Sunday 5 February 2017

David Rose's alternative reality in the Daily Mail

Peek-a-boo! Joanna Krupa shows off her stunning figure in see-through mesh dress over black underwear
Bottoms up! Perrie Edwards sizzles in plunging leotard as Little Mix flaunt their enviable figures in skimpy one-pieces
Bum's the word! Lottie Moss flaunts her pert derriere in a skimpy thong as she strips off for steamy selfie

Sorry about those titles. They provide the fitting context right next to a similarly racy Daily Mail on Sunday piece of David Rose: "Exposed: How world leaders were duped into investing billions over manipulated global warming data". Another article on that "pause" thingy that mitigation skeptics do their best to pretend not to understand. For people in the fortunate circumstances not to know what the Daily Mail is, this video provides some context about this Murdoch "newspaper".

[UPDATE: David Rose' source says in an interview with E&E News on Tuesday: “The issue here is not an issue of tampering with data”. So I guess you can skip this post, except if you get pleasure out of seeing the English language being maltreated. But do watch the Daily Mail video below.

See also this article on the void left by the Daily Mail after fact checking. I am sure all integrityTM-waving climate "skeptics" will condemn David Rose and never listen to him again.]

You can see this "pause" in the graph below of the global mean temperature. Can you find it? Well you have to think those last two years away and then start the period exactly in that large temperature peak you see in 1998. It is not actually a thing, it is a consequence of cherry picking a period to get a politically convenient answer (for David Rose's pay masters).

In 2013 Boyin Huang of NOAA and his colleagues created an improved sea surface dataset called ERSST.v4. No one cared about this new analysis. Normal good science. One of the "scandals" Rose uncovered was that NOAA is drafting an article on ERSST.v5.

But this post is unfortunately about nearly nothing, about the minimal changes in the top panel of the graph below. I feel the important panel is the lower one. It shows that in the raw data the globe seems to warm more. This is because before WWII many measurements were performed with buckets and the water in the bucket would cool a little due to evaporation before reading the thermometer. Scientists naturally make corrections for such problems (homogenization) and that helps make a more accurate assessment of how much the world actually warmed.

But Rose is obsessed with the top panel. I made the graph extra large, so that you can see the differences. The thick black line shows the new assessment (ERSST.v4) and the thin red line the previously estimated global temperature signal (ERSST.v3). Differences are mostly less than 0.05°C, both warmer and cooler. The "problem" is the minute change at the right end of the curves.

The mitigation skeptical movement was not happy when a paper in Science in 2015, Karl and colleagues (2015), pointed out that due to this update the "pause" is gone, even if you use the bad statistics the mitigation skeptics like. As I have said for many years now about political activists claiming this "pause" is highly important: if your political case depends on such minute changes, your political case is fragile.

In the mean time a recent article in Science Advances by Zeke Hausfather and colleagues (2016) now shows evidence that the updated dataset (ERSSTv4) is indeed better than the previous version (ERSSTv3b). They do so by comparing the ERSST dataset, which comes from a large number of data sources, with data that comes only from only one source (buoys, satellites (CCl) or ARGO). These single-source datasets are shorter, but without trend uncertainties due to the combination of sources. The plot below shows that the ERSSTv4 update improves the fit with the other datasets.

The trend change over the cherry-picked "pause" period were mostly due to the changes in the sea surface temperature of ERSST. Rose makes a lot of noise about the land data, where the update was inconsequential. As indicated in Karl and colleagues (2015) this was a beta-version dataset. The raw data was published; that is the data of the International Surface Temperature Initiative (ISTI) and the homogenization method was published. The homogenization method works well; I checked myself.

The dataset itself is not published yet. Just applying a known method to a known dataset is not a scientific paper. Too boring.

So for the paper NOAA put a lot of work into estimating the uncertainty due to the homogenization method. When developing a homogenization method you have to make many choices. For example, inhomogeneities are found by comparing one candidate station with multiple nearby reference stations. There are settings for now many stations and for how nearby the reference stations need to be. NOAA studied which of these settings are most important with a nifty new statistical method. These settings were varied to study how much influence that has. I look forward to reading the final paper. I guess Rose will not read it and stick to his role as suggestive interpreter of interpreters.

The update of NOAA's land data will probably remove a precious conspiracy of the mitigation skeptical movement. While, as shown above, the adjustments reduce our estimate for the warming of the entire world, the adjustments make the estimate for the warming over land larger. Mitigation skeptics like to show the adjustments for land data only to suggest that evil scientists are making global warming bigger.

