Science in Society Archive

With a Bang, Not a Whimper

The gigantic Antarctica iceberg floating loose has brought global warming back into the headlines. Scientists say sudden catastrophic changes are to be expected. Prof. Peter Saunders looks into the new report, and asks why the scientists are so complacent.

Hardly anyone doubts that the Earth is getting warmer and that the rate at which this is happening is increasing. What is more, the underlying cause is well understood. Over the past century we have been burning ever increasing amounts of fossil fuel and so releasing large amounts of carbon dioxide (CO2), too much to be absorbed by the plants and other carbon sinks. As a result, the CO2 concentration in the atmosphere is gradually increasing. Unfortunately, this makes the atmosphere less transparent to the light and heat that the Earth is radiating out into space, so less heat is being lost and the Earth is becoming warmer. It's the same principle that allows an unheated greenhouse to be warmer than the surrounding garden, which is why it is called the greenhouse effect.

There are other greenhouse gases, notably methane, but CO2 is the most important. The more fossil fuel we use, the more CO2 we release, and the more the Earth warms up. Since this is bound to have many harmful effects, such as a rise in sea level and desertification in many tropical regions, we ought to be substantially reducing our carbon emissions, by using less energy and by relying more on non-carbon sources. That much is agreed by almost all scientists who have studied the issue and by most of the governments of the world, with the notable exception of the USA.

It is, however, more complicated than that. The Earth is not an inert sphere covered by a thin layer of gas. There are many dynamical processes going on, involving the atmosphere, the oceans, the ice sheets, the land and the biosphere, and these have a significant effect on the climate. We don't yet understand how all this works, at least not well enough to model the climate accurately. But we do know quite a bit. We also know that a large, complex dynamical system can change in ways that are not easily predicted and are not simply proportional to the disturbances that led to them. In particular, a gradual change may become very rapid, apparently spontaneously.

If this were to happen to our climate, the consequences could be very serious indeed. So the US National Research Council established a Committee on Abrupt Climate Change to report on the current state of knowledge in the field and to suggest what further research should be undertaken [1].

The result makes sobering reading. Research over the past decade has established that the climate system is indeed dynamic and highly nonlinear, which means that abrupt changes are possible. What is more, many such changes have occurred in the past. The most dramatic one that we know of in detail was at the end of the "Younger Dryas" interval, which began about 12 800 years ago and lasted for about 1300 years. Much of the Earth was cold, dry and windy during that interval, although some southern regions were warm. The change in climate at the beginning and, even more so, at the end were abrupt; ice core data from central Greenland indicate that the temperature there increased by some 8o C in about a decade.

We don't have such good data for other regions, but we do know that the Younger Dryas affected most of the Earth and that sudden changes in temperature were by no means confined to one or two locations. We also know that there were about 24 other similar events during the 110 000 years covered by the Greenland ice cores. There have also been abrupt changes during the Holocene, i.e. the last 10 000 years. These include not only changes in temperature but also in precipitation and in the size and frequency of tropical storms and El Niño/La Niña events.

A recent abrupt change included in the report is the onset of the dust bowl in the American plains during the 1930s. Why exactly it began is not known, but the prevailing view is that such droughts are brought about and ended by random fluctuations, and maintained by positive feedback from the biosphere. Even a few weeks of very hot, dry weather can dry out the upper layers of the soil. The plants reduce the rate of transpiration, and this leads to even higher afternoon temperatures, and also to lower humidity, which reduces even further the amount of rainfall. If this continues for long enough, the plants die, so that even if it does rain, without a healthy root system in the ground, any water that does fall runs off rapidly.

This is uncomfortably close to the way that the daisies appear and disappear in Jim Lovelock's Daisyworld parable of how the planet might regulate itself based on feedback between organisms and the physical environment (see box). Fortunately for the United States, within a few years the rains came back and conditions returned to more or less normal. How long that might take, or indeed if it would happen at all, if a much larger region were involved, is hard to say. The example does, however, show how once things go wrong, whether by human actions or for reasons that have little to do with us, the way the system works can keep things wrong for a long period.

