The Real Science blog by "Steven Goddard" claims that NASA have altered their data in the USA and other countries to show global warming.

Right after the year 2000, NASA and NOAA dramatically altered US climate history, making the past much colder and the present much warmer. The animation [here] shows how NASA cooled 1934 and warmed 1998, to make 1998 the hottest year in US history instead of 1934. This alteration turned a long term cooling trend since 1930 into a warming trend.

If so, why would the adjusted data be considered superior to raw data?

  • You would, of course, need to modify more than just two entries over that range to reverse a trend. They appear to have done a decent job of lining up the scales in that graph, and it does appear that dates before 1970 were lowered slightly. About 1970 now seems to have pivoted slightly up. The original ("unaltered") graph does not have the post-2000 data (for what I hope are obvious reasons) - which appears to contain a large jump in the baseline. Note that even the original post-1930 temperatures were higher than the pre-1930 ones. Jun 24, 2014 at 10:19
  • It's more accurate to say that unadjusted climate data is unusable in analysis rather than that it does not show long-term global warming. The claim "unadjusted data does not show global warming" is just as correct as "unadjusted data does not show no global warming"
    – slebetman
    May 2, 2015 at 10:00

1 Answer 1


In a word, no. Weather stations are (as their names suggest) designed for collecting information for weather forecasting (i.e. short term variations in temperature, precipitation etc.), they were never intended for collecting information for climate research (long term statistical behaviour of the weather). As a result, the instruments used tend to have biases and discontinuities in the records, due to things like moving the station itself, or changing the time of day at which readings are taken, changes of instrument, that are not a problem for weather forecasting, but do pose a problem for climate research. To compensate for these known changes, homogenization steps are taken to correct for these problems. These are well documented and explained in the literature.

Trewin provides a nice overview of this research topic,

Blair Trewin, "Exposure, instrumentation, and observing practice effects on land temperature measurements", WIRES Climate Change, volume 1, issue 4, pages 490-506, Apr 21 2010, DOI: 10.1002/wcc.46.

The International Surface Temperature Initiative is a useful resource for those interested in investigating this topic in more detail.

Also of interest is the Berkeley Earth project, which was in initiative set up by initially skeptical scientists to investigate whether the global surface temperature datasets were in fact biased by these adjustments. Using their own independent method, they found that they got the same results as the other groups.

There is also the point that the satellite global temperature datasets give similar warming trends to the surface measurements, but the satellite measurements are not subject to the same biases. Note the UAH product is produced by noted climate skeptics John Christy and Roy Spencer.

surface (GISS and HadCRUT) and satellite (UAH, RSS) global temperature data

Essentially, it seems that "RealScience" has just cherry picked the station data to find ones with substantial calibration problems and not mentioned that there are good reasons for these adjustments.

  • 1
    Have you shown that this calibration is the cause of the (alleged) change in the cited graphs between 1999 and today?
    – Oddthinking
    Jun 23, 2014 at 17:04
  • GISS (NASA) provide both the code for their temperature time series (data.giss.nasa.gov/gistemp/sources) and the raw data is freely available, so if there were any other adjustment made, it would be easily detected, so I don't think NASA would be that stupid. There have also been independent implementations of their method (clearclimatecode.org/about). The methods of HadCRUT were also implemented independently by one of the panels investigating the email hack. To be sure I'd need to download the code and raw data and runs everything through and check, however that
    – user18604
    Jun 23, 2014 at 17:13
  • would be an egregious waste of my time, given that several groups have independently validated current homogenization methods (including BEST, who were initially in the skeptic camp on this question).
    – user18604
    Jun 23, 2014 at 17:15
  • @Oddthinking see also the comment threads from RealScience, where the reasons for the update were given (stevengoddard.wordpress.com/2014/06/23/…) and Goddard replied "Claiming a reason to alter data is light years away from demonstrating that what they are doing is legitimate." See also the comment a little further down from Steve Mosher pointing out the code is available and that it had been independently validated, Goddard gives evasive response...
    – user18604
    Jun 23, 2014 at 17:50
  • Steve Mosher says "This is old news recycled.", which sums it up nicely, this is a topic that has been endlessly recycled on climate blogs. See also stevengoddard.wordpress.com/2014/06/23/… and stevengoddard.wordpress.com/2014/06/23/… where Goddard replies to an answer he doesn't like with "Moron alert".
    – user18604
    Jun 23, 2014 at 17:51

You must log in to answer this question.