A Forecasting Expert Testifies About Climate Change
11/04/11 09:34 Filed in: Forced Guest Post
The New York Times
Mon, 11 Apr 2011
To the Editor:
In ''The Truth, Still Inconvenient'' (column, April 4), Paul Krugman
begins with a ''joke'' about ''an economist, a lawyer and a professor of
marketing'' walking into a room, in this case to testify at a
Congressional hearing on climate science.
I am the marketing professor, and I was invited to testify because I am
a forecasting expert.
With Dr. Kesten C. Green and Dr. Willie Soon, I found that the global
warming alarm is based on improper forecasting procedures. We developed
a simple model that provides forecasts that are 12 times more accurate
than warming-alarm forecasts for 90 to 100 years ahead.
We identified 26 analogous situations, such as the alarm over mercury in
fish. Government actions were demanded in 25 situations and carried out
in 23. None of the alarming forecasts were correct, none of the
interventions were useful, and harm was caused in 20.
Mr. Krugman challenged 2 of the 26 analogies, ''acid rain and the ozone
hole,'' which he said ''have been contained precisely thanks to
environmental regulation.'' We are waiting for his evidence.
''What's the punch line?'' he asked. I recommended an end to government
financing for climate change research and to associated programs and
regulations. And that's no joke.
J. SCOTT ARMSTRONG
Philadelphia, April 6, 2011
The writer is a professor at the Wharton School, University of
Schweinsgruber says: Why do marketing climatologists use cultist, derogatory attributes such as
Dr. Romney-Hughes says: Readers are referred to the Green, Armstrong & Soon 'forecasting' paper here (PDF here).
Our favourite lines are "the belief that "things have changed" and the future cannot be judged by the past is common, but invalid" in conjunction with "there is some positive trend so the benchmark is disadvantaged for the period under consideration". We also enjoyed how the authors placed great importance on the different magnitudes of their 'errors' at different timescales but neglected to mention the signs.
Had we spotted this when it came out, we would have included the International Journal of Forecasting in our honourable mentions for excellence in reviewing.