Monday, 17 January 2011

Being Certain about Climate Change Uncertainty

By Martin Sewell, Senior Research Associate, 4CMR

There are aspects of climate change about which we are almost certain (the physical chemistry), and areas in which uncertainty is rife (e.g. the effect of clouds, the ocean, the response of biological processes, climate change mitigation). My view is that we must explicitly engage with uncertainty, and the best way to do so is using a probability distribution, and the wider the distribution, the greater the uncertainty.

The 18th century philosopher (and economist) David Hume pointed out that ‘even after the observation of the frequent or constant conjunction of objects, we have no reason to draw any inference concerning any object beyond those of which we have had experience’. In other words, one can never generalize beyond one’s data without making subjective assumptions, so science always involves a degree of uncertainty.

What is the best way of communicating uncertainty? In March 1951, the CIA secretly warned US officials that a Soviet attack on Yugoslavia should be considered a ‘serious possibility’. When Sherman Kent, a CIA intelligence analyst, asked his colleagues what probability they attributed to the likelihood of an attack on Yugoslavia in 1951, he was shocked to hear such a wide range of responses that varied from 20% to 80%. In 1964 Kent wrote the seminal Words of Estimative Probability in which he attempted to quantify qualitative judgements and eliminate what he termed ‘weasel’ words. For example, he recommended that ‘probable’ meant 63–87%, and ‘almost certain’ 87–99%. Since then, the BBC and the IPCC have also given serious consideration to how to communicate uncertainty. My view is that we should use probability.