Sunday, February 20, 2011

Really Weird Final minus Raw

Those who never look at the data never get to see things like this.

At this CDIAC site you can download the monthly raw and monthly mean temperature records. This is accomplished by merely checking TMEAN and TMEANRAW in the box on that page. It will then give you the link to a CSV file that can be downloaded an the data put into an Excel spreadsheet.

I then opened the CSV file and subtracted the raw data from the final data. I think people should see what the climatologists think of as good corrections to the data.

I did this tonight for Albany Texas. Here is the picture of the total correction, by month for this temperature station.

Looks like crap.


  1. Not to nitpick, but with your considerable research skills did you think to take a look at _why_ adjustments to the data may have occurred? Clearly since you are aware of the source of the raw and treated data you are also aware of the METADATA that goes along with each station in the USHCN matrix. You can find all the station history in this nice file here:

    If you scroll down to this particular station number you can see if the station has moved in its history and if the instrumentation has changed. (check out this file from USHCN about what the various data positions are in the file itself:

    Now I don't claim to have all the information on this but when I compare your graph to the station history I see some changes in the location at points that show nice little spikes in the data (like in 1948 and 1974 the metadata indicate a small station move, and 1985 it looks like they installed a MMTS thermometer system, etc.)

    So rather than merely pointing at the graph and saying "looks like crap" you might wish to investigate what non-meteorological changes occurred at this station (or any station of the USHCN set) and then compare how these non-meteorological potential biases are dealt with.

    As I said I am relatively new to interpretting the station history metadata but just in case your readers want a bit more insight. This really isn't some fly-by-night organization playign fast and loose with the data. They've been keeping records and they are actually working very hard to show you how they processed the data and why.

    Just and fyi.

  2. Sigh, Nonymous you are an unthinking nuisance. And yes, you seem to like the nits.

    I know what corrections have been made. I have no objection to station move corrections. But I do think that it is quite fair to subtract final from input in order to see what the sum total of the corrections look like. And I do have an objection, a rather strenuous one to two things.

    1. if the corrections applied to the data are of a magnitude greater than a degree, it seems ridiculous to say that we have the precision sufficient to measure the amount of change that is claimed. The very fact that we must move the measured temperature by 2 degrees says it was 2 degrees inaccurate.

    2. Is it really sensible to change the observations taken 100 years ago and with confidence make it such that they read too cool (thus increasing the present calculated warming?

    3. I have a strenuous objection to the high frequency wiggle. That makes no sense whatsoever. Do you believe that the error in the thermometer record has a strong 12 month periodicity (before you misread things note that I said ERROR, not the temperature)? The error beats to a 12 month period

    Apparently you are incapable of actually understanding that if the error in the temperature varies that rapidly, then maybe we are not doing a good job of actually measuring the temperature.

    Finally nonymous, you seem never to understand any criticism. Is there ANYTHING that you can name that is wrong with the current system of temperature collection? From all your posts, one gets the opinion that you really think the climatologists shit don't stink. And I find it rather cowardly that you hide behind anonymity.

    As to the climatologists not being a fly by night organization, I think they are playing fast and loose with the data. With every edition of their global warming curves, the past gets cooler in absolute degrees. Note how the past tends to cool in the years between 1999 and 2011,r:20,s:0