RealClimate logo

Noise on the Telegraph

Filed under: — rasmus @ 11 February 2015

I was surprised by the shrill headlines from a British newspaper with the old fashioned name the Telegraph: “The fiddling with temperature data is the biggest science scandal ever”. So what is this all about?

The story makes serious allegations, however Victor Venema explains why the Telegraph got it wrong in Variable Variability, and makes the point that three hand-picked stations from Paraguay – out of thousands – hardly matters. He also shows the effect of post-processing on the global mean temperature: it reduces the global trend compared to raw data.

The story also sparked some discussion between colleagues at the Norwegian Meteorological Institute, where I work, as several of us know the scientist cited in the paper quite well. It would be completely out of character that he’d endorse the views expressed in the article, and this is also what he conveyed to us.

The records show that the recent high temperatures on Iceland are unprecedented, contrary to the main message from the Telegraph. And the evidence is not just in the temperature, but in a wide range of observations.

I like to look at the numbers myself, especially since the journalist responsible for the Telegraph story, Christopher Booker, bases some of his allegations on climate records with which I have some experience. Booker dismisses the data records and claims that

weather stations across much of the Arctic, between Canada (51 degrees W) and the heart of Siberia (87 degrees E). Again, in nearly every case, the same one-way adjustments have been made, to show warming up to 1 degree C or more higher than was indicated by the data that was actually recorded.

It is implied that such adjustments have been made to the Goddard Institute for Space Studies (GISS) data as well as the data from the National Climate Data Center (NCDC).

The purports about systematic one-way adjustments can easily be tested by comparing the trends in the GISS data with the independent North Atlantic Climate Data (NACD) or a more recent temperature analysis for Svalbard by Nordli et al. (2014).

It is straightforward to test Booker’s claim with open-source methods and data (see R-srcipt), and when we compare the independent Svalbard temperature from Nordli et al (2014) with GISS, we see that the GISS data has a smaller annual mean trend than the independent Norwegian data set for the same years (Figure 1).

Comparison between Nordli et al. (2014) and the GISS annual mean temperature for Svalbard. - See more at:

Comparison between Nordli et al. (2014) and the GISS annual mean temperature for Svalbard.

But is Svalbard representative for the this part of the Arctic? We can repeat the exercise for the most important temperature records from this region, and it is clear that there is no one-way adjustment, as purported in the Telegraph (Figure 2). In other words, our inspection of the actual data shows that Booker’s claim is false.

Comparison of the trend in annual mean temperature between NACD and GISS data for a number of locations in the North Atlantic and the Barents region. The size of the symbol indicates the length of the temperature record.

Comparison of the trend in annual mean temperature between NACD and GISS data for a number of locations in the North Atlantic and the Barents region. The size of the blue symbols indicates the length of the temperature record.

Another question is why there are differences between the different data sets. The GISS data are mostly taken from raw world weather records that have not been subject to the same quality control and homogenisation as the NACD or the Svalbard temperature of Nordli et al (2014).

Homogenisation is needed to remove effects from non-climatic artifacts, such as a change in the formula over time for estimating monthly mean values. Other conditions may include relocation of the thermometer, urban encroachment, or replacement of the instruments. GISS uses the GHCN adjustments, rather than doing it itself, but the way the GHCN corrects for artifacts may be different to the local met services.

Booker also accuses GISS and NCDC of using

“the warming trends to estimate temperatures across the vast regions of the Earth where no measurements are taken”.

A bizzare claim, and one may wonder why it seems so natural for Booker to make associations with misconduct. George Monbiot, on the other hand, has referred to him as a charlatan bent on spreading misinformation. In any case, a person who writes such a misleading story shows little respect for his readers.


  1. . Nordli, R. Przybylak, A.E. Ogilvie, and K. Isaksen, "Long-term temperature trends and variability on Spitsbergen: the extended Svalbard Airport temperature series, 1898–2012", Polar Research, vol. 33, pp. 21349, 2014.

123 Responses to “Noise on the Telegraph”

  1. 1
    toby52 says:

    Simple explanation: Christopher Booker is an [edit]

  2. 2
  3. 3

    Svalbard is very difficult to homogenize because the neighboring stations used to compare with to find non-climatic changes are mostly far away. That makes it even more important to work very carefully and gather as much information on the stations history, like the Norwegian weather service did.

