Howard Richman's  Instablog

Howard Richman
Send Message
Dr. Howard Richman (mailto:howard@idealtaxes.com) is one of three generations of a family of economists. Howard co-authors the Trade and Taxes blog (http://www.idealtaxes.com/) and co-authored the 2008 book, Trading Away Our Future, published by Ideal Taxes Association... More
My company:
Ideal Taxes Association
My blog:
Richmans' Trade and Taxes Blog
My book:
Trading Away Our Future:
  • Prediction Failures of AGW Models 1 comment
    Nov 21, 2009 4:32 PM

    Scientific theories should be judged on their ability to make predictions, so I am always on the lookout for papers that test theories against new data.

    This week, I found an interesting website about climate research. It's the Science and Public Policy Institute (scienceandpublicpolicy.org). So I looked for new developments and found a collection of papers from November 16, 2009: 450 Peer Reviewed Papers Supporting Skepticism of AGW-Caused Global Warming by Anthony Watts.

    So I looked through the list of papers, hoping to find a paper that tested the predictions of climate change models and found a July 2009 meta-paper that sounded intriguing: Climate Projections: Past Performance No Guarantee of Future Skill? by C. Reifen and R. Toumi published in Geophysical Research Letters.

    Turns out that the 17 climate models used by the Intergovernmental Panel on Climate Change (NASDAQ:IPCC) fit the data very well for whatever period they were designed to explain, but do no better than chance when applied to a different period during the 20th Century. They found:

    In our analysis there is no evidence of future prediction skill delivered by past performance-based model selection. There seems to be little persistence in relative model skill, as illustrated by the percentage turnover in Figure 3. We speculate that the cause of this behavior is the non-stationarity of climate feedback strengths. Models that respond accurately in one period are likely to have the correct feedback strength at that time. However, the feedback strength and forcing is not stationary, favoring no particular model or groups of models consistently. For example, one could imagine that in certain time periods the sea-ice albedo feedback is more important favoring those models that simulate sea-ice well. In another period, El Nino may be the dominant mode, favoring those models that capture tropical climate better. On average all models have a significant signal to contribute.
    The authors are not global warming skeptics. Their conclusion is that all of the models need to be taken as a group in order to have predictive power. But there is a much more obvious explanation that jumps out of the page: All of these models have many many variables (i.e., "feedback strengths") which allow them to fit whatever time period they were designed to fit, but they don't work when applied to a different time period.

    In other words, the feedback strengths in these models are all free parameters. If you have enough free parameters you can make almost any model fit anything. These models don't work.

    Furthermore, if the group of models is applied to a different time period as a whole, they don't work as a whole. Witness the article just posted this week at Spiegel Online: Climatologists Baffled by Global Warming Time-Out, which points out that global warming appears to have stopped during the 21st Century.

    As best as I can make out, the situation in climate prediction science is the following:

    • 1. The Anthropogenic Global Warming (AGW) theory was once the best theory available for explaining climate change.
    • 2. The new theory of cosmoclimatology, developed since 1996, is now the best theory available for explaining climate change. It involves three factors: cosmic rays, solar activity, and magnetic fields, without requiring innumerable free parameters that must be adjusted. So far it has made excellent predictions whether the period be months, years, or millennia.
    The IPCC has rejected cosmoclimatology, much to their shame. They rely on a paper which supposedly rejects solar-activity as a powerful predictor of climate. Here's a criticism of that paper from the website (weatheraction.com) of British Astrophysicist Piers Corbin, a cosmoclimatologist who actually gets paid by insurance companies to make his excellent predictions of North Sea storms:
    The attacks on what the Global Warmers deem as 'solar theory' are the product of disgraceful dishonesty which marks the integrity of the scientific establishment at its lowest level since the Papal Inquisition.

    The main periodical solar activity effect - the largest observed periodicity present in world temperature data - is the 22 year cycle (driven by sun-earth magnetic connectivity). Hence for about half the time, the 11 year cycle of solar activity of particles, sunspots and radiation will move with temperature and half the time move against it. This is well known to solar and climate scientists. All the pseudo-scientists have done is essentially choose time spans where the two move in opposite directions and ignore demonstrated correlations on longer time spans....

    The world's leaders are set to meet in Copenhagen in December to discuss a possible treaty to reduce world emissions of CO2, the gas crucial to photosynthesis which is thought by the AGW theorists to cause global warming. Before they do anything they should listen to a talk by cosmoclimatologist Henrik Svensmark of the Danish National Space Center at the Technical University of Denmark. He could tell them about the factors that make correct predictions about climate change.

    Disclosure: No Positions

Back To Howard Richman's Instablog HomePage »

Instablogs are blogs which are instantly set up and networked within the Seeking Alpha community. Instablog posts are not selected, edited or screened by Seeking Alpha editors, in contrast to contributors' articles.

Comments (1)
Track new comments
  • Craig Pirrong
    , contributor
    Comments (13) | Send Message
     
    Nice article, Howard. It seems, based on your description, that climate models are ill-posed problems. This makes them of dubious use, at best, for out of sample forecasting.

     

    von Neumann once said (I quote from memory) "Give me four parameters, I can fit an elephant. Give me five, and it will swing its trunk." Climate models have a plethora of parameters which allow you to "fit" just about anything. Fitting, though, is not the proper criteria by which to evaluate these models. Out of forecast skill is, and the metapaper you cite casts serious doubt on that.
    9 Dec 2009, 10:56 AM Reply Like
Full index of posts »
Latest Followers

StockTalks

  • Looks like Seeking Alpha will no longer republish my blog entries. You can read them at: http://tinyurl.com/cj7vtf
    Apr 6, 2009
More »

Latest Comments


Instablogs are Seeking Alpha's free blogging platform customized for finance, with instant set up and exposure to millions of readers interested in the financial markets. Publish your own instablog in minutes.