Those of us who live in northern Europe are especially interested in the Gulf Stream, which ensures our mean annual surface temperature is about 5-7 C warmer than we would otherwise expect. The Gulf Stream is part of a large thermohaline circulation (due to heat and brininess of the water) (THC) in the Atlantic Ocean. Warm water flows north and gives off its heat. This increases its density, so it sinks and flows back south in a deep current. The reason there is not a similar phenomenon in the Pacific is that the surface water is too fresh and therefore not dense enough to sink even if cooled. The strength of the Gulf Stream is thus sensitive to the proportion of fresh water in the North Atlantic, and one possible consequence of global warming could be to collapse the Atlantic THC and bring about a rapid cooling of northern Europe.

The report, which is almost 200 pages long, discusses both the theory and the observations in considerable detail, and is required reading for anyone seriously interested in climate change. The authors then go on to make a number of recommendations, largely about the direction of further research, which is what they had been asked to concentrate on. These are, as you would expect, about collecting data on past abrupt changes and improving modelling and statistical techniques to make them better adapted to situations in which abrupt changes may occur.

Their final recommendation is, however, profoundly disappointing, especially after having gone to such great lengths to demonstrate that the situation is almost certainly even more dangerous than most people think. They could not bring themselves to recommend more than what they call "no-regrets" strategies, by nudging (their word) research and policy in directions that will reduce vulnerability and increase adaptation at little or no cost.

So there is no call for a sharp reduction in the use of fossil fuels, still less for a reduction in energy consumption. Instead, they suggest that moving away from coal to natural gas would reduce greenhouse gas emissions and might also prove beneficial in the long term. They acknowledge that marginal regions face the possibility of droughts, so they suggest there should be markets in water. They also suggest the development of new instruments such as weather derivatives and catastrophe bonds. It is very hard to understand how they can be so complacent, and why their mentality can be so narrowly focussed on market 'solutions', given what is contained in the rest of their report.

The Earth's climate is a highly complex system and it is well known that the response of such a system is typically disproportional to the disturbance - a large perturbation in one variable may have little or no effect, while a small perturbation in a different one, or in the same one under slightly different conditions, may bring about a very large change. We also know that the Earth has shown that sort of behaviour in the past. Over the 3 billion years or so since life appeared on our planet, the sun's luminosity has increased by about 30% but the Earth's mean temperature (averaged over periods of, say, a million years) has hardly increased at all. On the other hand, during this time there have been significant fluctuations, some of them abrupt, and these have been the results of much smaller forces. Many of these forces were too small to leave a direct record - unlike the changes they triggered.

By pumping more and more greenhouse gases into the air we are pushing the system away from the state it has been in for many centuries. This is leading to global warming on a scale unprecedented in modern times. Recent calculations suggest that the average temperature on the Earth will rise by no less than 1.5 C and possibly as much as 6 C over the twenty first century. One consequence will be that the sea level will rise; the Pacific island of Tuvalu is already in danger and many seaports and much of Bangladesh could soon be submerged. Large areas that are currently suitable for agriculture could become deserts as the temperature rises. The Gulf Stream might disappear, leaving Northern Europe much colder than it is today.

Even if we were to make an immediate and drastic reduction in our use of fossils fuels today, it is probably too late to avoid many of the consequences of what we did in the last century. The question is how much worse will things get because of what we do - or fail to do - in this century. The report warns that the situation may be even more dangerous than we think, because a perturbation of the climate, whether due to natural causes or, as now, caused by human actions, can lead to a change which could be too great and too rapid for us to adapt to. It must be taken as a further warning that urgent action is required to reduce the emissions of greenhouse gases, and not, as the authors see it, largely as an argument for the creation of new financial instruments.

