Someone (I’m not sure if they want to be named, so I won’t, but they’re welcome to comment if they wish) sent me a copy of a document by David Wasdell. It’s called Sensitivity and the Carbon Budget, and is an attempt to determine the Equilibrium System Sensitivity (ESS). Both the person who sent it to me, and myself, think they should try and get it published, but they seem to want some kind of feedback. So, this post is an attempt to get my highly-educated and well-informed commenters to provide some kind of pre-submission review.
I’ll provide some comments to get it started. The basic premise is that doubling CO2 produces a change in forcing of 3.7 Wm-2 and, by itself, would increase surface temperatures by 1oC. The temperature change between the last ice age and the pre-industrial Holocene is about 5oC. This would then (given 3.7 Wm-2 ~ 1oC) be associated with a change in radiative forcing of 19 Wm-2.
Over the same time period atmospheric CO2 concentrations increased from 180ppm to 280ppm which would produce a change in radiative forcing of
This is 8 times smaller than the net change in radiative forcing, therefore (according to David Wasdell) feedbacks must amplify this by a factor of 8. If we now consider a doubling of CO2, which produces a change in radiative forcing of 3.7Wm-2, feedbacks would increase this to 30Wm-2 and – hence – the ESS would be about 8oC.
So, that’s roughly what’s presented in the paper. I have two basic comments. One is that the driver for the change from the last ice age to the Holocene was not simply increased atmospheric CO2. It was probably variations in our orbit and variations in the inclination of our axis. It’s thought that this produced an enhanced forcing in the northern hemisphere (above 65N) which melted ice and reduced the albedo. This then warmed the planet and released CO2. The CO2 then produced further warming, increasing the melting which then reduced the albedo further and further increased CO2 concentrations. So, it was a combination of changes in our albedo and increases in atmospheric CO2 – and other greenhouse gas (GHG) – concentrations.
If you look at Figures 5 and 6 in this paper (Hansen & Sato) the external forcings are taken to be both GHGs and albedo changes. My understanding is that it is hard to separate these two effects and, since they’re both slow, by combining them (to get a total change of around 5-6 Wm-2) you get an estimate of the Equilibrium Climate Sensitivity (ECS) which is essentially a response to fast feedbacks only. Given a change in temperature of 4-5oC and a change in forcing of 5-6 Wm-2, you get a climate sensitivity of 0.75oC per Wm-2; or around 3oC per doubling.
Since it’s hard to separate the GHGs and the albedo changes, it’s quite hard to actually determine – robustly – an ESS. The one extreme would be what’s been done in David Wasdell’s work. Assume that one of them is the main external forcing and the other is a feedback, which gives an ESS of 8oC. Alternatively, assume that they combine to give a net change in external forcing, which then gives an ESS of 3oC. As I understand it, the best that we can say is that it is probably somewhere inbetween (although, I’m happy to be corrected here as this is somewhat outside my comfort zone).
The other comment I was going to make is that if you assume that the feedback response is some fraction of the external forcing, then the feedbacks act to amplify the radiative forcing by :
If the feedbacks amplify the external forcing by a factor of 8 (as suggested by David Wasdell), then . This seems very close to a runaway process (i.e., if the sum above does not converge). Given that we’ve had major changes to our climate in the past without a runaway occurring, might suggest that cannot be this close to 1.
Anyway, those are my thoughts based on a fairly quick reading of the document. My gut feeling is that it is too simplistic and doesn’t appreciate that CO2 was not the only external forcing that warmed us from the past ice age into the Holocene. I’m also sure there’s much more that could be said, including the difference between the ECS and ESS timescales and whether that is relevant or not. So, maybe some others who know more about this than I do, could have a look and provide some feedback through the comments.