject the formula, that you accept it, or that you think the physical characteristics of greenhouse gases are as fickle as the human behaviour underlying economic rules?”

    Len, this is silly. An economist thought he had found a principle by numerical analysis but it turned out not to work. So it was abandoned. Nothing to do with fickle human behaviour.

    But your response shows a deep reliance on religiosity. A formula either works or it doesn’t. Isn’t that the difference between science and unscience? There is no point asking me whether I believe in Boyle’s Law. It works.

    So the formula you put up. When you learn Boyle ‘s law you get the surrounding stuff… Fixed mass of gas at constant temperature. At university you probably get to understand quite why it works.

    Let’s turn to climate. The formula you present… Where does it come from? How does it derive from other bits of physics? What are the limitations as to its usefulness? Does Pierrehumbert’s textbook describe its derivation?

    I do not know because I am not interested. However it strikes me that it cannot be correct because :

    If the CO 2 graphs are right, levels of the gas have been rising since whenever..

    The global temperature indices do not suggest a linear increase in temperature, even if you go full bore Rahmstorf or Foster

    The relationship claimed in the formula is a linear one


  87. “When averaging over multi-decadal timescales, the correlation between the global and local timeseries is r = 0.96.”

    [Ed of the Met, on correlation of Central England Temperature with global temperature, quoted by JAIME above at 16 May 17 at 9:52 am ]

    This is good to know, and surprising, since, over part of the time period there’s nothing to correlate with; over a century or so there’s only a few other series in Europe; and over the “global” period usually defined as starting in 1860 we’re comparing scientific data with a mishmash of data from all sorts of odd sources.

    So it would seem that Ed Hawkins is adding important information to Munshi’s article, suggesting that if there is no correlation between emissions and CET, then there is none between emissions and global temperatures either.

    So AND THEN THERE’S PHYSICS’ objection (15 May 17 at 9:33 am) that central England is a “really tiny region of the surface of the planet” fails.

    The point that everyone misses, I think, is that historical temperature measurements are historical. Long term temperature measurement covers a tiny part of the globe because only a tiny part of the globe contained a society able and willing to gather and preserve such data. This is not about cultural superiority. Maybe they did the same in ancient Mexico or China or Baghdad and we just don’t know about it. Too bad. What we know is that temperatures have been rising in the same places that CO2 emissions have been rising, and the causal link is this: scientific progress, which leads to increased interest in data collection and increased use of fossil fuels.


  88. “So AND THEN THERE’S PHYSICS’ objection (15 May 17 at 9:33 am) that central England is a “really tiny region of the surface of the planet” fails.”

    And yet the rings of one tree – YAD06 IIRC – on an obscure Russian peninsula is sufficient for the crackpot AGW alarmist “climate scientists” to eradicate a thousand years of accepted temperature data and proclaim that we must cease to use the most advantageous technology – the burning of fossil fuels – that has been far and away the greatest influence for good that humanity has ever experienced, all the while making damn good livings on the back of their hypocrisy.

    Liked by 1 person

  89. John, “Why do you think I am questioning what you call the validity of that formula, Len? ” The bold says it all.

    Jaime, rising CO2 and rising temperatures. The correlation is clear.

    Man in a barrel, economics is all about fickle human behaviour. That is why is is a *social* science. The CO2 response is physical data that you could reproduce yourself. That is why the CO2 forcing is generally believed and economics is not.


  90. CATWEAZLE666 (16 May 17 at 10:11 pm)

    And yet the rings of one tree on an obscure Russian peninsula … is sufficient…to eradicate a thousand years of accepted temperature data ..

    Be fair now, there were also a few dozen strip bark pines up on the treeline in the Rockies where nobody lives which generated the hockeystick which dictated a multi-trillion dollar programme to return us to the Middle Ages. It wasn’t all Keith Briffa’s fault. Indeed, some have suggested that his anguish over the hiding of his decline led him to leak the Climategate emails. Will we ever know the truth?

    I recommend everyone to listen to Professor Raleigh,
    a human being grappling with issues which may mean life or death for hundreds of millions of people, and who is bravely defending her work against forces which neither she nor we understand, but which threaten to return us all to the dark ages.


  91. LEN MARTINEZ (16 May 17 at 10:47 pm)

    …rising CO2 and rising temperatures. The correlation is clear.