1. A draft version of the report of the Committee on Abrupt Climate Change can be found at

The Daisyworld parable

About twenty years ago, Jim Lovelock and Andy Watson devised a very simple model which they called Daisyworld to show how it can self regulate, as Lovelock had proposed earlier in the Gaia hypothesis. The Daisyworld also serves as a very good illustration of the sort of behaviour that can be seen in the climate of the Earth.

Daisyworld is an imaginary planet which, like the Earth, is in orbit around an ordinary star. Its only life forms are two species of daisies, one with black flowers and one with white. The daisies grow best at 22.5 C, less well if it is warmer or colder than that. They do not grow at all if the temperature is below 5 C or above 40 C, though the seeds can survive in the ground. They also grow more slowly as the planet becomes crowded. And, like all organisms, individual daisies have a finite lifetime.

To see how regulation comes about, imagine there are seeds for black daisies in the ground but that it is too cold for them grow. Now suppose the sun slowly becomes brighter, as we would expect a 'main sequence' star (like our own Sun) to do. The planet will gradually warm up, and when it eventually just reaches the threshold temperature, 5 C, a few daisies will grow. Because they are darker than the bare soil, they will absorb more energy and so it will be warmer where they are than on the rest of the planet. That will make the daisies grow faster, which will increase the area they cover, making the area warmer, and so on.

The positive feedback does not go on indefinitely, however, for two reasons. First, there is only a limited amount of space on the planet, and as this is used up, the rate of expansion slows. Second, if the daisies manage to raise the temperature to 22.5 C, any increase beyond that makes things less favourable for them, not better. The feedback eventually becomes negative and the situation stabilises. In fact, what happens in the model is an abrupt change in the temperature of the planet, from 5 C to about 20 C (see Fig. 1).

If the sun's luminosity continues to increase, the temperature rises, though slowly, and the area covered by daisies is reduced. If there are seeds for white daisies as well, they soon appear and start to replace the black ones. Since they absorb less energy, this has the effect of cooling the planet. The result is that over a long period the temperature of the planet remains very nearly constant; in fact it actually falls slightly. Eventually, however, when the sun is very much brighter and the black daisies have completely disappeared, the temperature starts to rise, first slowly and then more rapidly. Finally, the temperature reaches the point where the white daisies are only just able to cool the planet enough to ensure their own survival. Any further increase in luminosity triggers an abrupt disappearance of the daisies and a correspondingly abrupt rise in the temperature.

It is significant that after this has happened, if the solar luminosity should fall again, the daisies will not reappear and the planet will not become noticeably cooler. For values of the luminosity less than where the sudden jump occurred, the climate has two possible steady states, one at a low temperature state with daisies and the other at a high temperature without them. Systems with multiple steady states typically show hysteresis, this means that if for any reason they switch from one state to another it can be surprisingly difficult to get them to switch back.

Fig. 1: As the sun becomes brighter, Daisyworld gradually warms up, until there is an abrupt increase in temperature at about L=0.7. The temperature then remains roughly constant until there is a second abrupt increase at L=1.6. The upper portion of the loop demonstrates hysteresis: Once the daisies have died, if the sun's luminosity were to decrease they would not reappear and cool the planet until L had fallen to 1.2.

Another interesting feature of such systems is that as they approach a point at which a sudden jump will occur, they become less stable, i.e. they take longer to recover from a perturbation. So unusually large variations in climate can be warnings of an imminent sudden change even if they are not in the direction in which the change will occur.

Of course Daisyworld is far too simple to be a model of the real Earth. It does, however, capture the main features by which a planet such as ours can be self-regulating, that is, the temperature (more generally, the climate) can be very largely determined by processes on the planet itself. It also shows how the climate can exhibit abrupt and largely irreversible changes of state.

A relatively accessible account of the Daisyworld model and its implications can be found at

The figures to go with it are available at

A simpler, but less complete, account is at

Article first published 31/03/02

Got something to say about this page? Comment

Comment on this article

Comments may be published. All comments are moderated. Name and email details are required.

Email address:
Your comments:
Anti spam question:
How many legs on a duck?