Replies

TChris, that's still essentially a weather prediction, not a climate prediction. Take a look at the link I posted in #27. Given a liberal margin-of-error, I could easily predict the average temperature for any state one month in advance; I just look at what the average temperature IS for that state and that month. (That, in essence, is climate. Climate = average weather.)

First off, I agree that weather and climate are different, but the article (which I did read) misses the point about global warming.

Yes, repeatable trends, like the difference in temperatures between winter and summer, are predictable at least as an average. But the global warming advocates are 'predicting' a change to the previously repeatable situation.

And yes, one can 'predict' based on an assessment of the trends, but that does not make the predictions accurate.

Here are two issues that need to be addressed:

1) What is the uncertainty associated with the prediction? How does that uncertainty compare to the effect being predicted? If our scientifically justifiable accuracy is no better than the effect being predicted (+/- 5 degrees of uncertainty on a prediction of +3 degrees) then we hardly have a basis for embarking on policies that guarantee near-term economic disruption.

2) What is the demonstrated validity of the models used to make the prediction? Even if we have analyzed our uncertainty and feel there is a real trend, we still may not be making accurate predictions. There is a simple test for this, and all current global-warming predictions fail. Apply the model to the conditions of 1900, and predict the conditions of 2000, using the sampled proxy evidence and assumptions for the future that the models employ. Not a single one of them correctly predicts the year 2000. An example of this is that a few years ago the same sorts of models (less sophisticated in a computer complexity sense, but still based on the same underlying assumptions) predicted an ice age by now.

Bottom line: Using a demonstrably repetitive cycle, like the winter-summer temperature variation, to justify an open-loop extrapolation of a change to that very cycle is hardly convincing. It does show that there is a difference between climate and weather, but it tells us little about whether climatologists have any basis for alarm over global warming.

There is a simple test for this, and all current global-warming predictions fail. Apply the model to the conditions of 1900, and predict the conditions of 2000, using the sampled proxy evidence and assumptions for the future that the models employ. Not a single one of them correctly predicts the year 2000.

I would be interested in your comments on this:

Junkscience is... junk (please ignore the provocative title)

The article links to this:

http://www.grida.no/climate/ipcc_tar/wg1/450.htm#fig127