In my ongoing criticism of Michael Crichton’s book, I made the point that parameterisation links small and large scales, and I emphasised the IPCC point that clouds and humidity remain sources of significant uncertainty.

I’ve just started reading a gem of a paper by Graeme Stephens on the current state (circa end of 2003) of knowledge about cloud climate feedbacks. I suspect I will be making lots of notes from this, and excerpting lots of bits. However, wrt to Crichton, and the “no-one knows” issue, Figure One in Stephens’ paper is a comparison of future climate predictions from a number of models. In each of these models, the climate heats up with a 1% CO2 increase year on year. He shows the low cloudiness in two versions of models that lie at either end of the range of the global mean warming responses (ie just under 2C and just over 4C). He says:

The reduced warming predicted by one model is a consequence of increased low cloudiness in that model whereas the enhanced warming of the other model can be traced to decreased low cloudiness.

Why do I bring this up? It shows that while the cloudiness is a key part of the response, it isn’t changing the sign within our current understanding… we clearly have a lot to do to get it right … but it doesn’t change the result: increased carbon dioxide implies a warmer climate, how much warmer is a fair question though!