Temperature adjustments: Dodds, Mosher and Venema cannot be happy

As always, the climate orthodoxy possesses little understanding of how the internet works. They struggle to understand how ‘small, inconsequential issues’ seem to get magnified and blow up in their faces. There is undeniable shock at the Booker articles on temperature adjustments in the Telegraph which have over 35,000 comments and 100,000 social media ‘shares’. Here’s one old explanation (from yours truly):

…deficiencies and uncertainties in climate science are not allowed to become part of everyday discussion…

It is for this reason primarily that Trenberth’s Travesty, Mike’s Nature Trick and Jones’ Hide the Decline are all memes of the post-Climategate age. They carry memetic value because their opposites are pushed relentlessly as part the dominant paradigm, even as scientists apparently discuss doubts in private.

The current episode was set off by relentless ‘hottest year ever’ oversell perpetrated by such sources as AP’s Seth Borenstein. It led Paul Homewood to look at hot areas in the global average starting with Paraguay.

‘…if you want to properly understand an argument or debate you need to look at the primary sources’ reminds Kevin Marshall at manicbeancounter, just as Steven Mosher and a clueless Lubos have not done. Marshall traces the evolution of the story:

With the Booker story exploding, a clutch of scientists and enthusiasts decided they needed to spread their own memes. They picked up a rash of ad hoc excuses to fight the mighty Booker.

Excuse #1 – the adjustments produce no change, by Mosher with BEST data.


Messing with temperature trends strikes anyone as a problem. It did climate consensus supporter Andrew Dodds:


Dodds is everyman – he has the questions anyone would have. Going by Mosher – if an adjusted and unadjusted record are roughly the same, there’re no problems. Which means, if an adjusted record is not the same as its unadjusted progenitor, and actually its opposite in trend, there is a problem.


Numerous stations show change in sign, shape, gaps, everything … by way of adjustment. Many of them end up with a warm trend. Accordingly, there is skepticism. Which brings us to Excuse #2.

#2 – we make cooling adjustments too! – by Victor Venema and others.

To prove the adjustments don’t just warm records all over the place, Venema whips out this graph of sea temperatures, where he says ‘scientists’ cooled a steep warming (grey dashed line) to a more gentle one. Hurray!

ocean raw adj1-01

What about the Dodds criterion? The adjustments change the shape of the trend dramatically. Surely, this is a case for natural extreme skepticism.

Dodds may not know this but the global record is subject to such large-scale trend-altering adjustments at several points. With every adjustment, a piece of the climate orthodox narrative fits better with the instrumental temperature record.

Venema’s graph shows a 1910-1940 adjustment. Here’s an adjustment Phil Jones proposed for 1940-1960:


Here is the result of an adjustment Kevin Cowtan and Robert Way propose 1997 onward:



Global temperatures are not derived hands-off, they are beaten into shape. They are malleable and they accommodate everything from 19th century proxy records to sulphate aerosols to the ‘pause’. As Steven Goddard notes, ‘the fact that they can provide theoretical justifications for data tampering, tells us absolutely nothing about the correctness of what they are actually doing’.

All Booker and Delingpole have demonstrated is this. In the global warming narrative, if you adjust something, something else breaks: Dodds, Mosher and Venema cannot all be happy at the same time.




  1. Pingback: Shub’s Thoughts On The Great Debate | NOT A LOT OF PEOPLE KNOW THAT
  2. Brad

    If Mosher says the changes make no difference, why are they doing it, just to collect a paycheck? I’d like to hear BEST say – “what we do makes no difference, so we are all quitting and going to find a new career.”
    The issue of justifying your existence often hinges on what you do for a living. To admit that what you do makes no real difference in the world is a soul-wrenching experience.

  3. manicbeancounter

    Victor Venema follows the same method as ATTP, citing secondary sources.
    But there is another twist. He quotes out of date secondary sources. His article is dated 10/02/15, by which time Paul Homewood had moved from the 23 temperature stations in Bolivia and Paraguay to looking at 19 temperature stations

    Venema makes a similar claim about Australia. Linking to the vicious Hot Whopper site, he claims that there have been just two sites in Australia adjusted. Hot Whopper has just one – Rutherglen.
    There is also Bourke and Bathurst Goal – long records with raw data showing cooling trends
    Also Williamstown RAAF
    Also Deniliquin NSW
    This last, it is worth comparing the NASA GISS raw data for 1880-1992
    With the adjusted data
    Hot Whopper’s article is 27/08/14.

  4. Victor Venema

    The trend of the homogenized data of GHCNv3 and Berkeley Earth is very similar.

    That Berkeley Earth on average corrects the data less is because it has a much larger dataset and their raw dataset has a smaller bias. Also the trend of the raw dataset of the ISTI is stronger. That does not make the homogenization of BEST irrelevant. You would not know this before you remove non-climatic effects and there is more of interest that just the global mean temperature signal.