    The above post shows the difference between GISS and the Norwegian weather service since about 1980. I was curious about the long-term differences and had a look at Berkeley Earth.

    Oyvind Nordli and colleagues find a trend in Svalbard Airport between September 1898–December 2012 of 2.6°C per century.

    Whereas Berkeley Earth finds a trend since about 1910 of: 1.11 °C per century. (If I understand their table right.)

    Thus if there is a conspiracy it includes all the national weather services as well. Both GISS and Berkeley Earth underestimate the trend relative to the values of the national weather service, which I would see as the most reliable estimate.

    I know of another study where the Alpine region was compared to CRU and the the trend of the Austrian weather service was 0.5°C per century larger.

    Does anyone know of more examples of such comparisons? I would like to make a study out of this and would be very grateful for any hints.

  4. 4
    Ray Ladbury says:

    Hmm, How is it possible to discuss Christopher Booker without the use of epithets? Such a discussion would at the very least seem incomplete.

  5. 5
    MARodger says:

    Whilst in his descriptions of C. Booker, toby52 @1&2 appears to have used language a little too lively for the moderator’s tastes, it is evident that C. Booker simply acts here as an echo-chamber as he is but repeating assertions made in a blog by P. Homewood. This Homewood blog rambles on as such blogs usually do. The guts of the accusation is that Homewood has found 19 weather stations in the Arctic GHCN have been adjusted to increase AGW. NOAA graphical records are presented appended to the blog to show the impact of these adjustments (although not all appear to show an increased trend over the full station record results from adjustemnts).
    The NOAA graphical records for all GHCN stations can be found from this directory and stations and their ID numbers can be found on the NASA click-on-map station-finder on this page. Somebody may wish to demonstrate how much (or how little) Homewood has resorted to cherry-picking.
    The previous C. Booker piece in the Telegraph repeated Homewood’s assertions about adjustments in Paraguay. It is all very silly and nonsensical but C. Booker has been writing stuff like this for years.

  6. 6

    Hi Rasmus,

    When you say that GISS data are mostly taken from raw temperature records and not subject to homogenization, that isn’t strictly true. Since 2011 or so, GISS has used GHCN v3 adjusted data as its primary input stream. This data is processed via the pairwise homogenization algorithm described in Menne and Williams (2009), which compares stations to their nearby neighbors to detect localized breakpoints associated with station inhomogenities.

    As Victor points out, however, these automated homogenization techniques only work well in areas with dense station coverage, and for somewhere isolated like Svalbard manual homogenization like that performed by Nordli and colleagues will likely yield more accurate results.

    See this related discussion by Cowtan and Way about homogenization and the arctic:

  7. 7
    Jon Keller says:

    Good work. In case anybody missed it, here is Kevin Cowtan explaining the Paraguay adjustments. And here is a post about temperature homogenization from …And Then There’s Physics.

  8. 8
    Darrel says:

    Sourcewatch gives a nice summary of this fellow:
    “To highlight the level of inaccuracy and falsehood in skeptical journalism the Guardian launched a prize in 2009 to be “presented to whoever crams as many misrepresentations, distortions and falsehoods into a single article, statement, lecture, film or interview about climate change”. This was called the “Christopher Booker prize” [6] The first nomination was inevitably Christopher Booker for an article about arctic sea ice with six errors in 900 words. [7]

  9. 9
    Sydney Bridges says:

    Not for nothing has the “Telegraph” been known as the “Torygraph” for the last fifty years-maybe longer, but that’s as far back as I can remember. As for doing its generally well-heeled readers a disservice, I suspect it is telling them exactly what they want to hear, which certainly isn’t the truth. There was a time when the Torygraph was a good source of news, as long as you filtered out the obvious political bias. My father used to read it when I was growing up and I too read it. Now, when I’m in the UK, I don’t even bother to glance at its front page on a news stand and I certainly wouldn’t pay for a copy.