This is no longer the case. A recommendable overview paper by Philip Jones, The Reliability of Global and Hemispheric Surface Temperature Records, analyzed the new NOAA dataset. The results for land are shown below. The new ISTI raw data dataset shows more warming than the previous NOAA raw data dataset. As a consequence the homogenization now does not change the global mean appreciably any more to arrive at about the same answer after homogenization; compare NOAA uncorrected (yellow line) with NOAA (red; homogenized).

The main reason for the smaller warming in the old NOAA raw data was that this smaller dataset contained a higher percentage of airport stations. That is because airports report their data very reliably in near real time. Many of these airport stations were in cities before and cities are warmer than airports due to the urban heat island effect. Such relocations thus typically cause cooling jumps that are not related to global warming and are removed by homogenization.

So we have quite some irony here.
Still Rose sees a scandal in these minute updates and dubs it Climategate 2; I thought we were already at 3 or 4. In this typical racy style he calls data "wrong", "rogue", "biased". Knowing that data is never perfect is why scientists do their best to assess the quality of the data, remove problems and make sure that the quality is good enough to make a certain statement. In return people like David Rose simultaneously pontificate about uncertainty monsters and assume data is perfect and then get the vapors when updates are needed.

Rose gets some suggestive quotes from an apparently disgruntled retired NOAA employee. The quotes themselves seem to be likely inconsequential procedural complaints, the corresponding insinuations seem to come from Rose.

I thought journalism had a rule that claims by a source need to be confirmed by at least a second source. I am missing any confirmation.

While Rose presents the employee as an expert on the topic, I have never heard of him. Peter Thorne, who worked at NOAA, confirms that the employee did not work with surface station data himself. He has a decent publication record, mainly on satellite climate datasets of clouds, humidity and radiation. Ironically, I keep using that word, he also has papers about the homogenization of his datasets, while homogenization is treated by the mitigation skeptical movement as the work of the devil. I am sure they are willing to forgive him his past transgressions this time.

It sounds as if he made a set of procedures for his climate satellite data, which he really liked, and wanted other groups in NOAA to use it as well. Was frustrated when others did not prioritize enough updating their existing procedures to his.

For David Rose this is naturally mostly about politics and in his fantasies the Paris climate treaty would not have existed with the Karl and colleagues (2015) paper. I know that "pause" thingy is important for the Anglo-American mitigation skeptical movement, but let me assure Rose that the rest of the world considers all the evidence and does not make politics based on single papers.

[UPDATE: Some days you gotta love journalism: a journalist asked several of the diplomats who worked for years on the Paris climate treaty, they gave the answer you would expect: Contested NOAA paper had no influence on Paris climate deal. The answers still give an interesting insight into the sausage making. What is actually politically important.]

David Rose thus ends:
Has there been an unexpected pause in global warming? If so, is the world less sensitive to carbon dioxide than climate computer models suggest?
No, there never was an "unexpected pause." Even if there were, such a minute change is not important for the climate sensitivity. Most methods do not use the historical warming for that and those that do consider the full warming of about 1°C since the 19th century and not only short periods with unreliable, noisy short-term trends.

David Rose:
And does this mean that truly dangerous global warming is less imminent, and that politicians’ repeated calls for immediate ‘urgent action’ to curb emissions are exaggerated?
No, but thanks for asking.

Post Scriptum. Sorry that I cannot talk about all errors in the article of David Rose, if only because in most cases he does not present clear evidence and because this post would be unbearably long. The articles of Peter Thorne and Zeke Hausfather are mostly complementary on the history and regulations at NOAA and on the validation of NOAA's results, respectively.

Related information

Buzzfeed (October 2017): This Is How A Bogus Climate Story Becomes Unstoppable On Social Media

New York Times (September 2017): British Press Watchdog Says Climate Change Article Was Faulty

2 weeks later. The nailing New York Times interviewed several former colleagues of NOAA retire Bates: How an Interoffice Spat Erupted Into a Climate-Change Furor. "He’s retaliating. It’s like grade school ... At that meeting, Dr. Bates shouted that Ms. McGuirk was not trustworthy and belonged in jail, according to an internal log ..." Lock her up, lock her up, ...

Wednesday. The NOAA retiree now says: "The Science paper would have been fine had it simply had a disclaimer at the bottom saying that it was citing research, not operational, data for its land-surface temperatures." To me it was always clear it was research data, otherwise they would have cited a data paper and named the dataset. How a culture clash at NOAA led to a flap over a high-profile warming pause study

Tuesday. is a balanced article from the New York Times: Was Data Manipulated in a Widely Cited 2015 Climate Study? Steve Bloom: "How "Climategate" should have been covered." Even better if mass media would not have to cover office politics on archival standards fabricated into a fake scandal.

Also on Tuesday, an interview of E&E News: 'Whistleblower' says protocol was breached but no data fraud: The disgruntled NOAA retiree: "The issue here is not an issue of tampering with data".

