What is climate sensitivity, and why does it matter?

Climate sensitivity lies at the heart of the climate change debate.

If it is high, we have a big problem. If it is low, we do not have a problem.

Sceptics claim that climate sensitivity is low. Their whole case rests on this simple, testable proposition.

Climate sensitivity is both simple and complicated. It is a measure to the degree to which the earth will warm up for a given increase in the energy transfers in and out of the planetary system.

It is derived from these steps:

• We are confident of the value of changes in Earth's energy transfers ("energy budget").
• We also know that a given increase in the energy budget will set off "feedbacks" that amplify or reduce the global climate change that result from that increase.  It is here that uncertainties lie - but they are now marginal uncertainties. Climate science has enough knowledge to be sure that we have a problem.

All that remains is to refute the sceptics' claim that sensitivity is low.

This is how the scientific case is constructed:

• We know from textbook physics that CO2 retains heat.
• We know from observation that we have increased CO2 in the atmosphere by 40% since 1750.
• We know from basic physics that a doubling of CO2 by itself is sufficient to increase the global average temperature by 1.2*C when the system eventually settles down .

However, the oceanic feedbacks, through storing heat, delays this increase, while other feedbacks will enhance it.

These feedbacks include:

• Increase in water vapour in the atmosphere. This is a positive feedback, increasing the heating.
• Increase and changes in cloud cover. This has both positive and negative feedbacks.
• decrease in ice cover at the poles and on mountains. Positive feedback.
• changes in the ocean heat system (since most of the increased temperature is stored in the oceans). Negative feedback in the short term, delaying the global increase. 
• Man-made particulate aerosols, negative feedback of uncertain size.
• In the longer term, releases of methane from oceans and tundra. Seriously positive.

There are other, independent changes in the system that will affect the climate, primarily

• Volcanoes - negative feedback in short term, due to particles, positive in long term due to their CO2.
• Solar variations - which, hopefully, may reduce warming to a greater or lesser extent in the coming decades.

Climate models are designed to factor in all these variables. The models can be tested by putting in the data and running them, and comparing the results with observed temperature records.  The correspondence between model outputs and observed records is convincing. Not perfect, because nothing is perfect, but as they are refined with new knowledge, the relationship between models and observations gets ever tighter.

There are 3 lines of evidence that give climate sensitivity.

1. Models themselves. (They are reliable, btw) (as is demonstrated by these 3 graphs)
2. Observations of how recent global climate reacts in the short term to variations in things like solar output and volcanoes.
3. Studies of the ancient climate, matching known proxies of temperature against known levels of CO2 &c.

These lines of evidence show remarkable consistency. The image here (click on it to expand) summarises some of the results. None give a sensitivity below 1.5*C for a doubling of CO2, and the most likely figure is around 3*C. 

This means we have a serious problem

Against this impressive consistent result from a variety  of detailed, serious work, the climate sceptics have a handful of little papers that claim a low climate sensitivity. 

The whole debate, and indeed the whole future of human civilisation, hinges on this point.

In future blogs I will be examining the sceptics' arguments in more detail.

My blog posts here are highly simplified summaries. For more in-depth analyses I suggest you start here.

And there is an excellent in-depth review by Stephan Rahmstorf here.