    As for Christopher Booker, he is not one to let fact-checking get in the way of a good storyline. His libel against Rajendra Pachauri, Chair of the IPCC, was, from what I recall, expensive for his employers and the story was withdrawn, in other words it was a lie. His views on asbestos, smoking, and intelligent design round out a resume of ignorance and/or incompetence. But, although he has been condemned by at least one High Court Judge for his misleading stories (“Mr. Booker’s articles comtain significant factual errors and omissions”), he is praised by fellow Torygraph hack and climate denier, James Delingpole, which says about all I need to know on this particular story.

  10. 10
    Lars Karlsson says:

    If one wants to understand Booker, one should read his article “Charles Darwin zealots have made science a substitute religion”.

  11. 11
    John Mashey says:

    Rasmus: Having purchased Booker’s “The Real Global Warming Disaster”,
    I have one real concern about your post:

    “especially since the journalist responsible for the Telegraph story, Christopher Booker”

    Booker is a journalist? I hadn’t realized that.
    I see Journalism ethics and standards and there is dissonance.

    “While various existing codes have some differences, most share common elements including the principles of—truthfulness, accuracy, objectivity, impartiality, fairness and public accountability—as these apply to the acquisition of newsworthy information and its subsequent dissemination to the public”
    I know journalists like that who deserve the name.

    Surely there is a better term for writers like Booker?

  12. 12
    rasmus says:

    I may have misjudged his role, but I assumed he’s a journalist since he writes for the Telegraph.

  13. 13
    rasmus says:

    Thanks. -rasmus

  14. 14
    tamino says:

    One of the best refutations of the “adjustments are evil” meme comes from none other than Steven Mosher:

  15. 15

    Yes, I’ve been seeing some of this chatter on, among some of the usual suspects & wannabes. The laughable Steve Goddard has been pushing this line for quite a while, and one ‘chiefio’ used to peddle it for WUWT–IIRC, the kerfuffle last time was anent Bolivia, not Paraguay, but whatever…

  16. 16
    Jim Eager says:

    Rasmus wrote: “I may have misjudged his role, but I assumed he’s a journalist since he writes for the Telegraph.

    Rasmus, apparently you are not aware of the difference between a journalist and a columnist. The former deals with news and facts, at least ideally. The latter openly deals with their opinion of the news and ‘facts.’ Fair to say the latter counts on people not knowing the difference.

  17. 17
    Hank Roberts says:

    This isn’t journalism, where you can deny — and disprove — falsehood.

    This is politics. They want you to deny it.

    The prissy filter here won’t allow this as an excerpt. Read it:

  18. 18
    MARodger says:

    I might have missed some, but I count 135 GHCN stations above 63ºN. Homewood flaunts the records of just 17 of these stations, with two more on the Scottish isles. Some of the 135 stations will have significantly incomplete records. Likely some not being flaunted by Homewood will have adjustmments that increase warming trends above the the trends seen in the raw data. However, for the record, the longitude/latitude of these 135 stations are plotted here.

  19. 19
    John Mashey says:

    Sad to say, “journalist” is a term oft abused, like “skeptic.”
    Pseudoskeptic has a long history, as per Pseudoskeptics are not skeptics, but I’m not sure if there’s a well-accepted term for writers like Booker (or Ridley)

    Advocacy journalist or pseudoj ournalist might fit.
    Google pseudo journalism, getting among others Pseudo-Journalists Betray the Public Trust, at LA Times.

  20. 20
    notarobot says:

    Having traced the origins of the widely-reported “scandal” to the blog mentioned above. there were also allegations about leaving-out climate data from the 1940’s period. Can anyone here please respond to those claims?
    And by widely-reported I am referring to this topic being featured on Drudge and many international news outlets.

  21. 21
    Euan Mearns says:

    Roger Andrews has a post on my (our) blog that summarises 800 raw (un homogenised) temperature records.

    How Hemispheric Homogenization Hikes Global Warming

    This painstaking work has been done over many years. Roger finds that in N hemisphere his selected subset of raw records matches GISS very well. But in the S hemisphere that match is not there at all. There seems to be a large number of S hemisphere records where the distant past has been cooled or the recent past warmed relative to the raw records.

