Tuesday, January 05, 2010

TRF's 20-day-old UAH temperature estimate exactly confirmed

On December 19th, I calculated the UAH December 2009 anomaly to be
0.28 °C,
plus minus a small error, as the sum of the 2009-2008 temperature differences averaged over the known days (Dec 1st - Dec 17th), plus the known final December 2008 anomaly (0.18 °C) - an algorithm that had previously been proven very successful.

You might have said that it was a bold prediction because the figure was 0.22 °C cooler than the 0.50 °C anomaly measured in November 2009.

Roy Spencer at WUWT just reported the final December 2009 UAH global temperature anomaly. Guess how much:
0.28 °C
That's right: TRF can get a better than 0.01 °C accuracy for the final temperature of a month in the middle of the same month. Read TRF to learn the truth weeks, and sometimes centuries, before others. :-)

This also strengthens the validity of the UAH rankings of 2009.


  1. Are you going to take a shot at January too?

  2. I think it may be a good idea to publish the estimates around the 20th of each month. But such numbers are of course much less important if the year is not just ending - like after December.

    The method is not breathtaking - maybe I should convince the UAH folks to do it themselves.