Skip to content

German Professor Uncovers Massive GISS Tampering

November 20, 2015

By Paul Homewood  




From No Tricks Zone:


Veteran journalist Günter Ederer* writes a piece reporting that massive alterations have been found in the NASA GISS temperature data series, citing a comprehensive analysis conducted by a leading German scientist. These results are now available to the public.



Professor Dr. Friedrich Karl Ewert. Source: University of Paderborn


Ederer reports not long ago retired geologist and data computation expert Professor Dr. Friedrich Karl Ewert began looking at the data behind the global warming claims, and especially the datasets of NASA’s Goddard Institute of Space Studies (GISS).

Ewert painstakingly examined and tabulated the reams of archived data from 1153 stations that go back to 1881 which NASA has publicly available – data that the UN IPCC uses on which to base its conclusion that man is heating the Earth’s atmosphere through the burning of fossil fuels. According to Ederer, what Professor Ewert found was “unbelievable”:

From the publicly available data, Ewert made an unbelievable discovery: Between the years 2010 and 2012 the data measured since 1881 were altered so that they showed a significant warming, especially after 1950. […] A comparison of the data from 2010 with the data of 2012 shows that NASA-GISS had altered its own datasets so that especially after WWII a clear warming appears – although it never existed.”

Ederer writes that Ewert particularly found alterations at stations in the Arctic. Professor Ewert randomly selected 120 stations from all over the world and compared the 2010 archived data to the 2012 data and found that they had been tampered to produce warming.

The old data showed regular cycles of warming and cooling over the period, even as atmospheric CO2 concentration rose from 0.03% to 0.04%. According to the original NASA datasets, Ederer writes, the mean global temperature cooled from 13.8°C in 1881 to 12.9°C in 1895. Then it rose to 14.3°C by 1905 and fell back under 12.9°C by 1920, rose to 13.9°C by 1930, fell to 13° by 1975 before rising to 14°C by 2000. By 2010 the temperature fell back to 13.2°C.

But then came the “massive” altering of data, which also altered the entire overall trend for the period. According to journalist Ederer, Ewert uncovered 10 different methods NASA used to alter the data. The 6 most often used methods were:

• Reducing the annual mean in the early phase.
• Reducing the high values in the first warming phase.
• Increasing individual values during the second warming phase.
• Suppression of the second cooling phase starting in 1995.
• Shortening the early decades of the datasets.
• With the long-term datasets, even the first century was shortened.

The methods were employed for stations such as Darwin, Australia and Palma de Mallorca, for example, where cooling trends were suddenly transformed into warming.


The full story is here.


It is particularly gratifying to see Prof Ewert’s confirmation of the massive tampering with Arctic temperatures, which I highlighted earlier in the year (and which NOAA/GHCN have STILL failed to give me any explanation for, despite many promises to do so).


It should also be noted that his analysis only runs to 2012, so does not pick up on the recent “Pausebuster” scandal now being investigated by the US Congress.





It should be pointed out that Prof Ewert is actually referring to the GHCN V2 raw data, used by GISS prior to 2012, compared against the GHCN V3 Adjusted data used since. Hence the same adjustments made by GHCN are also used by NOAA and HADCRUT.

  1. November 20, 2015 7:37 pm


    • Ian Magness permalink
      November 20, 2015 7:49 pm

      Not really David. Not since it became clear many years ago that the Hockey Stick was deliberately fabricated and had no scientific or statistical validity whatsoever, has any misbehaviour by the Green Blob and their supposed scientific and political collaborators come as any surprise.

  2. Ian George permalink
    November 20, 2015 8:00 pm

    And here’s my favourite ‘adjusted’ site – De Bilt in Holland, which I believe is the only site in Holland that GISS use.
    Version 2 (‘raw’ data)
    Version 4 (adjusted data)

    How can they get away with this?

    • Billy Liar permalink
      November 21, 2015 2:07 am

      Those are big numbers! (the ‘adjustments’)

      Quite astonishing.

    • CheshireRed permalink
      November 21, 2015 11:38 am

      Nice links Ian. Quite astounding data manipulation.

      Both our host here and Tony Heller at deserve massive credit for their efforts. Meanwhile our ‘leading’ science establishments should hang their heads in shame.