    Furthermore Roger finds that the temperature gradient through GISS N hemisphere is IDENTICAL to the gradient through the S hemisphere which I find highly suspicious / bordering on physical impossible I’d expect N and S hemispheres to respond quite differently to both natural and Man forced warming.

    I sent email to Gavin Schmidt yesterday asking for an explanation, awaiting a reply. We are simply seeking truth and will readily accept an acceptable explanation ;-)

    PS cross posted this comment to my own blog.

  22. 22
    Russell says:

    How dare you disparage the greatest living Englishman !

    He writes like an angel- it’s just the content that presents problems.

  23. 23
    Tony Weddle says:

    Wow, the Telegraph article is getting lots of attention. Phil Plait, over at Bad Astronomy, has also weighed in on the matter. Sadly, Telegraph “readers” will never see the real story.

  24. 24
    John Benton says:

    Why do you cherry pick data for Svalbard only back to 1980? Let’s see the data from 1900.

  25. 25

    #24–And why do you choose a pejorative verb in order to ask what might otherwise be a reasonable question?

    You might at least pretend to ask in good faith.

  26. 26

    #22–Even greater than Oliver Goldsmith, apparently, as Goldsmith took two separate operations to accomplish what Booker does in just one.

  27. 27
    MARodger says:

    Euan Mearns @17.
    I think you’re gonna have to talk me through this coz I’m missing something here. You told your chum Roger “it should be obvious that the S hemisphere has been homogenised to match a nearby set of stations – the N hemisphere.” Why? What on Earth would make any sane person come to such a conclusion?

    You chum Roger should be perhaps congratulated for his herculean efforst at creating a global temperature series from the ‘raw’ GHCN station data. The global result does appear to exagerate the adjustments a little, perhaps 50% too high, but that’s not so bad. The graphic here uses data from a more authoritative source (but not meant to be a public one – the URL is now defunct).
    Roger does however make exactly the same sort of mental leaps that you appear to. He says in his blog post “Is there any chance this added warming is real? None.”
    Why ‘none’? Is this what we are supposed to conclude from Roger’s Figure 7? I don’t see why. So what if GISS SH land temperatures show a 1.3ºC rise 1890-2010, CIMP5 multi-model means for SH* shows a rise of 1.0ºC 1890-2010 and Roger’s rather inaccurate efforts show a rise of 0.6ºC 1890-2010?
    (*I assume this is SH. It sure ain’t SH land. Have a butchers at IPCC AR5 Figure 10.7 if you’re not convinced. Which continent is the one causing the low average rise in Roger’s CMIP5 trace? South Pacifica, Indiania and South Atlantis, probably.)

    Surely a more sensible comparison for the GISS adjustment methods would be to compare GISS SH land temperatures with the independently-derived BEST SH land temperatures which shows a rise 1890-2010 for 1.3ºC. Snap!!

    So where have I gone wrong? Was there some inference I failed to make? Or was my mistake bothering to visit the Mearns/Andrews website?

  28. 28
    Nick Odoni says:

    Booker’s approach to my mind is hardly better than someone clutching at a piece of seaweed to make a weather forecast. I think those of us who are serious about the science and what we think it shows should ask – demand? – a right of reply to Booker’s Telegraph article. I find it exceedingly disappointing that a paper such as the S.T. should allow him to write such stuff in the first place, but if that cannot be avoided then we must have a right to set the record straight and put down the facts as we see them.

    Okay, such an article might have to have some figures and graphs in it, but so what? The whole of the modern world relies heavily – one way or another – on science and technology, and on people handling difficult numbers and maths, and sometimes struggling with counter-intuitive concepts. But it’s just because of that detail and complexity that planes don’t tend to fall out of the sky that often (the few exceptions noted), and ships are mostly able to withstand great storms at sea without breaking up (the exceptions again noted), and we can put satellites into orbit, or probes onto Mars, or build lasers to perform micro-surgery, or derive a new fertilizer to boost crop yields or clear up oil spills, and so on and so on … Climate modelling is all part of the same great endeavour.