Associated Press: Major global warming study again questioned, again defended. "The study has been reproduced independently of Karl et al — that's the ultimate platinum test of whether a study is to be believed or not," McNutt said. "And this study has passed." Marcia McNutt, who was editor of Science at the time the paper was published and is now president of the National Academy of Sciences.

Daily Mail’s Misleading Claims on Climate Change. If I were David Rose I would give back my journalism diploma after this, but I guess he will not.

Monday. I hope I am not starting to bore people by saying that Ars Technica has the best science reporting on the world wide web. This time again. Plus inside scoop suggesting all of this is mainly petty office politics. Sad.

Sunday. Factcheck: Mail on Sunday’s ‘astonishing evidence’ about global temperature rise. Zeke Hausfather wrote a very complementary response, pointing out many problems of the Daily Mail piece that I had to skip. Zeke works at the Berkeley Earth Surface Temperature project, which produces one of the main global temperature datasets.

Sunday. Peter Thorne, climatology professor in Ireland, former NOAA employee and leader of the International Surface Temperature Initiative: On the Mail on Sunday article on Karl et al., 2015.

Phil Plait (Bad Astronomy) — "Together these show that Rose is, as usual, grossly exaggerating the death of global warming" — on the science and the politics of the Daily Mail piece: Sorry, climate change deniers, but the global warming 'pause' still never happened

You can download the future NOAA land dataset (GHCNv4-beta) and the land dataset used by Karl and colleagues (2015), h/t Zeke Hausfather.

The most accessible article on the topic rightly emphasizes the industrial production of doubt for political reasons: Mail on Sunday launches the first salvo in the latest war against climate scientists.

A well-readable older article on the study that showed that ERSST.v4 was an improvement: NOAA challenged the global warming ‘pause.’ Now new research says the agency was right.

One should not even have to answer the question, but: No, U.S. climate scientists didn't trick the world into adopting the Paris deal. A good complete overview at medium level.

Even fact checker Snopes sadly wasted its precious time: Did NOAA Scientists Manipulate Climate Change Data?
A tabloid used testimony from a single scientist to paint an excruciatingly technical matter as a worldwide conspiracy.

Carbon Brief Guest post by Peter Thorne on the upcoming ERSSTv5 dataset, currently under peer review: Why NOAA updates its sea surface temperature record.


  1. If we go back and consider the actual substance of "Climategate," we see that it was only some minor FOIA transgressions and a few arguably embarrassing comments made in what had been thought to be private. It only got traction because 1) Phil Jones, presumably on the advice of some idiot in the UEA press office, completely clammed up rather than provide the press with the perfectly reasonable explanations available to him, and 2) the Guardian, fresh off its success in exposing the phone hacking scandal (a large and real one), detected in Phil's silence a potential cover-up and pulled out all the stops in order to flush out the truth. (Plus we should not discount the prurience value of hacked emails, as we saw again last year.) Then, because it was the climate science-friendly "left" Guardian pushing it, the scandal acquired enough seeming credibility to get the various inquiries launched. British tabloids in general dream of this sort of thing, thus Rose's ridiculous attempt to launch "Climategate 2."

    Erratum: You want "transgressions" rather than "digressions."

  2. Worth mentioning David Rose's illustrative plot

    The difference is almost entirely due to the different anomaly bases (HADCRUT 1961-90 vs NOAA 1901-2000).

  3. Victor

    John bates has a comprehensive article up over at Judith.

    He seems to have excellent credentials so your comments on his piece would be interesting


  4. Note, the Mail is not a Murdoch paper as stated.

  5. Climategate may have been the Russian petrostate's first successful use of hacking to influence world politics.

  6. Tony, I would rather no comment on people, so I will not publish your second comment. If only because that would be quoted elsewhere to pretend I only attacked the person.

    I never expected the headlines in the Daily Mail song to be real; it is a song. If the Daily Mail becomes an official satire magazine, I will not complain anymore about the content being wrong.

    Yes, der Spiegel is a decent journal in Germany. I hope you are willing to put yourself in the position of Germany, who already had a similar problem, which devastated the country and much of Europe and the world. To prevent this from happening again, Germans are well educated to recognize the start of fascism. It is thus logical that the immune system of Germans for this kind of problems responds faster, especially faster than that of fearful people who do not mind a bit of authoritarian leadership that much and somehow do not fear how it ends.

  7. I don't know if you're aware of the history of the Mail.

    If not, this may help.

    Note that the Mail is still owned by the Rothermere family today.

  8. Just listened to the Daily Mail song again. Hard do believe that people really believed that "I am not racist, but" is a real DM front page.

    More, also satirical, context on The Daily Mail provides the video: Charlie Brooker's Daily Mail Island. I guess Daily Mail readers will not like it.