  3. November 20, 2015 8:01 pm

    I can’t wait for Hazzbin to report this on the BBC. I can just see the BBC headline. Oh wait, sorry, I must have been dreaming.

  4. Gary H permalink
    November 20, 2015 8:07 pm

    One assumes that Chairman Larmar Smith’s committee is on board.

  5. Don Keiller permalink
    November 20, 2015 8:23 pm

    Just how do we get this into the MSM?
    Are there any honest, investigative journalists left?

    • Ben Vorlich permalink
      November 20, 2015 8:38 pm

      Don’t hold your breathe, Steve Goddard has been publishing exposes of this for at least 2 years.

      • A C Osborn permalink
        November 20, 2015 11:04 pm

        Yes it confirms what he has been saying.
        But since 2012 it has got even worse.

    • November 21, 2015 6:27 am

      I’m sure Delingpole and Booker are on the case, but it needs somebody honest in the BBC.

      • Ben Vorlich permalink
        November 21, 2015 8:06 am

        Andrew Neil or Micheal Portillo are the best best bets.
        I have sent the story and a link to the BBC at suggesting that one of their four experts (Shukman, Harrabin, McGrath or Amos) check the story out. I have no expectation of anything happening.

      • Ben Vorlich permalink
        November 21, 2015 8:12 am

        I have just sent the link to the Daily Politics on their contact us web page and got a 405 error.

  6. November 21, 2015 12:20 am

    Paul, how about a link to your earlier post about adjustments?

  7. November 21, 2015 3:11 am

    Caution is needed here. This may be a result of comparing unadjusted GHCN v2 data with adjusted GHCN v3 data, both of which can be downloaded from the GISS website, but which are not validly comparable. The proper comparison for the unadjusted GHCN v2 data is with the unadjusted GHCN v3 data, which is not available from the GISS website but can be downloaded at

    “Between the years 2010 and 2012 the data measured since 1881 were altered” suggests that it is the “raw” data which is being discussed. GHCN v3 adjusted data is used as “raw” data by Gistemp, but it is not “the data measured” as it has already been adjusted by NOAA.
    I have emailed Prof Ewert to alert him to this.

      • November 22, 2015 9:14 pm

        v1 to v4 are better described as below to avoid any confusion (GHCN has had versions v1, v2 and v3, and now v4 beta)

        v1: GHCN-M v2 unadjusted
        v2: Gistemp homogenized with GHCN-M v2 unadjusted as input
        v3: GHCN-M v3 adjusted
        v4: Gistemp homogenized with GHCN-M v3 adjusted as input

        To open .tar.gz files, try 7-Zip

        Here are De Bilt plots from the two GHCN-M versions downloaded from, (dated 20151009 as I have those still expanded on my disk. These are large files with data for all stations, so you need to select and plot the data for the station you are interested in.

        GHCN-M v3 unadjusted (the one you should compare with GHCN-M v2 unadjusted from GISS, your v1 above):

        GHCN-M v3 adjusted (corresponding to GHCN-M v3 adjusted from GISS, your v3 above):

        And for completeness, Gistemp homogenized, corresponding to your v4 above, based on data to September 2015:

      • November 22, 2015 10:48 pm

        Just to clarify Peter.

        Both the GISS homogenised versions are to compensate for UHI

      • Ian George permalink
        November 23, 2015 12:47 am

        Thanks for that Peter.
        According to the GISS data I have used above.
        V1 Data set is ‘after combining sources at same location’.
        V2 data set is ‘after cleaning/homogeneity adjustment.’
        V3 data set is ‘after removing suspicious records’.
        V4 data set is ‘after GISS homogeneity adjustment’.
        I suppose my point is that there seems to be a cooling of the past and an increase in the present over the 4 versions. As they say, if you torture the figures for long enough, they’ll ‘fess up to anything

  8. AndyG55 permalink
    November 21, 2015 3:49 am

    How about this for data FABRICATION !!

  9. Pethefin permalink
    November 21, 2015 6:43 am

    All those interested in this theme should follow the postings of Tony Heller at this site:

    Tony has been digging in to this for a long time and some of his findings would be devastating for real scientist but climate science seems to a “post modern science” based on computer simulations and alternative realities.

