Thermometers Show That The US Has Been Cooling For Nearly A Century

US measured temperature data (blue below) shows that we have been cooling for nearly a century. But the temperatures reported by NCDC (red below) show the opposite, that the US is warming.

ScreenHunter_2185 Aug. 22 06.07

Measured : ushcn.tavg.latest.raw.tar.gz
Reported : ushcn.tavg.latest.FLs.52i.tar.gz 

The animation below shows the change from the thermometer data, to the fake data which NCDC releases to the public.


They accomplish this transition by an impressive 1.6 degrees of data tampering, which cools the past and warms the present. Their hockey stick of adjustments is shown below.

ScreenHunter_2183 Aug. 22 05.54

Note the sharp rise in data tampering shown in the graph above. This is largely due to just making up data. Since 1990, they have been losing station data exponentially, shown below. When they don’t have data for a particular month, they just make modeled temperatures up for that station, and infill the missing data with imaginary temperatures.

ScreenHunter_2182 Aug. 22 05.50

The amount of fake data is currently 40%, and if trends continue, nearly all of their data will be fake by the year 2020.

About Tony Heller

Just having fun
This entry was posted in Uncategorized. Bookmark the permalink.

30 Responses to Thermometers Show That The US Has Been Cooling For Nearly A Century

  1. omanuel says:

    Thank you, Steven, for having the courage to report what is:

    1. Global temperatures are falling
    2. Stars make & discard hydrogen
    3. Neutrons repel neutrons
    4. We live in a democracy

    Instead of what is reported:

    1. Global temperatures are rising
    2. Stars collect & consume hydrogen
    3. Neutrons attract neutrons

  2. tom0mason says:

    And that is why we pay taxes.

  3. Jason Calley says:

    Tony, I have a question concerning the infilled temperature that perhaps you know the answer to. My understanding is this: when a temperature reading is missing (for whatever reason) a new temperature is estimated and infilled. This new temperature is based on the historic pattern of what that location read in the past compared to the surrounding stations. My question is this; when the pattern of temperature relationships is looked at so that the infilled temperature can be estimated, do they use only the pattern of raw readings, or do they base new infilled temperatures on the pattern of ALL previous readings — including earlier infilled readings? Here is why it matters to me. Suppose that the algorithm for creating infilled data is based on all earlier readings (including earlier infilled readings for that same location.) Suppose that the initial algorithm creates an infilled reading that is slightly too high (and that certainly looks like something that you have already demonstrated to be true.) Once a slightly too high infilled temperature is input, the next time the algorithm attempts to create infill temperature for that location, the historical pattern of temperature relationships will already be biased slightly too high (because of the earlier too-high infilled temperature) — which means that the next infill be be even a bit higher than the first infill, and so on and so on and so on, with each infill biasing the patterns to make the adjustments increase logarithmically. If infills were based strictly on earlier raw readings (without earlier infills), you might conceivably (though unlikely) get a system that was relatively accurate, but if the infill algorithm is used as part of an iterative process, the adjustments suffer from positive feedback and become increasingly wrong and increasingly gigantic.

    • Jason Calley says:

      Oh! I almost forgot to say, “thanks”. THANKS!

    • I don’t know. As far as I know their source code is not available.
      I do know that almost all of the warming since 1990 is due to infilled data.

      • Jason Calley says:

        Tony, it is just stunning what they are doing — and they still claim to be scientists! If they are, in fact, using a system of infilling that incorporates a sort of positive feedback, that might explain not only the magnitude of their corrections, but the sheer volume of infilled data as well. As you have pointed out, they seem to be infilling not only data that is truly missing, but often seem to be replacing existing real data with infilled readings as well. Because their readings are being continually biased upward, they may be dropping existing readings because they appear to be outlying data. In other words, the algorithm initially produces a small (but continuously growing) bias to infilled readings. The erroneous infilled reading then produces a trend that makes real data look like outliers, so the new (and real) data is increasing dropped and then infilled with continually increasing numbers.

        Rinse and repeat.

        If they have a subroutine which routinely checks older historic data for newly created supposed outliers, then that would even explain the never-ending adjustment and re-adjustment of temperatures from decades ago.

        This is crazy! And without transparent publishing of their methods there is no way to verify their output. This is not even ethical, and it for darned sure is not science.

        • hifast says:

          Jason–great insight. Your hypothesis describing the unstable nature of NCDC adjustments is quite plausible.

        • mjc says:

          Also, by doing it that way, if/when they get ‘caught’, they just blame the algorithm and say that anything was an unintended consequence of a poorly written program. Then make an update and a host of arbitrary corrections to the existing mess…and carry on.

      • PersephoneK says:

        Hi Tony,

        I’m new to your website and finding it interesting. I consider myself a Global Warming agnostic for the most part (but a skeptic in general), since whether or not GW is or isn’t happening (in the variations of what that means to people), my response is largely the same (ie less government involvement usually = better). But I’m always interested in learning the truth about pretty much everything, and definitely want to understand more about any false data.

        That said, I’m curious, and sorry if this is a stupid question, but why is temperature data sometimes missing at stations in the first place?

      • rah says:

        Except they already said their algorithm is working just fine.

  4. Headline on Drudge about a possible 10-year pause in “warming.” I think they mean a 10-year continuation of the cooling trend.

  5. E Martin says:

    How about doing an entire paper on the data fiddling and infilling?

  6. Andy DC says:

    Billions for climate “research” with a predetermined conclusion. But we can’t afford a few dollars to keep thermometers operating, in order to determine what is actually happening with the climate in the real world.

  7. ntesdorf says:

    Those adjustment graphs also measure the Warmistas increasing desperation as the inexorable pause continues to drive the CAGW theory into the dust.

Leave a Reply

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