A report released this week by the National Oceanic and Atmospheric Agency (NOAA) called it "the warmest year ever for the nation." Experts agree that 2012 was a hot year for the planet. But it’s that report -- and the agency itself -- that’s drawing the most heat today.
"2012 [wasn't] necessarily warmer than it was back in the 1930s ... NOAA has made so many adjustments to the data it's ridiculous," Roy Spencer, a climatologist at the University of Alabama in Huntsville, told FoxNews.com.
A brutal combination of a widespread drought and a mostly absent winter pushed the average annual U.S. temperatures up last year, to 55.32 degrees Fahrenheit according to the government. That's a full degree warmer than the old record set in 1998 -- and breaking such records by a full degree is unprecedented, scientists say.
But NOAA has adjusted the historical climate data many times, skeptics point out, most recently last October. The result, says popular climate blogger Steve Goddard: The U.S. now appears to have warmed slightly more than it did before the adjustment.
"The adjusted data is meaningless garbage. It bears no resemblance to the thermometer data it starts out as," Goddard told FoxNews.com. He's not the only one to question NOAA's efforts.
"Every time NOAA makes adjustments, they make recent years [relatively] warmer. I am very suspicious, especially for how warm they have made 2012," Spencer said.
The newly adjusted data set is known as "version 2.5," while the less adjusted data is called "version 2.0."
NOAA defended its adjustments to FoxNews.com.
Government climate scientist Peter Thorne, speaking in his personal capacity, said that there was consensus for the adjustments.
"These have been shown through at least three papers that have appeared in the past 12 months to be an improvement,” he said.
NOAA spokesman Scott Smullen agreed.
"These kinds of improvements get us even closer to the true climate signal, and help our nation even more accurately understand its climate history," he said.
One problem in weather monitoring occurs when there is a "break point" -- an instance where a thermometer is moved, or something producing heat is built near the thermometer, making temperature readings before and after the move no longer comparable.