Silly Singer

Enough mawkish maudlin sentiment. Time to jump on the wackos again. This time its UnSciAm, who are foolish enough to publish a letter from Singer without fact-checking it first (I don’t think you can blame Singer for getting his facts wrong, since thats the entire point). You can (and I do) blame SciAm for not checking first. Its not as if the data is obscure or hard to find or anything.

Singer claims “the models predict that temperature trends will increase with altitude by 200 to 300 percent” (and then goes on to say “the data from both weather balloons and satellites show the opposite”; what is the opposite of increasing by 200-300%? Decreasing by the same? Of course they don’t). Anyway, no, the models don’t preditct this at all. They predict about 30% – see for example Santer et al, 2005 – though its moneywalled, and I couldn’t find a public version. How about ripping off one of their figures:



  1. #1 cce

    SciAm replied in the print version.

    “Contrary to Singer’s assertions, the IPCC and CCSP are in very close agreement regarding tropospheric warming trends. The CCSP states: ‘For globally averaged temperatures, model predicted trends in tropospheric lapse rates [vertical gradients in atmospheric temperature] are consistent with observed results.’ The recent confirmation of consistency among global models and observations represents a major advance in climate change science. As Singer notes, the CCSP reports less consistency among the long-term observational and model-predicted trends in tropospheric lapse rates for tropical regions. He has, however, omitted the April 2006 report’s conclusions on this subject. The explanation favored by CCSP is that these discrepancies arise from ‘significant nonclimatic influences remaining within some or all of the observational data set’s rather than ‘errors common to all models.'”

    Also, completely unrelated, but interesting. The AMS hosted a presentation on polar ice last week that was recorded by C-SPAN.

  2. #2 Martin Lewitt

    Yes, ALL the models couldn’t be wrong, so the data must be. This faith in the models was punctured by Roesch’s diagnostic study of surface albedo, where not only were all the models wrong, they were all biased in the same direction.

    Perhaps CCSP should just note the discrepency, calculate the error it might imply for the models and await resolution.

  3. #3 Gavin

    Monckton makes the same mistake (among many others):

    All the IPCC’s computer models predict that at the tropical mid-troposphere, roughly
    coincident with the characteristic emission level, the change in temperature over the decades
    should be two or three times greater than the change in temperature at the surface. Averaged
    over the whole mid-troposphere, the rate of change at altitude should be – and, on our
    figures, must be – about half as much again as the change in surface temperature.

    I wonder if these things could possibly be connected? :)

  4. #4 Luboš Motl

    A more refined journal version of Singer’s argument was just published in (peer-reviewed) International Journal of Climatology,

    [I think “Tropical vertical temperature trends: A real discrepancy?” by Thorne et al (GRL VOL. 34, L16702, doi:10.1029/2007GL029875, 2007). Singers approach to error analysis is rather casual -W]

The site is currently under maintenance. New comments have been disabled during this time, please check back soon.