Some implications of a new approach to the water vapour feedback

The water vapour feedback is the largest physical climate feedback. It also gives the second-largest contribution to the range of uncertainty in climate sensitivity in General Circulation Models (GCMs). Tracing these differences back to their physical causes in the hope of constraining climate sensitivity requires an appropriate quantification. Yet the Intergovernmental Panel on Climate Change judge that the conventional diagnosis of a water vapour feedback and a lapse rate feedback provides little insight into differences between GCMs' climate sensitivities. We show that the conventionally diagnosed water vapour feedback is in fact formally useless for investigating differences between GCMs' climate sensitivities-the anticorrelation between conventional water vapour feedback and lapse rate feedback makes the correlation between the water vapour feedback and their sum insignificant: i.e. statistically, knowing this feedback allows one to conclude nothing about the sum and thence about climate sensitivity. This follows primarily from how little relative humidity (RH) changes with climate change in GCMs. A more detailed physical analysis concludes that the overall mean decrease of RH on warming seen in GCMs is robustly physically based. This and other physical arguments then suggest that the stronger the positive water vapour feedback, the less sensitive climate can be expected to be. A diagnosis based on the partly-Simpsonian model of water vapour feedback avoids these problems. On the conventional view of the water vapour feedback, naive extrapolation of variations within present-day climate suggests that parts of our planet are close to locally reaching conditions that would allow a run-away water vapour greenhouse effect once they were extensive enough. Of course this has never occurred in geological history, and is not seen in Earth-like GCMs. Again, the partly-Simpsonian approach provides a simple qualitative explanation, by showing that the water vapour feedback can only cancel part of the basic Planck's Law negative feedback.