SST bucket problem, the magnitude of the difference

Posted: October 24, 2011 by tchannon in climate, Ocean dynamics, Uncategorized

Figure 1

A recent post on Tallbloke’s showing data created by Roger Andrews about SST and sea level. At the time I experimentally extracted a time series from official HADSST2 gridded data here and compared it with the ICOADS data.

It seems a shame to lose the result so I am posting a plot. The gridded version is fixed up later by the publishers but the original can be seen here.

I think it is understandable there is concern over what exactly caused this, a very large change.

Tim Channon

  1. Roger Andrews says:


    It was never posted as a thread, but my (rather long) analysis of the SST bucket-intake bias issue is at
    [Moderator edit: Please accept my apologies for misrepresenting. Sort this out by email if you wish — Tim]

    Here’s a brief summary:

    1. Conventional wisdom has always held that an abrupt and permanent change from bucket to intake SST measurements around 1941 caused an artificial and permanent upward shift of around 0.5C in the raw (ICOADS) SST record, and a 0.5C stair-step adjustment is therefore applied to remove this artificial shift (see Figure 1).

    2. But there is and never has been any evidence to confirm that there was a permanent bucket-intake change around 1941, and in fact it’s now generally accepted that there wasn’t. There’s also no evidence for a permanent upward shift in the ICOADS SST record. What at first glance appears to be an upward shift is actually a short-duration spike caused by bad data during World War II.

    3. In other words, the 0.5C shift correction is invalid. Which means that HadSST2 and HadCRUT3 (which is about 70% based on HadSST2) are invalid too. Yet HadCRUT3 is accepted as the world’s “official” surface temperature time series.

    4. If the world’s official surface temperature time series is invalid, which it is, and nobody has noticed, which it seems they haven’t, then it’s time someone blew the whistle.

  2. Brian H says:


    There, done.

  3. Lord Beaverbrook says:

    Just thinking out loud here, the events around the end of the war in1945, Hiroshima and Nagasaki, that raised a lot debris into the atmosphere, equivalent of a small volcano perhaps, would they trigger a cooling dip in the data?

  4. Ray Tomes says:

    There were many more atmospheric nuclear tests, some much bigger, in later years.

  5. malagaview says:

    On the one hand they had to hide the decline

    On the other they had to make in incline

    The mainstream data series aren’t just flawed… they are F…ed

  6. toto says:

    Malaga: the SST are adjusted to be *lower* after WWII. Look at the graph again.

  7. Roger Andrews says:

    Toto: Yes, and it results in SST warming being underestimated by about 0.5C during the 20th century, not overestimated.

  8. malagaview says:

    @ Toto

    My general point is about the integrity of the mainstream temperature data sets.

    Data points have been randomly adjusted upwards and downwards based upon fairy dust science…
    Missing data have been invented… measuring locations have been cherry picked and truncated…
    Data is also mangled and tortured by harmonizing, extrapolating, gridding, weightings and numerous adjustments….

    When I look at this SST graph all I see is data torturing…. varying annual adjustments roughly in the range -0.1 C to +0.4 C… adjustments well beyond the accuracy of the underlying data readings… this is about the generation of false precision… at best the underlying data is probably accurate to plus or minus 1 C… in reality the error range is probably far worst… but how knows?

    In 1950 the adjustment was about 0.4 C… and it then declines to around 0.2 C by 1970… so here HADSST have been making the incline by adjusting downwards older data points… there is method in their madness… unfortunately this method is driven my the madness of AGW… and this madness has resulted in a vast array of adjustments in both directions… unpicking all the details of this can of worms is a very long process… this is all part of the AGW method: if you can’t convince them then confuse them… and just reading the postings and comments on the blogs you can see that this tactics really works.

    Like I said previously: On the one hand they had to hide the decline

    Like I said previously: On the other they had to making the incline… [typo corrected]

    Like I said previously: The mainstream data series aren’t just flawed… they are F…ed

  9. Green Sand says:

    Been looking at HadCRUT3 via CRUTEM3 and HadSST2.

    Interesting looking at the hemispheres, southern land to ocean ratio 1 to 4, recorded contribution of each each element is level 1850 through to today.

    Northern land to ocean ration 1 to 1.5, with the land making an increasing contribution over ocean? Eyeballs at 0.2C over the full series.

    More work to do…

  10. adolfogiurfa says:

    What is it temperature?, is it literature?, prose?, verse? or just a mad mathematician´s dream?