  10. Bloke down the pub permalink
    November 21, 2015 11:18 am

    Yet another example of research by a retired professor. It’s funny how so many scientists feel unable to publish work that doesn’t support cagw until their livelihoods no longer rely on it.

  11. CheshireRed permalink
    November 21, 2015 11:30 am

    That this scandal was deliberately started and has been allowed to perpetuate speaks volumes of the dreadful standard of senior politicians at the top table of global politics. There is no excuse – they know of these adjustments but choose to ignore them. It’s malfeasance in public office.

    Please also note how despite this professor making the most outrageously damaging allegations against their professional integrity NASA/GISS will not take legal action against him as to do so would invite cross examination in an open court, which is the very last thing these criminals want.

    NB: Amber Rudd and George Osborne: If you both demonstrate the courage to tackle this blatantly fraudulent issue head on you will be the political beneficiaries. This is what’s called a career open-goal.

    • Bloke down the pub permalink
      November 21, 2015 11:42 am

      This is what’s called a career open-goal.

      Unfortunately, as these adjustments of doubtful integrity have been done with the knowledge of, and possibly at the behest of the president of the USA, any UK politician who exposed them as lies would find life extremely difficult. It’s the sort of situation that may take a generation to pass before people can see for themselves what should be patently obvious.

      • A C Osborn permalink
        November 21, 2015 5:14 pm

        Give the info to Putin, I am sure that he would just love to embarress Obama.

  12. November 21, 2015 11:58 am

    It is an interesting comment on the approach of the meja that a German prof saying what Paul has been explaining for years is published as believable. The Beeb wont believe him tho’.

    Perhaps Paul should change his name to Paulus von Heimvood?

  13. November 21, 2015 3:12 pm

    The “sham scientists” have taken over and managed to totally overcome the “real scientists” who insist on the scientific method. They are called “unemployed.”

  14. November 24, 2015 5:24 pm

    Regarding Paul’s clarification:

    “Just to clarify Peter. Both the GISS homogenised versions are to compensate for UHI”

    Both GISS homogenised versions are intended to compensate for urban effects (UHI but also cooling, which can occur, although whether as often as GISS output suggests must be open to question). GISS does this by replacing the urban trend by an averaged trend of ‘nearby’ rural stations:

    “The adjustment to the urban record is a two‐segment
    broken line that approximates the difference between the
    averaged regional rural record and the urban record.”

    This implies an ability to identify urban and rural stations. To do this using night time luminance from satellite images requires knowledge of the actual station locations. The claim in Hansen et al 2010 that “Station location in the meteorological data records is provided with a resolution of 0.01 degrees of latitude and longitude, corresponding to a distance of about 1 km” is only true in the sense that these locations are provided as values to one hundredth of a degree. It is false if read as implying that station locations are known to within one hundredth of a degree of latitude and longitude, and was known to be false in this respect by at least one of the co-authors on the date the paper was accepted for publication. The excuse is that “Unfortunately, we [GISS] don’t have the manpower to check out all entries of that file”. As excuses go, this is rather a week one. Had it come from an undergraduate student doing project work under my supervision he or she would have been advised to sample the stations in the file to assess the frequency of errors which would have led to wrong classification before proceeding. As nearly half the stations in the inventory are WMO stations it should not take a spreadsheet-literate intern more than an afternoon to compare the locations known to GISS with those known to the WMO, and rank these stations by the distance between the two locations. I have found errors in the WMO inventory, which were promptly corrected when reported, but these seem far fewer than in the GHCN-M inventory which GISS uses, where a few errors I reported more than five years ago were corrected then, although others were left, including physically impossible locations.
    Another complication is that some of the locations in the GHCN-M inventory may be correct even if different from the current WMO location, if these stations are not currently reporting in the GHCN-M data but the GHCN-M location was still correct at the time of the last data reported. Possibly up to 20% of locations for non-North American stations (North American station locations seem more accurate) could be sufficiently inaccurate to switch the urban/rural classification. I’ll try to dig out more accurate figures and include them in the post referred to below.

    This comment is already becoming to long. A blog posting showing the adjustment contributions of both GHCN and Gistemp for the same stations would seem useful (and I have not even touched on the volatility of GHCN adjustments. I will write such a post now, and post it as (this will link to a placeholder post for now. Expect actual content Thursday)

Comments are closed.

%d bloggers like this: