• a) They employ a "base temperature" of 255K (-18C or 0F). The base temperature is the mean planetary temperature that would arise if all greenhouse gases were removed from the atmosphere, but the surface properties remained the same. This temperature is calculated ignoring the fact that clouds reflect both incoming and outgoing radiation. On correcting for this error, a base temperature is found to be 267K (-6C or 21F). It is important that the base temperature is correct because it is the starting point for all calculations of warming caused by greenhouse gases.

• b) They assume that, in the post-industrial period, there is no underlying natural temperature change. Thus, the increase in temperature over the last 100 to 150 years is caused solely by man-made releases of greenhouse gases. Study of ice-core data shows that there has been no period of 100 years or over without natural climate change. They also show that the direction of these geologically short-term changes are unrelated to carbon dioxide concentrations. During warming periods, the average rise expected is about 0.7C (1.3F).

• c) They ignore (or at least do not highlight) the fact that water vapour is a far more potent greenhouse gas than carbon dioxide. For example, to achieve an atmospheric absorptivity of 20%, would require about 30 times as many molecules of carbon dioxide as of water vapour.

1) The role of Water Vapour in Global Warming

This paper introduces a steady-state model to estimate the impact of atmospheric composition on mean global temperatures. As a steady-state model, it calculates the mean global temperature that would result from any atmospheric composition if the composition were held constant for long enough for temperatures to reach a steady value. The advantage of a steady-state model is that it decouples the impact of atmospheric composition from the natural variations in temperature that occur without changing atmospheric composition. The model is general and, in principle, applicable to any planet with an atmosphere containing greenhouse gases.

The title of the paper has been revised to reflect that it concentrates primarily on the balance between warming caused by carbon dioxide and that caused by water vapour. It omits consideration of amplification effects such as methane release and ice retreat because these are equally applicable to primary warming triggered by any greenhouse gases. The paper divides global warming into two parts (a) Total Global warming - that is the warming caused by all greenhouse gases combined, which amounts to about 22C (39F), and (b) Incremental Global Warming – that is the temperature change resulting from incremental changes in atmospheric composition such as post-industrial carbon dioxide releases, which amounts to around 1C (2F). The conclusions are:

• i) Water vapour releases are probably responsible for as much incremental warming as are carbon dioxide releases, which has relevance to a possible hydrogen economy

• ii) Water vapour is responsible for over 90% of total global warming and, in this sense, is the most important greenhouse gas.

The paper introduces a theoretical correlation between global temperature and carbon dioxide concentration that fits the observed measurements well. The model also accurately predicts the time lag between the atmospheric temperature and composition.

3) The natural variability of climate

The paper analyses the variability of global temperatures recorded in the Vostok ice-core. The variability found in the 400,000 year record matches that of the last 10,000 years when allowance is made for a smoothing of apparent variability with time. The smoothing has two causes; the first results from slow diffusion of trapped gases through the minute cracks in the ice and the second results from much longer periods of time between samples in the very old sections of ice. The diffusion results in a time difference between the temperature and carbon dioxide records at any given depth. The longer periods mix ice laid down over a long period of time so that the resulting measured temperature and carbon dioxide values are averaged over many more years. This averaging smoothes the record for both temperature and carbon dioxide concentration. Once correction is made for the smoothing, it is found that the variability of climate has remained the same throughout the 400,000 year period of the ice-core.

4) The current correlation between temperature and carbon dioxide concentration

It is noted that throughout the period in which both carbon dioxide and weather records have been available, both temperature and carbon dioxide concentration have steadily increased. The two data series are thus serially correlated and, in itself, such a correlation has no statistical significance. (For example, the divorce rate and price of fish are similarly serially correlated). In the circumstances, the only statistically meaningful conclusion results from model comparison. The paper compares the measured correlation with the three correlations derived from three alternative models. The models are:

• Simple steady-state model. The rate of temperature rise is found to be much larger than predicted by the steady-state model. The discrepancy cannot be the result of a time lag because such a lag would result in measured temperatures lagging behind the steady-state values that would arise given infinite time to settle

• The Vostok ice-core correlation. The observed temperature change is found to be less than one third of that which would be anticipated if we were to assume that carbon dioxide drives temperature change and the same model applies now as has applied for the last 400,000 years

• The natural variability of climate to be anticipated where the variability is not driven by changes in atmospheric carbon dioxide concentration. This model agrees well with the currently observed temperature rise.