Leftists choose their words very carefully, in an effort to make themselves appear intelligent and virtuous.
They claim their surface measurements global average temperature "predicts" what would have been measured at all un-sampled locations, if they had been using a perfect measuring system.
That's propaganda.
In fact, their surface temperature data are handicapped by having up to half the planet (divided into "grids") with no data!
They wild guess the temperatures in the grids with no data, to compile a global average surface temperature.
The result is that the surface temperature measurements are a large pile of farm animal digestive waste products!
And here's why:
(1) Surface data are mysteriously different than weather balloon (radiosonde) and weather satellite data, which happen to be similar to each other.
Surface data are an outlier from the two other measurement methodologies.
Being an outlier should make any real scientist suspicious of the surface data (but not the leftist climate modelers, of course).
(2) There are only two basic types of surface data
(other than land measurements and sea measurements):
(a) Real measurements, and
(b) Fake measurements.
The fake measurements are the wild guess data placed in the empty grids.
Infilling a grid when having no actual data for that grid, is a guess.
I call it wild guessing because the people doing the infilling are government bureaucrats who can't be trusted.
Those bureaucrats can't be trusted because they were originally hired to play computer games, and make scary global warming predictions.
They WANT to see the rapid warming they predicted with their computer games (i.e; confirmation bias).
So it is no surprise their half infilled, half real measurements surface data claim significantly more warming than radiosonde and weather satellite temperature data, and the gap is growing.
Government bureaucrat computer games (falsely called "climate models") predicted a fast rate of warming.
The bureaucrats want their computer game predictions/projections/simulations to look accurate!
And that's exactly what they are doing with repeated data "adjustments" that create more warming out of thin air, and that means their infilling also can not be trusted.
Leftists say "un-sampled" locations (grids)
are "predicted" (wild guessed).
Those are deceptive propaganda words
to put perfume on a pig.
Infilled data are not real predictions.
A real prediction can eventually be compared with reality.
Infilling can never be verified.
Infilling can never be falsified.
No one can ever know if the infilling process is unbiased science, or a biased process to deceptively increase global warming.
That's why I call infilling "wild guesses" -- there is no feedback to find out if the guesses are even 'in the ballpark' of reality.
Infilling is only subject to the "leftist verification test" -- the usual, confident, arrogant public statement that basically says: "It's Right Because We Are Brilliant Scientists, and We Say So !"
Leftists use surface temperature data for two reasons, and those reasons have nothing to do with real science:
(1) The Climate Computer Gamers (government bureaucrat climate modelers) own the surface temperature actuals, so can easily and frequently "adjust" them to show more warming, and
(2) Surface data, after many "adjustments", show more warming than satellite data. If the opposite was true, smarmy leftists would only use satellite temperature data, which happens to have far less infilling than surface data.
NASA and NOAA both claim their annual average global surface temperature estimate has a margin of error of +/- 0.1 degree C., or less, -- that's nonsense, completely unrelated to measuring instrument margins of error, and also ignores the large amount of grid infilling / wild guessing!