Not a 50% chance. A 50% conditional probability. | Brussels Blog

Not a 50% chance. A 50% conditional probability.

posted by on 19th Oct 2013

Below is a reply to Lord Deben, the chair of the Committee on Climate Change.

The Committee has focused on “an approximately 50% chance of a global average near surface temperature increase of 2ºC above pre-industrial levels” (see below).

Their reasonong is flawed.

To get their “50% chance”, the Committee had to exclude the effects of feedbacks missing from climate models. See Missing Climate Feedbacks in my previous post.

This means the Committee really considered a “50% conditional probability“. It is conditional because this probability depends on the assumption that the missing feedbacks would have no net effect on global warming. This 50% conditional probability is not the same as a 50% chance.

I think scientific judgements (even at that time) predicted the net effect of the “missing feedbacks” would increase global warming. This means that taking these missing feedbacks into account, the chance of exceeding a 2ºC limit would be greater than 50%.

I don’t remember Lord Turner (then chair of the Committee on Climate Change) making the conditionality clear when he spoke at the meeting of the All Party Parliamentary Group on Climate Change on 2nd July 2008.

The reply to Lord Deben

Lord Deben
The Committee on Climate Change

19th October 2013

Dear Lord Deben,

A “50% chance”?

Thank you for your reply. I am pleased to hear that the CCC is in the process of considering the effect of permafrost melt and other feedbacks on future global warming. May I draw your attention again to the work of Kevin Schafer. There is an accessible report of his concerns in Science Daily, Thawing of Permafrost Expected to Cause Significant Additional Global Warming, Not Yet Accounted for in Climate Predictions.

Nov. 27, 2012 — Permafrost covering almost a quarter of the northern hemisphere contains 1,700 gigatonnes of carbon, twice that currently in the atmosphere, and could significantly amplify global warming should thawing accelerate as expected, according to a new report released November 27 by the UN Environment Programme (UNEP).


Policy Implications of Warming Permafrost seeks to highlight the potential hazards of carbon dioxide and methane emissions from warming permafrost, which have not thus far been included in climate-prediction modelling. The science on the potential impacts of warming permafrost has only begun to enter the mainstream in the last few years, and as a truly “emerging issue” could not have been included in climate change modelling to date.

I read the title of this article as meaning the same as “There is a substantial probability that thawing of permafrost will cause significant additional global warming”. Do you read it differently?

If it is read as I suggest then it contradicts the Written evidence submitted by the Met Office to the Parliamentary Committee On Energy and Climate Change on 5 June 2013. An alternative explanation would be that that the statement itself is self contradictory where it says

The Committee on Climate Change has focused on an approximately 50% chance of a global average near surface temperature increase of 2ºC above pre-industrial levels, and a much higher probability, >90%, of limiting warming to less than 4ºC above pre-industrial levels. Based on this climate constraint a set of global emissions pathways were then derived.

And in a later paragraph

Even with the improvements in modelling many uncertainties in future projected warming and climate change still remain. These include uncertainty in the precise magnitude of the response of clouds to a warming climate, uncertainty in the response of vegetation to CO2 increases in combination with climate change, and uncertainty on how climate change will alter the frequency and severity of extreme weather events. Some processes, such as loss of carbon from melting permafrost, is not yet included even in the latest complex climate models.

The Committee on Climate Change makes headline claims about “an approximately 50% chance of a global average near surface temperature increase of 2ºC above pre-industrial levels”. This 50% chance is not a normal probability it is a conditional probability unworthy of the simple term “chance” used in normal English usage.

If the Committee, with Met Office assistance, had concluded that there were several feedback effects, positive and negative that cancelled each other out then it might have been possible to plausibly assert the “50% chance”.

In this case, I would have expected different responses from your committee to those I received following my initial questions in 2008. I think scientific judgements even at that time would have expected the net effect of the “missing feedbacks” to be significantly positive.

Is this your view?

Best wishes

Geoff Beacon

There are no comments.

Please Leave a Reply

TrackBack URL :