[dubcomesaveme]

http://en.wikipedia.org/wiki/Milankovitch_cycles Global warming WILL occur, regardless of man.

The question is what is the value of the subcomponent of temperature variation caused by man. I would argue that it is nearly impossible to isolate this value. It would require models of every other natural cause of temperature variation, from the galactic position of the solar system to the water cycle, and confidence estimates of these models. Weather is a prototypically complex and chaotic system.

Rather than allocating resources based on an answer to a nearly impossible to answer question, we should tackle questions that are more solvable like "relation between cancer rates and environmental pollution" (aka things that could have the same policy effect i.e. curbing emmisions) and remember that acute, immediate events can have a far more deleterious effect on global temperature http://en.wikipedia.org/wiki/Nuclear_winter

[Tichy]

THe Milankovitch cycles don't seem very relevant here, don't they appear across hundred thousands of years (I just skimmed the Wikipedia article)? Also, if Milankovitch can predict cycles, wouldn't it contradict your assumption that we can't infer anything about climate at all?

[grandalf]

Superb point about resource allocation. That such an approach hasn't been taken is the root of my skepticism about the global warming apocalypse.