On February 14th 2006 a large group of journalists, academics and representatives of the BBC assembled at the windswept visitors' centre at the Thames Barrier, east London. The occasion was the press briefing for the launch of the most ambitious global climate modelling experiment ever attempted. It marked a unique collaboration between the BBC and climateprediction.net (or CPDN for short).
That briefing witnessed the start of what has been an exciting year for climate research, one that’s seen the UK public debate on climate change move on to a level that has proved to be world leading.
"the original concept was to involve ordinary people"
Back in 2001 I was invited to join the project by Myles Allen, the Oxford atmospheric scientist whose vision CPDN was. The original concept was to involve ordinary people in trying to predict what a future world will be like if carbon dioxide and other greenhouse gases continue to pour into our atmosphere.
It’s easy to prove that this sustained assault on the natural systems is largely due to human activity - simply by looking at the amounts of carbon-based fossil fuels being burnt. The underlying argument for engaging the public in this cutting edge research was that if we’re all part of the problem then let’s all be part of the hunt for a solution.
However, before anyone can know what the appropriate solutions are, we need to understand how the greenhouse gases we’re producing are likely to interact with the myriad of interrelated natural processes that make our planet work in the way that it does. Whether planning attempts to mitigate climate change through cutting back on greenhouse gas emissions, or simply learning to live with the consequences, we need to know the scale and patterns of future global warming.
The early years of the project saw the conversion of the model (the same one used by the Met Office for their climate research) so that it would run on domestic PCs instead of supercomputers. This was not a trivial exercise. Testing to make sure the model was stable, and the download process was secure, took a lot of work.
By September 2003 we went public with the first experiment that explored the likely outcomes from a simple doubling of carbon dioxide levels in the atmosphere.
Now, you might think that for such a simple scenario the outcome would be easy to predict, but it’s not. The reason is that even the most complex climate model, like the one CPDN uses, is a simplification of the real world. When reality is simplified the shortcuts used might produce believable results, but in time these will be proved wrong.
To explore all the ways in which the simplifications could produce the wrong results, the model had to be run many hundreds of thousands of times, each run with a slightly different set of simplifications. The model versions were tested by starting them with climate data recorded in the past, and seeing how close they came to predicting the present-day climate. Versions that performed well went on to explore the effect of doubling carbon dioxide on global warming.
To do this on a supercomputer would take decades and be prohibitively expensive. As I write this piece in January 2007 over 170,000 versions of the model have been run. The equivalent of around 20 million years worth of atmospheric processes have been simulated thanks to people all over the world donating time on their computers. Already we've learned a great deal about the strengths and weaknesses of the model that are being factored in to the ongoing research.
This first experiment only concentrated on the atmosphere, but in reality climate is influenced by many factors including the way that ocean currents move heat around our planet. Any realistic climate model must be able to simulate not only the way the air behaves but also the oceans.
When the BBC approached me to see if the CPDN experiment could be used to predict the most probable future climate of the UK, it became clear to the team that the time had come to launch the most ambitious public climate experiment ever attempted.
In this experiment a fully dynamic atmosphere was coupled with fully dynamic oceans. Not only were a range of realistic past and possible future greenhouse gas emissions included, but also the global cooling (dimming) effects of sulphur from volcanoes and human activities. It was this experiment that was launched in February 2006.
"over 100,000 people downloaded the model across 143 countries"
Within days of launching, this coupled experiment the project developed a life of its own through bbc.co.uk and the internet. Over 100,000 people downloaded the model across 143 countries within the first two weeks. By the end of 2006 over 4 million years worth of atmospheric processes had been simulated. What would have taken decades to do on a conventional supercomputer was achieved in just a few months. International interest was phenomenal and the team found themselves giving interviews to news channels around the world.
We’d originally only planned one hour-long BBC/OU programme on BBC FOUR, “Meltdown”, to encourage participation in the experiment. However, over the course of 2006 a host of other programmes related to the experiment were broadcast across BBC ONE, BBC TWO and BBC FOUR as part of the “Climate Chaos Season”. You can watch the shorter programmes, Reports From The Front Line Of Global Warming, online.
In May and June, two one-hour programmes presented by Sir David Attenborough raised public debate to new heights. At the end of October, the publication of the Stern Review saw the true costs of inaction quantified for the global community for the first time. Moreover one of the scientific sources used in the Stern Review was the first paper published in Nature to come out of our original CPDN experiment.
This coupled experiment did not go smoothly, however. Within weeks of the launch the first results started coming back. Something unexpected was happening – the predictions were suggesting that our world would experience future warming far greater than anyone had thought possible.
Potentially this was disastrous news, so the team began picking their way though the details of the information returned. Quickly it was discovered that there was a quirk in the sulphur emissions file supplied to the project. It was being misread by the model, so future human emissions of sulphur pollution were not being factored in.
As a result the cooling effect of sulphur pollution in the atmosphere (and which gives rise to acid rain) was not included and the models were “overheating”. Inadvertently the project had started the largest ever experiment on global dimming! Unfortunately, switching off sulphur pollution is not realistic, as most of it comes from the burning of the same fossil fuels that are the source of most of the greenhouse gases. The experiment had to be restarted with the correct sulphur inputs.
Since then the project has run without a hitch, and the results from thousands and thousands of people around the world are now being processed and analysed.
What do the results reveal? Well, broadly speaking in the UK by 2020 we can expect just over a 1°C mean annual temperature rise compared to that seen around the 1970s. In the next 15 years or so we can expect an equivalent change in the weather to what we've seen over the last 30 – increases in summer drought and winter flooding, for example.
By 2050 the rise is likely to be around 2.5°C, and by 2080 around 4°C. These figures are conservative, however, as they're based on carbon emissions in line with population growth. Exceptional industrial growth rates of China and India, for example, are not factored in.
These results are broadly in line with what other researchers have predicted, so you could ask, “What was the point?”. But the fact that we've been able to look at tens of thousands of different versions means we’ve got far greater confidence in the results that any previous research has ever been able to demonstrate – giving the critics far less room to hide in.
The first wave of the results can be seen on the BBC climate change results website.
The BBC and the Open University are not responsible for the content of external websites.