This is the Cato Daily Podcast for Thursday, May 22nd, 2008.

I'm Caleb Brown.

Is there something systematically wrong with those computer models that predict global warming.

Cato Institute Senior Fellow in Environmental Studies, Pat Michael's comments.

Well, there are dozens of computer models of climate for the future.

They all tend to predict warming, in fact, constant rates of warming,

but how much warming is forecast is really dependent upon what model you choose.

Now what's more interesting is... forecast is really dependent upon what model you choose.

Now what's more interesting is what recently appeared in Nature magazine by Kenley

Side at all.

These were some scientists from Leipzig, is that they were looking at

patterns of temperature in the ocean and decided that we're not going to see

any more warming till at least the middle of the next decade. That would mean that

we essentially will have gone two decades with no surface warming.

Now, there's not one computer model in the world that was used by the United Nations Intergovernmental Panel on Climate Change,

the so-called consensus of scientists of which I'm a member that has in it

two decades where there is no net warming.

Now if these models are what serves as the basis for policy, that means we really

don't have very much science behind our policies anymore. And the ramifications of this are very

large. It not only tells us the shortcomings of the computer models, but it tells us that there

may be a systematic error. Because the way global warming works, carbon dioxide initially

warms up the air, and then the surface of the ocean warms up a little bit which allows more water to evaporate and water is a much more powerful warming greenhouse gas than carbon dioxide. If you delay the warming of the ocean

...