Lindzen on Climategate

Richard Lindzen has an excellent editorial showing that the Climate Science Isn't Settled.

I'm not a climate scientist, but from the outset the idea that the earth's atmosphere could provide significant positive feedback has always seemed very counter-intuitive to me. I just couldn't see how it wouldn't "blow up" with the various changes that the Earth has seen in its billions of years. So these paragraphs resonated with me:

The notion that the earth's climate is dominated by positive feedbacks is intuitively implausible, and the history of the earth's climate offers some guidance on this matter. About 2.5 billion years ago, the sun was 20%-30% less bright than now (compare this with the 2% perturbation that a doubling of CO2 would produce), and yet the evidence is that the oceans were unfrozen at the time, and that temperatures might not have been very different from today's. Carl Sagan in the 1970s referred to this as the "Early Faint Sun Paradox."

For more than 30 years there have been attempts to resolve the paradox with greenhouse gases. Some have suggested CO2—but the amount needed was thousands of times greater than present levels and incompatible with geological evidence. Methane also proved unlikely. It turns out that increased thin cirrus cloud coverage in the tropics readily resolves the paradox—but only if the clouds constitute a negative feedback. In present terms this means that they would diminish rather than enhance the impact of CO2.

There are quite a few papers in the literature that also point to the absence of positive feedbacks. The implied low sensitivity is entirely compatible with the small warming that has been observed. So how do models with high sensitivity manage to simulate the currently small response to a forcing that is almost as large as a doubling of CO2? Jeff Kiehl notes in a 2007 article from the National Center for Atmospheric Research, the models use another quantity that the IPCC lists as poorly known (namely aerosols) to arbitrarily cancel as much greenhouse warming as needed to match the data, with each model choosing a different degree of cancellation according to the sensitivity of that model.