NOAA Quadrupling Radiosonde Temperatures By Data Tampering

Screen Shot 2016-04-30 at 8.18.13 AM

NOAA 850-300 mb radiosonde data shows about 0.2C net troposphere warming from 1958 to 2010, but after NOAA gets done tampering the net warming is about 0.8C.

Screen Shot 2016-04-30 at 8.00.29 AM

Adjusted : RATPAC-A-annual-levels.txt     Original Angell data : global.dat

They accomplish this by a spectacular hockey stick of data tampering.
Screen Shot 2016-04-30 at 8.15.38 AM

This entry was posted in Uncategorized. Bookmark the permalink.

9 Responses to NOAA Quadrupling Radiosonde Temperatures By Data Tampering

  1. AZ1971 says:

    Any reasons why adjustments were necessary according to NOAA?

    • Billy Liar says:



    • AndyG55 says:

      TOBs? ;-)

    • Rud Istvan says:

      They claim to correct for inhomogeneities. Specifically different instruments at different places and times, using a first differences method.
      This makes no sense. There are a finite number of radiosonde instrument makers in the world, and a finite number of instrument changes. For each, there is a careful lab based calibration curve. For example the cold dry bias of each successive Vaisala hygrometer instrument version is established in the lab. The proper procedure is to take the reported data by altitude, apply the appropriate instrument calibration curve based on the temp measured for the altitude, and record that as the correct ‘raw’ data. There would be no inhomogeneities to adjust. There are only 85 stations in this series.
      The proper calibration fix is childs play to program. The calibration curves are available from the manufacturers. The stations will know when they switched instruments. Its in their purchasing records. And the errors between instruments for temp are much smaller than for humidity in the first place.

  2. Steve Case says:

    Off Topic:

    This is interesting over at
    There’s a box just to the left of their chart
    that says:


    GMSL Rates
    CU: 3.3 ± 0.4 mm/yr
    AVISO: 3.4 ± 0.6 mm/yr
    CSIRO: 3.3 ± 0.4 mm/yr
    NASA GSFC: 3.4 ± 0.4 mm/yr
    NOAA: 3.2 ± 0.4 mm/yr (w/ GIA)


    They note that NOAA includes GIA but for CU they don’t and we know that they do include a GIA adjustment. They are getting away with a lie of omission and we all seem to be sitting still for it.

    Of course we know that they all include the GIA adjustment, I wonder how long it will be before all mention of GIA will be scrubbed from their website.

    • Rud Istvan says:

      SC, good catch. GIA model estimates are mainly relevant for the closure problem. Note also that it seems reasonably geostationary tide gauge estimates have apparently been ‘disappeared’. For sure, CSIRO spliced sat altimetry onto tide gauges to claim acceleration, pulling a very visible “Mike’s Nature trick”. Opening graphic in essay PseudoPrecision in my most recent ebook.

  3. John of Cloverdale, WA, Australia. says:

    Of course the adjustments make it warmer. Why am I not suprised.

  4. AndyG55 says:


    What columns from the RATSAC data should I be using ?

    I’m averaging the 850 to 300 columns. Not doing any smoothing.

    I’m only getting a trend in RATSAC that is about TWICE that of the Angell data.

  5. pinroot says:

    I like the acronym that the adjusted data starts out with:

    It couldn’t be more appropriate.

Leave a Reply to Steve Case Cancel reply

Your email address will not be published. Required fields are marked *