The Big Correction That Wasn't

The author of today’s post, Lisa Moore, Ph.D., is a scientist in the Climate and Air program.

This past week there’s been a lot of buzz about a small correction that NASA made to U.S. temperature data. Some have said, incorrectly, that the new data show that 1934 edges out 1998 as the warmest year on record, rather than 1998 as previously thought. Actually, 1934 edged out 1998 in the old U.S. record, too, although the difference was not statistically significant. My favorite quote on all this is from Tim Lambert, who said in his coverage of the issue that "1998 and 1934 went from being in a virtual tie, to being in a virtual tie".

Climate change deniers have been all over the NASA correction, saying it proves that global warming isn’t happening. Of course, that’s ridiculous. For one thing, U.S. temperature isn’t global temperature. Globally, the warmest year on record is 2005, and the second warmest is 1998. But what should we make of those high U.S. temperatures in the 1930s?

If you look at the corrected U.S. temperature record (and to a lesser extent, the corrected global record), you see that warming rose through the 1930s, then stalled or slightly reversed from 1940 to1970. The pattern was the same in the uncorrected data. The reason for the pattern is that industrialization started to take off during the 1930s, but without pollution controls. Aerosol pollution built up to create a "global dimming" effect that offset global warming until the Clean Air Act of 1970.

The NASA correction has been covered in numerous blog posts. RealClimate has its typically excellent coverage, as well as Grist and Ecotality. But I’ll let NASA scientist James Hansen have the final word. Check out his online letter about the corrections [PDF], especially the graphs comparing the corrected trends to the old records.