I'm interested to read about this debate on a more empirically focused level.

(1) The last survey I read about scientists' (experts in that field) opinions was a Gallup poll from 1992. Even then, 60% agreed that the global temperatures had risen but only 19 % attributed the increase to human activities. The opinions may have changed since then. Does anyone have newer polls?

(2) Last time I read about the climate models these models were tested by looking at their ability to predict the past changes starting back from now. The models usually were off from the mark at least by a factor of 2. Does someone have newer information on that question?

(3) There is not the slightest disagreement that carbondioxide in the atmosphere is increasing. There is also no disagreement since at least Arrhenius more 100 years ago that this factor viewed in isolation predicts an increase of the temperature. His prediction of the actual temperature increase was far off, but his argumentation that CO2 viewed alone leads to a temperature increase stands unchallenged. The debate is whether other factors resulting from a temperature increase (like increased rainfall, clouds etc.) will work in the direction of higher or lower temperatures. If someone could point me to newer results I'd like to read them.

Basically the question here is whether the earth climate works in a negative (like for instance in the Gaia hypothesis) or in a positive feedback mode. In a negative feedback mode the effects leading to higher temperatures will on the other hand trigger effects leading to lowering the temperatures, keeping all in all a balance (notwithstanding extremely large natural catastrophes like a hit by a big celestial body). Does anyone have newer informations on a consensus on the mode of feedback?

One last question that is often in my mind in these discussions: Everybody knows that a weather prediction for let's say the weather in two weeks is extremely difficult to make and should not be trusted a lot. Why do some people who on other occasions don't trust long-term (14 days) weather predictions and have a deep mistrust in scientific modeling and predicting in other fields trust here so completely in extremely simplified and simplifying global climate models predicting the weather not in 14 days but in 100 years?

Having said that I still think that acting as if the danger was true is the better option in the light of conflicting theories about the future. But I must say when I look back at the 'ice age' predictions from the 1970s I'm a bit skeptical about the new kids on the block predictions, especially if they come in some cases from the same people who have dismally failed in 1970 to predict the state of the world in 2000.

Wolfgang