    One thing I would add is that I think part of Booker’s view is probably related to his suspicion, rightly or wrongly, that many scientists and ‘supporters’ of global warming are left-wing, anti-capitalists or similar. Quite possibly some of the more vocal commentators are, but I suspect also that many are not. I do not myself like the capitalist system and think there is a hell of a lot wrong with it, but, as with Churchill’s views on democracy, I think it sure as hell beats the alternatives. So, as both a scientist and someone who holds views from a right-wing perspective (whatever the latter really means), I think we should get our reply in to the D.T./S.T. as soon as possible, and make quite clear that what Booker is writing is rubbish, and dangerous rubbish at that.

  29. 29
    Euan Mearns says:

    M A Rodger

    “it should be obvious that the S hemisphere has been homogenised to match a nearby set of stations – the N hemisphere.” – I said that to Roger rather tongue in cheek via email, was a bit surprised that he stuck it in his post. Though we both have a sense of humour. I personally know next to nothing about homogenisation and am simply the messenger on this occassion. Evidently homogenisation works by comparing adjacent records and applies adjustments to make them more similar. And so when I see a set of records for the S hemisphere that show identical warming to the N hemisphere and am made aware that all these records are adjusted to smooth out differences it raises suspicion in my mind. Especially when the default position should be that the hemispheres should differ – should they not?

    In an earlier post Roger outlined a detailed example for Alice Springs where a raw record that had absolutely flat temperature was adjusted so that it then produced a warming trend. This makes me immediately suspicious. And it is multiple examples such as this that Paul Homewood and Christopher Booker are highlighting. Is there a remote possibility, for example, that the guys on the ground in Alice had already adjusted the records to correct for station moves etc and that applying homogenisation is correcting an already corrected record? I’m a geochemist by training (so is Roger in a sense) and we are both used to analysing large complex data sets. If we had bad data we would simply not use it. Hence Roger’s approach of selecting what he thought were good records and excluding the bad ones. I’m kind of surprised / shocked to learn that wholesale and automated corrections are applied to raw data.

    If I read the Cowtan chart you link to correctly, then it seems to show that temperature adjustments to raw data are overwhelmingly positive and result cumulatively in +0.3˚C warming since 1900. Why is it that the adjustments are so heavily skewed positive? That’s a genuine question, but if you are unable to answer in a civil and professional way, then please don’t bother answering at all.

    And you may think you have answered my main query, but you haven’t. Note that Roger points out that this discrepancy has no major impact on the over global temperature picture. But it is an interesting anomaly which I believe merits and explanation.

    1) Why do GISS N and S hemispheres show identical warming? (do they indeed show identical warming or has Roger made a mistake?)

    [Response: They don’t. GISTEMP LOTI: , Met stations index zonal mean changes can be seen here. – gavin]

    2) Why does Roger’s N hemisphere analysis match GISS almost exactly and the S hemisphere not?

    [Response: He can put out his code and data as is available for GISTEMP and people might be able to help him out. But without that, no-one has a clue. – gavin]

    Enquiring minds would want to know the answers to put sceptical minds at rest.

  30. 30
  31. 31
    Pragmatist says:

    Simple question: Is anybody going to explain the specific adjustments made by the NCDC algorithm for the temperature records from the 19 stations in Paul Homewood’s analysis? Until that happens, the firestorm will continue.

  32. 32

    The Arctic is way more complex a job than for cherry pickers. There are some serious dynamical meteorological changes which may distort the temperature trend in some regions. If the winter near surface lapse rate changes from stable to adiabatic, the cloud system changes, winter becomes warmer and summer cooler, the end result is still warming but appears different. Like now, for instance, Arctic sea ice extent is nearing all time low , despite a much larger 2014 September minima than 2012. How is a newspaper chap suppose to understand that? Even more how can he conclude that there is a scandal when not grasping the entire subject well? It may be still very cold in the Arctic, but what is missed is the location of the cold,
    rather nearer the sub-Arctic below 65 degrees North. Again a dynamical change from the presence of thinner sea ice. But it is quite obvious why contrarian journalists attack the Arctic temperature record. It is, after all a place devoid of any thermometer misreadings or pavements. However if any journalist wants a simple story proving Global Warming beyond a shadow of a doubt, I suggest joining the paparazzi industry.