    And so is the correlation with rising population, rising life expectancy, and rising just about everything else that makes us happier and better off in every way.

    .economics is all about fickle human behaviour. That is why is is a *social* science. The CO2 response is physical data that you could reproduce yourself. That is why the CO2 forcing is generally believed and economics is not.

    Economics is all about fickle human behaviour, which is why we fickle humans attach so much more importance to it than we do to the exact percentage of a trace gas in the atmosphere and its possible effects in a hundred years’ time.

    Your interventions in defence of some divine principle of scientific equilibrium of trace gases in the atmosphere only reinforce my Marxist belief in science at the service of mankind – the onlie begetter and only judge of what is science. I like what we’re doing to the planet, until someone gives me a reason to dislike it. Go on.


  92. I’ve just noticed to my embarrassment that Jaime had already linked to the UK Met Office’s UK Climate Projections with her ’04 May 17 at 9:57 am’ comment. I made such similar remarks and links on ’15 May 17 at 12:16 pm’ [in response to ATTP’s peculiar claim (‘ 15 May 17 at 9:33 am’ ) ‘AGW does not suggest that there should be a correlation between emissions and temperature of some really tiny region (CET) of the surface of the planet’ ] that I think must have been influenced by Jaime’s comment but managed to forget about it during the 11 day gap.


  93. “Jaime, rising CO2 and rising temperatures. The correlation is clear.”

    No it isn’t, not even close..

    Stop making things up.

    And even if it was, it would be meaningless.

    And write out 100 times “correlation does not imply causation”.


  94. Len,

    “Jaime, rising CO2 and rising temperatures. The correlation is clear.”

    Actually, we were talking GHG emissions, but no matter, there appears to be a very close relationship between GHG emissions and atmospheric CO2 (at least since 1958)

    As Rog Tallbloke pointed out above, there is a very impressive correlation between 10Be concentration (a solar activity proxy) and CET. There’s also an impressive correlation between previous solar cycle length and North American temperature. The same can be demonstrated for global temperature.

    As you can see, the ‘correlation’ with CO2 doesn’t look very impressive during the 1920s/40s warming (if you accept the CO2 Antarctic ice core data – which all AGW believers do) and the subsequent 1950s/70s cooling. This is especially so if you look at Arctic warming, where the Arctic in the 1930s/40s was as warm, if not warmer than it is today. So it would be wise to heed Catweazle’s remark above that correlation is not evidence of causation, especially when correlation is only good for the period of really rapid warming 1980 to 2000, or thereabouts.


  95. Jaime, tallbloke’s correlation is not so great.As I said to him above, “Tallbloke, looks as if you [have] a little “divergence problem” post-1970 in your 2nd graph. And there’s plenty of anti-phase activity and even CET preceding solar. Looks like England controls the Sun at times.”

    As for your graph, very pretty. But is it meaningful? Filters can alter phase, and the relative sizes of the curves can change the impression of one curve preceding or following another. So reading much into it seems fraught with difficulty​. Any relationship seems to break down after 1970 anyway, so what’s the point?

    Geoff, do we attach such importance to economics? Austerity at the zero lower interest rate bound suggests not. As to liking what we are doing to the world, that’s easily said when you live in a privileged part of the world; most people don’t.


  96. What a jumble, Len. You are in a bit of a shambles. But hopefully help is to hand. But not from me this time. I have some other matters to deal with that are time-consuming. So please excuse me for a while.


  97. “As for your graph, very pretty. But is it meaningful?”

    Not to you, as you clearly lack sufficient knowledge and understanding of the underlying science to comment one way or the other.

    All your clown dancing, dodgy rhetorical devices, goalpost changes etc. are never going to compensate for the obvious fact that you are purely regurgitating/C&P’ing second hand concepts – many of which are, due to their origin completely incorrect – that you do not remotely understand.

    Liked by 2 people

  98. How do you explain the total lack of correlation of temperature to cycle length after 1970? This is just more GHE denial, which seems to summarise many or perhaps all of you. You’re the “anything but carbon’ brigade.


  99. Len but how do you explain the total lack of correlation between temperature and ghgs for much of the record? If you can explain why ghgs are a reliable predictor then maybe discussion is worthwhile. Until then, we are all just flailing about in almost total ignorance. Just because you have a religious belief in the impact of CO2 does not make it an important determinant of the climate.


Leave a Reply

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

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

Facebook photo

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

Connecting to %s

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