    Thus the claim that may not be “right” would be the one of Dodds. Given that the data of the global mean temperature before 1950 has been adjusted 0.1 or 0.2°C upwards, the signal has a little more curve. But given how small that difference is, you probably need to be a mitigation sceptic to make a big deal out of that.

    A far cry from the false claims of Homewood and Booker.

  5. catweazle666

    “A far cry from the false claims of Homewood and Booker.”

    Yes, you would say that, wouldn’t you…

  6. Brad

    Please explain in detail how talking about BEST having a much “larger data set and smaller bias” justifies GHCN adjustments. Two wrongly-determined trends shouldn’t make a right in anybody’s book..

    The part I really hate about government-funded climate science is that there are no repercussions for being wrong while In the private business world you would get fired. Oh wait,, you continue to get government funding to continue the cover-up until you can retire and write a book.

    Dang, where do I apply?????

  7. Jaime

    That BEST adjustment graph leaves a lot to the imagination. Is there an expanded Metadata only vs. raw for the period since 1980 and in particular since 1998?

  8. Shub Niggurath

    Jaime, good question. I asked a variant of the same: What does the record look like, with error bars, with stations requiring any form of adjustments excluded?

  9. manicbeancounter

    Compare my criticisms of Victor Venema (rightly held up in moderation due to all the links)
    VV claims corroborates GHCNv3 in general, because it is sort of in line with BEST in general. He fails to go straight to the raw data and show that the claims of Paul Homewood are false – after all what Homewood shows are a series of graphs that vaguely appear to be telling the same story in the same area.
    I did check the nine surface Paraguayan temperature stations that Homewood claimed seemed adjusted in the same way. They were all adjusted. It was to offset a consistent “anomaly” in the data. 8 of the 9 thermometers showed a sharp fall in average temperatures in the late 1960s. (The ninth had fragmentary data, but shows a similar pattern).
    Such a sudden change goes against the clear consensus of scientists that the cause of changes in temperature is due to changes in greenhouse gases. The data is therefore wrong, so was adjusted to the known apriori truths about the world. To me, with similar cooling occurring at a similar time in the Arctic it suggest that the theory does not allow for the richness and diversity of the real world.

  10. catweazle666

    “The data is therefore wrong, so was adjusted to the known apriori truths about the world.”


    Who would trust $10 thermometer data when they have $100,000,000 supercomputer computer game climate model data?

    Be reasonable!

  11. Lars

    How can they (GISS/BEST/etc.) really still get away with the old “but we have thousands of quality station records . . . so here look at our global graphs”? I say NO, because you don’t have thousands of comparable temperature records to what is being discussed in these adjustments . We all know the global construction is biased towards the last 40 years and towards the Northern Hemisphere (http://data.giss.nasa.gov/gistemp/station_data/stations.gif), and extremely sensitive to changes to rural sparse data where adjustments/homogenizations then paint very very large swathes, like the ones at hand. So the stations being discussed are very important and there aren’t “thousands” of similar records. They can make their adjustments but the overall global reconstruction graph needs error bars growing wider as we move back in time in the instrumental record (god knows how the proxies get to 0.5 degrees in resolution!) because of the increasing sparseness of long temperature records. On top of that, the error bars need to also, just for honesty’s sake, include the actual value of adjustments for these long/rural records which compound potential errors in human adjustments! And of course this debate is healthy and useful and should be increasingly vocal in the scientific community as the models diverge from reality – hmm, which inputs into the models might be the largest culprit? Perhaps the modeled/homogenized/adjusted long temperature records you’re feeding into them?

  12. catweazle666

    “the overall global reconstruction graph needs error bars growing wider as we move back in time”

    The average “climate scientist” would not recognise an error bar if you beat him over the head with it, how else do you explain hysterical claims of “hottest year evah!” when the claimed temperature increase is a good order of magnitude smaller than the claimed value? Charlatans the lot of them.

    These people are a joke in very poor taste, hopefully the work of Homewood, Goddard, Booker, Delingpole, Shub and many others who are revealing their mendacity are rapidly putting the skids under them and their cosy little overpaid sinecures.

  13. manicbeancounter

    Lars and Catweazle
    You are right to talk about error bars as being wrong, but not the nature on how wrong.
    Homewood has highlighted that there are a number of adjustment problems in the temperature record. Rather than reducing biases from the instruments – a valid thing to do – there appears to be the introduction of bias by eliminating quite valid cooling. In both Paraguay and the Arctic there seems to have been significant drops in temperature over a number of thermometers.
    If bias is introduced into the data, then all errors bars are irrelevant. You can only guess. Given that recent figures have a small bias upwards, and early twentieth century figures a bigger known bias downwards, any errors bars will be asymmetric.
    The way to give validity back to error bars is to remove all known biases in the data. But that needs to be done by people who wish to extrapolate all the scanty data is able to yield, and not by people who subscribe to a particular view of the causes of temperature change.