This one is from The PAST is Not What It Used to Be where I discuss the Data Bias that I think has exaggerated Global Warming since 1880 by about 0.3ºC.
The blink graphic switches between two different sets of US Annual Mean Temperature data covering the same years published by NASA GISS (Goddard Institute for Space Studies). My annotations remain fixed to show the changes.
Notice how the 1999 version has the years before 1960 warmer, by about 0.1ºC as compared to the same data published this year! Note also how the 2011 version has the data for 1980 and afterwards warmer, by up to 0.2ºC as compared to the same years on the 1999 version! Why would data that has been in hand for decades have to be adjusted in this "see-saw" way, if not to exaggerate the amount of warming in recent years? (That is Dr. James Hansen, head of NASA GISS, juggling the Earth and his impact on our economy.)
The second graphic is from Do We Care if 2010 is the Warmist Year In History?. It reproduces a NASA GISS email from Dr. Makiko Sato to Hansen, released in an FOIA request, in which she recounts the seven versions of US Annual Mean Temperature published for 1934 and 1998.
I've annotated that email with a graph showing how 1934 was warmer than 1998 by more than 0.5ºC when the first version was published in 1999. When the exact same raw data was reanalyzed and republished in 2001, 1934 temperatures (which were by then old enough to collect Social Security :^) were somehow reduced by almost 0.2ºC and 1998 data got boosted by nearly 0.1ºC, to reduce 1934's lead.
The graphic shows how 1934 continued to get cooler and 1998 got warmer (though 1998 had a surprising setback in the 2006 version :^) until, in the last analysis, they were in a virtual tie. Note that all this data is from an official NASA GISS email.
Sadly, after the date of the email, further reanalysis was done to make 1934 about an eighth of a degree cooler than 1998. Sad loss for the old guy. This clearly demonstrates the NASA GISS desire to make the more recent year warmer to prove Global Warming. It was embarassing to have the older data beat the more recent in the race to be warmer. Kind of like losing a footrace to your grandfather! So they did a Photoshop on the photo finish image to correct their earlier error.
Question: If the final version (1934 cooler than 1998) is correct, then the initial seven versions must all be in error?