  33. 33
    William says:

    Or was my mistake bothering to visit the Mearns/Andrews website?

    Yes, that was it. Euan is the sort of person who assumes that if his results and those of climate science differ, it is not because he could be wrong or doesn’t understand but that climate science has failed. Euan has been blogging for a long time and one of the skills he has acquired is plotting graphs with misleading scaling. Point this out and you’ll get banned from commenting and he’ll lie about you to justify the ban. But his failings are easily understood when you consider that he heaps scorn on Skeptical Science but considers WUWT only “variable” and to have some “top notch” articles. He should be pitied.

  34. 34
    Euan Mearns says:

    Gavin, thanks for the response as it stands. As I mentioned already I am no expert in this particular field. I have asked Roger to respond and to make his spread sheet available.

  35. 35
    Killian says:

    Some background: Mearns was an Oil Drum contributor. We went round a few times about FFs depleting before climate change could become dangerous.

    He was a strong supporter of work on coal resources by Prof. David Rutledge, published on theoildrum, that supposedly showed climate concerns just weren’t a big deal. He was wrong, obviously. Did his own work on it, too, using the same model Rutledge did, summarized here:

    Luis De Sousa and Euan Mearns (2008) took an approach similar to the papers by Brecha and Rutledge, but arrived at somewhat different conclusions in terms of policy recommendations. For estimating emissions, they use a model that they developed earlier and that they term “Olduvai Revisited” (2008). The model is based on resource estimates and forecasts which assume “bell shaped” behavior of the production curves. They find a global peak for fossil fuel production by 2018. Using the MAGICC software package, they find that, for this scenario, CO2 concentrations should not rise over 450-500 ppm and that temperatures should not rise over 1 deg. C. De Sousa and Mearns conclude that fossil fuel decline will keep CO2 concentrations below levels that are or were considered dangerous by climate experts and that there is no need to burden the OECD and non-OECD countries with artificial measures to mitigate emissions to achieve this end.

    One of the major errors at the time was the stubborn use of 3C as their sensitivity and excluding Earth System sensitivity. I suggested they at least run their models *also* at 4.5C and 6C, but they refused to do so. Why? Good question.

    Anywho, that history underlies his current observations and self-described “…seeking truth…” questions. Perhaps he has moved past this stance and come to understand the importance of climate change…? He does acquiesce to the NH numbers, after all.

    And y’all think ahm jes an ignint tree hugger.


  36. 36
    Euan Mearns says:

    And in response to William @ 2:37

    He is not banned to commenting on my blog, his comments are simply subject to moderation as mine are here. He accuses me of lying here. The post to which William refers is this one:

    Northern Hemisphere Snow Cover

    I thought I played this pretty well straight down the middle of the fair way. William made a ton of critical accusations about the post and then admitted he hadn’t even read it. It turns out that some of the Y-axis labels on my charts are wrongly labeled, the million is missing by accident/mistake. The scaling thing is a matter of taste. I use the same scale on all charts. Good practice in many circles.

  37. 37
    Euan Mearns says:

    @ killain 5:40

    Yes, at The OIl Drum our traffic peaked at about 180,000 unique visits per day. For the last couple of years I was one of 4 principles running that site. I wrote the last few posts. The second last post was an obituary to my friend and environmental campaigner Randy Udall who would be arguing like hell against me right now, in his house in an incredibly well informed and civil manner.

    The issue of FF depletion in relation to climate models is incredibly important. In fact it lies at the heart of the whole debate. In around 2010 myself and Dave Rutledege were invited to met Hadley where we gave a couple of talks on FF resource constraints on CC and policy. One to two hundred people attended.

    I have modified my views substantially from the Olduvai Revisited days, but continue to update those views. $50 / bbl – where does that leave the US shale industry and tar sands production? But we sure as hell have a hunger for those C-H bonds that goes beyond what I imagined.

    I’m actually biased negative on shale, tar sands and insitu combustion of FF knowing that this opens up a whole pandoras box of CO2. Hence, I am strongly pro-nuclear as the lesser of energy evils.

    There is an unmeasurable probability that we are close to peak oil now. The only thing that has dented CO2 emissions in recent history was the finance crash of 2008. All my charts look the same now as then.

