Uncertainty in climate science
Image - UncertaintyIPCC (note)
Different future IPCC scenarios for temperature, with the uncertainties in paler colours.
There have been a couple of interesting discussions of uncertainty in climate science over the last few days – see this article in the Guardian, and Richard Black’s BBC blog from last Monday.
One of the favourite key messages of climate sceptic lobbyists is that the science of climate change is uncertain or unproven. Climate models – the computer programmes that underline many projections of how the climate may change in the future – are a particular target for this line of argument. Nigel Lawson, chairman of the Global Warming Policy Foundation, even went so far as to say:
‘The only certain thing is the science is uncertain’
The difficulty is that making general warnings about ‘uncertainty’ leaves deliberately hazy what exactly is uncertain. Climate science comprises a vast body of knowledge, much of which is very solidly understood. Scientists have, for example, known that greenhouse gases warm the planet since the 1860s. The observation that the globe is warming is supported by numerous independent lines of evidence and is not reliant on outputs from climate models.
But inevitably as scientists push the barriers of knowledge of climate change science, there is major uncertainty in some areas.
The Guardian article flags responses to the web magazine Edge, which conducted a survey asking which scientific concept, if widely understood, would be most useful to improve our ‘cognitive toolkits’. Many responses, the Guardian reports, pointed out that the public often misunderstands the nature of scientific doubt. Neil Gershenfeld, director of the Massachusetts Institute of Technology’s Centre for Bits and Atoms, said:
Uncertainty is intrinsic to the process of finding out what you don’t know, not a weakness to avoid. Bugs are features – violations of expectations are opportunities to refine them. And decisions are made by evaluating what works better, not by invoking received wisdom.
Or as Andrew Dessler of Texas University recently put it more pithily, scientists “view uncertainty like a starving wolf views red meat.”
A good example of this was contained in Richard Black’s blog last week, where he discussed a recent paper on mountain glacier melt.
The outputs from the models used resulted in a three- or even four-fold difference between the minimum and maximum projected loss.
As Black notes:
No-one acknowledges the limitations of computer climate models more readily than modellers themselves, who will frequently bemoan the roughness of the resolution at which they have to work given the tools available.
But does this mean that we have to wait for complete certainty before acting? Black asks,
Are you willing to accept the risk that the European Alps could lose 90% of their ice by 2100, or New Zealand’s ranges 85%?
In such a complex system, our knowledge of many parts of the climate system is unlikely to ever be definitive. However, it’s clearly the case that even uncertain answers can help us to work out the scale of the risks we face from climate change, and how to manage them.