Wednesday 25 August 2010

Climate Science - why I am a sceptic

I have been asked on a number of occasions to defend my scepticism about anthropogenic global warming and the role that CO2 emissions play in determining climate change. This was started by a letter published in the Guardian Newspaper (16 December 2009) in which I drew some parallels between the state of climate science currently and those of finance in the 1970’s and 80’s. To amplify that letter here is a fuller explanation of my views.
I am a sceptic because I have always accepted Popper's refutational logic - it’s the job of good science to attempt to refute rather than to confirm. In the 'hard sciences' like chemistry and physics this is relatively straightforward, but subjects like cosmology, climate science and economics have very limited opportunity for direct controlled experiment hence the reliance on the analysis of past data and the development of models that attempt to link theories of causation with observational data. The problem, as Imre Lakatos a student of Popper made clear, is that theories are inevitably surrounded by ceteris paribus clauses and observations themselves are laden with theory. This central problem in science that observation is never pure and every test of a theory is always a joint test led Lakatos to the idea, very much in tune with Thomas Kuhn's work, that science is cyclical.
According to Lakatos scientists in a discipline 'buy into' certain core concepts which they accept as 'givens' and then the science progresses or degenerates as it gathers empirical content. Successful programmes of research progress by gathering empirical evidence supporting the core but eventually contrary evidence emerges. The response from those committed to the programme is to create ad-hoc modifications to the central theory. As a result a layer of theoretical adjustments and even denial of contrary empirical evidence occurs until a point is reached where the research programme is no longer tenable and it goes into decline as rival programmes begin to emerge. Unfortunately, one revelation from the recent email leak from the University of East Anglia is the lengths to which adherents to the ‘core programme’ in climate science have attempted to exclude those who disagree with them from the peer reviewed literature. This is not an uncommon defensive strategy by those in their position. In the mid 70’s for example, the editor of one of the principal financial journals stated that submissions to the journal that were contrary to the Efficient Markets Hypothesis would be very unlikely to be published. This is one of the weaknesses of the peer review system is that it is very hard to get fundamental criticism of the dominant research programme into the literature. Indeed, there is some evidence that where contrary evidence has been found authors have found it necessary to dream up more or less plausible explanations of their findings in terms acceptable to the mainstream.
Climate science, like monetarism in economics, has certain core concepts which all members of the club agree with. In the case of climate scientists, there is consensus that there is a positive feedback mechanism amplifying the impact of changes in CO2 on average planetary temperatures. Surprisingly few engaged in this topic would disagree that there has been some planetary warming over the last 100 years. How much is debated, but certainly enough to cause a recession of arctic ice and some rising of sea levels. None would disagree that altering the concentration of CO2 will cause some warming. It is also agreed that on its own the impact of CO2 would be very limited. But, any warming of the atmosphere should translate into a slight warming of the ocean with a consequential release of water vapour. The argument is about what happens next. Climate scientists propose that the increase in atmospheric water vapour (which in the volumes released is much more significant than the increase in CO2 level) will induce commensurately more warming, increasing ocean temperatures further and hence the warming in what could be, ultimately, a catastrophic global temperature shift. What they have developed then is a theory of causation which cannot be proven by evidence but is supported by modelling of the climate system.
Now the sceptics argue that this theory of causation is incomplete and the climate models so far developed are very poor at predicting climate changes. There is little dispute that the existing models are severely handicapped by the relatively poor understanding of cloud formation and precipitation. Sceptics argue that increasing water vapour raises precipitation and clouds causing a homeostatic process to occur which mitigates the warming effect. Clouds at certain levels are highly reflective and the increased cloud cover can bounce away a greater proportion of sunlight hence cooling the planet. So if this negative feedback effect does occur (and such homeostatic systems are the rule rather than the exception in natural systems) then effectively the planet engages in a process akin to 'sweating' to keep its temperature stable.
If this is the case, why has there been any warming at all? This is where the sceptics suggest that solar forcing, long term oscillation in the heat flows within the Pacifica Ocean and a myriad of other more minor effects begin to play a significant role. In this we can see another problem of resolving the debate. We have a classic clash between those who subscribe to Ockham’s reductionism and those who do not. The climate scientific community has proposed a simple and quite brutal process which they argue will lead to catastrophic changes in planetary temperature. The core contrary argument is that planetary temperature is moderated by a system of subtle interactive effects which when taken in isolation appear to be minor but when put together create a system which is far more adaptive and complex than just the sum of its parts.
So the real nub of the debate about global warming is not the empirical evidence of a warming planet, retreating ice caps and rising sea levels but rather a complex, dynamic theory of causation which climate scientists see in very simple terms and which sceptics argue is far from simple. In fact this debate goes to the heart of the problem with the conventional approach to science. Climate scientists (like many in other systems) stick rigidly to the application of Ockham’s Razor: simple theories are better than more complex ones and the scientific process is one where we ‘reduce’ complex behaviour to simpler and simpler levels of explanation. However, there is no law that says that if you have two competing theories the simpler one is necessarily truer than the more complex one.
It is true, as some claim that some sceptics have vested interests in the outcome of this debate. Inevitably the same is true of global scientists as well who benefit from the reputational rewards of their scientific club and considerable levels of funding to support their science. For my part I take the view that there is an urgent need to develop energy technologies which create a long term sustainable future. We need oil for certain energy applications (it is highly concentrated energy and is mobile). We need better static sources for power generation and renewables and nuclear are better suited to this purpose. But, if the theory of positive causation is wrong what we do not want to do is to choke off economic development or divert resources into carbon capture and such schemes if they are unnecessary. The resolution of the global warning debate is not a choice between conservation and gluttony but rather between alternative strategies for a sustainable future.

4 comments:

H.R. said...

Caught your comments over at WUWT and followed your link here.

Excellent post. I've bookmarked your site for more visits.

FijiDave said...

I agree with Judith Curry, a very interesting essay.

Now have you bookmarked.

Brian H said...

Have you heard of Nordhaus' DICE (and RICE) model? It looks at cost-benefit for the economy given various mitigation approaches. Here is a review by Dyson:
http://www.nybooks.com/articles/21494
and Nordhaus' own site:
http://nordhaus.econ.yale.edu/RICEmodels.htm

John F. Pittman said...

Good thinking and good writing. I note that there is also a contingent of frequentist and Bayesian approaches that agree with your overall assessment.