How do we obtain a figure for Climate Sensitivity?

In a recent post I sketched out the importance of climate sensitivity.

I am going to re-iterate the argument here, because it is so important. I am simplifying the argument put by Stephan Rahmstorf here. It covers the whole argument, and is worth reading in its entirety, but time is short, so I will do a short summary as a service to my time-pressed readership.  If you have any doubts about what you find here, you will have to go back to Rahmstorf's page.

How do we know that doubling CO2 in the atmosphere will raise global climate by 3*C +/- 1.5*C?

We know from basic physics that in and of itself, a doubling of CO2 will produce a 1*C increase in the planet's temperature.

This increase will be influenced by feedbacks, primarily water vapour, clouds, and distribution of warm air in the short term.

We know from seasonal variations how these change with temperature, and climate models must be validated by being able to match observed seasonal variation.

Early models give a CS range of 2.5 - 4, and this has been narrowed down in improved models to 2.6-4.1. The most likely value is 3*C.

The second line of approach is through studying ice cores, which yield data for CO2, methane, dust and ice sheet extent. The result here is a CS of 3-4*C. This calculation is independent of climate modelling.

Models can be run with a wide variety of values to see which will best match observed data. A run using 20th century data gave a lower end for CS in the region of 1.5*C. The highest climate sensitivity - the most dangerous results for mankind - are much more uncertain than this lowest value, but it is the lowest value which is relevant to the present argument, since the sceptics have to claim a very low value. (Or they can accept a higher value, combined with a very strong cooling effect from aerosols).

CS can also be calculated from observed measurements.

Earth has heated by 0.8*C since the late 19th century. We know from observations the rate at which the oceans have been absorbing heat. We have also a good idea of the extent to which aerosol particles cool the planet. We can also factor in variations in solar input. Simple calculations of this interaction give a CS of 2.3*C.


There are a few papers purporting to support a lower value, and we need to look at these in detail.

[to be continued]