  38. 38
    MARodger says:

    Euan Mearns @29.
    So, messenger, perhaps we take this one step at a time. Roger’s figure 5 is fundamental to your assertion that GISS NH & SH temperature series “show identical warming? (do they indeed show identical warming or has Roger made a mistake?)”
    This figure 5 purports to plot “GISS North and South Hemisphere surface air temperature time series.” Air temperature means land temperature and I assume this is what is plotted. It is a good match for NCDC annual NH & SH land temperature anomalies which I have to hand. What is strange is the anomally base period which seems to be way back in 1900. Presumably the data has been shifted vertically on the graph for some reason.
    And so to these trend lines. They do appear very similar in slope for NH & SH in Roger’s figure 5. That they are so similar in slope is surely the crux of your assertion. I assume they are linear regressions. The GISS (&NCDC) data stretches from 1880-2014 and the NCDC NH & SH trends calculated over that period are far from being that similar. So for what period do Roger’s trends apply? And why choose that period?

  39. 39
    Mal Adapted says:


    Simple question: Is anybody going to explain the specific adjustments made by the NCDC algorithm for the temperature records from the 19 stations in Paul Homewood’s analysis? Until that happens, the firestorm will continue.

    Surely you don’t think any amount of explanation will put an end to the firestorm, “Pragmatist”? If simple questions like yours were inspired by an innocent quest for understanding, the fire would have been out 20 years ago. The flames are fanned by pragmatists who will lose money if fossil-fuel use is curtailed. From their PoV they’d be fools not to. From the PoV of the rest of us, who stand to lose more than just money if business continues as usual, we’d be fools to trust the so-called “pragmatists”

  40. 40
    Roger Andrews says:


    Let me try to simplify things a bit. The issue has nothing to do with methodologies, code or what’s on my spreadsheet. It’s simply that the GISS homogeneity adjustments add no appreciable amount of warming to the unadjusted GHCN records in the NH but more than double the amount shown by the unadjusted records in the SH. Since I can find no justification for this added warming I have provisionally concluded that it was manufactured by the GISS homogenization algorithm. However, I don’t have the data I need to say for sure whether it was or not, so I’m hoping you will be able to provide a more definitive explanation.

  41. 41
    Lawrence Coleman says:

    The fact that it was published in the first case means the damage is already done. The denialists only cherry pick and read the articles they want that substantiate they flawed argument thus reinforcing their ignorance. It is the laypeople perched on the edge who might agree that temp seems to be rising but are not convinced that humans are behind it and might believe that some scientists play by different rule books. They are the ones who might get duped by Christopher Booker’s lack of understanding (or more sinister motives). These rogue journalists must get held accountable for their deceiving articles. However no penalty ever gets enforced upon them and so this will continue to happen into perpetuity.

  42. 42
    toby52 says:

    Zeke #30

    Seems like the old denier bait and switch.

    Start by yowling “Phrawd!”, then when you get attention go into the change-the-goalposts routine, viz:

    “I make no claims about the effect of (temperature) adjustments globally,” Homewood said. “I feel that by identifying specific examples, we have moved the debate forward by challenging how adjustments work in practice, and whether we can always rely on them.”

  43. 43
    Ray Ladbury says:

    Roger Andrews,
    The graph Gavin shows inline to #29 shows far more warming in the NH than the SH. Since you are contending that the SH trend is more than doubled by the adjustments, you seem to be saying the SH has basically experienced no warming at all.

    Given that we are seeing dramatic retreat of glaciers in the SH as in the NH, how reasonable do you think this assertion is?

  44. 44
    William says:

    [edit – discussions about comment moderation are boring. Discussions about comment moderation on another blog are doubly so. Please focus on science issues here. Thanks. ]

  45. 45
    Euan Mearns says:

    @ Gavin, the chart you present I believe is LOTI while Roger’s query is specifically about the land air thermometer record. I wonder if this screen cap from the GISS temp home page might show something similar to Roger’s observation? The chart used for the link does not appear inside.

