That's not correct. Just to name a few examples, we have quite good data of CO2 concentrations and CO2 infrared behaviour. But what's also strange in your argu

That's not correct. Just to name a few examples, we have quite good data of CO2 concentrations and CO2 infrared behaviour. But what's also strange in your argument, you say that if we know less about the effects, then we should freely change the climate system more. Here's a free online book by Spencer Weart of the history of climate science for anyone who is interested: http://www.aip.org/history/climate/simple.htm http://www.aip.org/history/climate/simple.htm And a blog by Steve Easterbrook about the current state of software in the more complex models: http://www.easterbrook.ca/steve/?p=2828 http://www.easterbrook.ca/steve/?p=2828 Probably many software-oriented people here might find the latter blog interesting: "We found that the climate models all have very low defect densities compared to well-known, similarly sized open-source projects."