  9. Victor

    For what its worth I thought the daily mail song amusing, lf unrelated to reality

    For your personal amusement you might enjoy this trump spoof

    My county is Devon. The spoofer is from next door Cornwall. It is obviously very quiet there without the tourists


  10. tonyb: reading Karl et al would have answered Bates' own question about using 90% confidence intervals, rather than 95%.

    Hint: comparison with IPCC report, which happens to use 90% CI (as noted in Karl et al).

    I am also rather surprised Karl is accused by Bates of having pushed his co-authors to make decisions that would increase warming, since Karl was not directly involved in the underlying datasets, ERSST.v4 and the ISTI database. There are no specific decisions in Karl et al that change anything about these two datasets and thus the observed warming.

  11. Lars Karlsson, I would prefer not to publish your link to the blog of Curry. I am sure people interested in misinformation will be able to find it and prefer not to send people to deceptive information. She also explicitly refuses to engage with my arguments.

  12. Huh, according to Bates was demoted by Karl in 2012. So we're in familiar "disgruntled former employee" territory.

  13. It seems that testing the pairwise homogenization method used by GHCN should be possible: just (!) set up a data set simulating the real data with random variation, occasional biases (site changes, TOBS changes etc) plus a known trend and see how well the method does at removing the known biases. Has this been done?
    Regards, William

  14. Exactly, that is very easy. NOAA naturally did so. There was also large study with many participants including NOAA's pairwise method doing exactly that, I was the first author of that study. That was how I started working on this topic.

    There is a new similar study coming up in the International Surface Temperature Initiative.

  15. A detailed refutation of Bates's claims about the process, by an indepedent scientist who worked with NOAA and co-authored all the relevant papers underlying Karl et al 2015.

  16. But Victor, Judy points out that you are unprofessionally smearing The Daily Mail. That is truly unheard of. Are you not ashamed of attacking such a reputable and esteemed newspaper?

  17. Victor, as everyone knows, sometime around 2013 there started to be this meme of a hiatus in global warming, because the temperature record for several years was tracking below the aggregate model projections. It became noticeable enough that climate researchers started to explain it. This article by Carl Mears is one example. Mann and Steinman published on natural cooling in the Pacific as an explanation for the discrepancy.

    When Karl et al published their 2105 paper, they specifically presented it as addressing this issue: “Much study has been devoted to the possible causes of an apparent decrease in the upward trend of global surface temperatures since 1998…..These results do not support the notion of a “slowdown” in the increase of global surface temperature.”
    You know all of this, of course. I bring this up as the background to this question: If the prospective temperature record had instead been tracking the models fairly closely, would Tom Karl have gone through the same process of retrospectively discovering and correcting “artifacts of data biases” in the NOAA temperature record? Is there a risk here of post hoc data adjustments that are biased in one direction? Is this a valid question?

  18. This is probably not a good example because the difference between buoys and ships was known already in 2008 (likely earlier I was not in the field then and am not an expert on SST). Because of the fast shift from ships to buoys it became more urgent. This update of ERSSTv4 was published in 2013 and the work started earlier; probably immediately after ERSSTv3b.

    That being said, such risks naturally exists. WWII sticks out like a sore thumb. If it had been a cold period for natural reasons and the warm bias of the measurements in that time would have masked that and made the graph nice and smooth, this problem likely would have gotten less attention than now where it is a peak in the raw data at an historical time.

    Scientists should always be aware of possible biases and fight them. The ones who do that the best are most successful. I would almost wish my colleagues had large biases, that would make it easier for me to write big publications. But they are pretty good and smart.

  19. Thanks for the detailed comment. Please delete the reference to the Mail as a Murdoch paper (I know it's crossed out but 'merchants of doubt' like Rose would seize on that, ignoring the detailed scientific evidence, and use it against you).
    The term 'merchants of doubt' was coined by Naomi Oreskes in her excellent book of the same name, also made into a film (available on youtube) which I recommend to provide some historic context to this debate.

    Another source on the NOAA debate in response to the Rose article is provided by potholer54, another good source for exposing climate change piffle (and 'crabzilla!)

    Newsflash! the Mail was recently banned by wikipedia as a source for 'sensationalism' (Fox News and Russia Today remain info sources for wiki, illustrating the depths that have to be reached before such a sanction is imposed)

    re: your twitter vote on whether Rose's article is more deceptive than Kelly's, I'd opt for Rose on the basis that Kelly's organ has not yet been banned by wiki.


Comments are welcome, but comments without arguments may be deleted. Please try to remain on topic. (See also moderation page.)

I read every comment before publishing it. Spam comments are useless.

This comment box can be stretched for more space.