    [Response: That appears to have generated more than a decade ago, and in any case it is also of the LOTI index. There have been many updates in coverage, data sources and algorithms since then and so it is not in the least bit relevant to your claims this week. – gavin]

    @ MA Rodger

    It is a good match for NCDC annual NH & SH land temperature anomalies which I have to hand. What is strange is the anomally base period which seems to be way back in 1900. Presumably the data has been shifted vertically on the graph for some reason.

    I agree that it is not ideal that Roger may have used a different base period etc. But you seem to confirm that Roger’s Figure 5 matches the data you have to hand. If someone could post a chart of the official GISS land air temperatures with both N and S hemispheres that would be helpful and a step forward.

    That would leave us with two things:

    1) GISS N and S land air temperature time series are very similar. I feel intuitively they should not be.
    2) Roger’s hand selected sub-set of uncorrected S hemisphere records differ to the N hemisphere and to GISS S hemisphere. The homogenisation procedure appears to have introduced bias to the S hemisphere.

    The other thing that needs to happen to advance this is obviously for Roger to let us see his spread sheet, or at least have a full list of the 800 stations he has selected.

  46. 46

    “Until that happens, the firestorm will continue.”

    Oh, sure. But I’ll be using the other teapot until it cools down.

  47. 47
    Kit says:


    He should be pitied.

    Indeed. Euan seems to believe he and Roger are acting rationally and without bias, yet it’s clear that neither are. I don’t think they realise it. Poor ducks :(

  48. 48
    MARodger says:

    Euan Mearns @45.
    I have reproduced Roger’s figure 5 here and, except for the curious vertical displacements and the pruning of the start and the end of the record, Roger’s figure 5 is what ir says on the box.
    The linear trends 1890-2010 are also as per plotted in Roger’s figure 5 but there is a rather curious thing with these trends. As I said @38, the 1880-2014 trends for NH & SH are significantly different. That the 1890-2010 trends are almost the same (within 2%) demonstrates that these trends are sensitive to the selection of start-point & end-point. I put the GISS data through a spreadsheet and obtained a measure of how sensitive these trends are, as graphed here. You will see that if for some reason you want to get NH & SH trends identical, a really good start-point is, guess what, 1890. And while the best end point is 2007, 2010 isn’t a bad substitute.
    I would suggest that youir assertion @29 “GISS N and S hemispheres show identical warming” is wholly start and end point dependent.

    Of course, all this blather only makes any sense if there is some reason to doubt the homogensation process carried out on GHCN data. So I repeat the comment @27 but this time the illustrative graphic compares SH Land temperature anomalies from GISS & NCDC and also BEST which do not use GHCN homogenisation and as well (for completeness) UAH which doesn’t even use met data (although it is measuring a different part of the atmosphere). This graphic illustrates conclusively that GISS, NCDC, BEST & UAH show the same basic trend in SH Land temperatures. If, as indicated by Roger’s figure 4, GHCN homogenisation does add ~0.6ºC/century to SH Land warming trends (which is over half the total trend, so I am sceptical), the resulting temperature series are not inconsistent with other temperature series derived using independent methods.
    So are BEST (& UAH) also screwing up their SH Land temperatures in some bizarre conicidence? Or can we draw a line under this business?

  49. 49
    Hank Roberts says:

    > euanmearns …/wp-content/uploads/2015/02/GISS_temp_N_S.png
    [Response: That appears to have generated more than a decade ago, and in any case it is also of the LOTI index….. — Gavin]

    Hm. Mr. Mearns presents a screen grab dated “2015/02” stripped of the axis labels for time or temperature as though it’s new. Just an accident no doubt.
    Where did Mr. Mears get that? Who did he trust to think it’s current?

    Fortunate that Gavin recognized it came from “more than a decade ago” eh?

    One should be very careful about rebunking this sort of picture.
    “Trust, but verify” — R. Reagan

  50. 50
    Hank Roberts says:

    P.S. — OOPS! image search finds images resembling that chart Mr. Mearns posted: These

    The first one is the thumbnail at the upper right corner of the GIS page used for navigation:

    So he can complain he was misled, assuming he never clicked past the home page for GISS and that he took the click-through navigation button picture as actually presenting the most up-to-date science.

    As that button lacks any labels on the axis, and is ancient, perhaps a note to the webmaster?