As world leaders gather in Japan to discuss what many now see as the biggest problem facing the planet, geophysicist Peter Sammonds sifts through the myths and realities of global warming

Wouldn't a bit of global warming be a good thing?

Global warming is considered different from local, more manageable environmental issues like deforestation, land degradation or pollution of the sea. It is seen as a global manifestation of our pollution of the planet, stoking fears of an unknown, invisible danger all around us. The solutions proposed range from dramatic changes in lifestyle to deindustrialisation and population control. Even then, many fear that any action we take might already be too little too late.

What are the facts? There is much anecdotal evidence that global warming is occurring - hurricanes in the Caribbean, the break-up of Antarctic ice shelves and warm summers in England. But what are the likely implications of warming, and is the science of climate change really now beyond dispute?

Global warming theory became scientific orthodoxy with the publication of the first report of the Intergovernmental Panel on Climate Change (IPCC) in 1990. In its 1992 report the IPCC predicted a best estimate of a global mean temperature rise of 0.5°C by 2010, and 2.5°C in the next 100 years. Assuming this to be the case, does it really matter?

The Earth's climate is forever changing. We currently live in what geographers call the Holocene, a warm period that has lasted for about 10 000 years. Prior to the Holocene was the Pleistocene, when the climate alternated between ice ages, lasting about 100 000 years, and short inter-glacials, lasting about 10-20 000 years. In fact most Earth scientists regard the Holocene as an inter-glacial, with the onset of a new ice age not too far distant.

The record of past temperatures - the palaeo-climate - is preserved in the ice sheets of Greenland and Antarctica. In the last few years much effort has gone into extracting continuous ice cores of 3 to 4 km depth which hold high resolution records of the palaeo-climate of the last 100-250 000 years. These tell us that compared to the more distant past, we have had remarkably stable temperatures during the Holocene. However, even in the Holocene temperature change is not unusual.

Change is the norm, and nature has no preferred state - but should we, from a human-centred perspective, have an ideal climate in mind? Reconstruction of palaeo-climates and palaeo-environments shows that the ice-age Earth was a truly dreadful place. As well as mean global temperatures more than 4°C colder than today, ice-age climates were more arid, there was no Indian monsoon, vegetation survived in isolated refuges and there was higher atmospheric circulation. The Amazon forest was almost wiped out.

By contrast, 6-7000 years ago was what geographers term the mid-Holocene thermal optimum, when global mean temperatures were some 2 to 3°C warmer than present. With warmer climates comes higher rainfall; in the Holocene optimum, global net precipitation was 9 per cent higher. The Sahara desert, which was considerably more extensive during the Pleistocene ice ages, effectively did not exist during this period. The Savanna which replaced it was hospitable to human life, and it was only with the general cooling after the Holocene optimum that the Sahara grew to its present extent. There were some areas of increased aridity, such as Turkey and the Rockies, but in general the warmer climate of the Holocene optimum meant a more fertile and productive Earth.

So why should we have anything to fear from global warming? Some areas of productive farmland would be lost, but they would be more than replaced by new areas for agriculture. A flourishing Sahara would surely be a good thing on a global scale. Just as 'scare' stories about Britain acquiring a Mediterranean climate make me say 'roll on global warming', so everything history has to tell us seems to point to the potential benefits of a warmer world.

However, even if the general outlook is bright, what about the local extremes of climate that the environmental disaster lobby predict will be caused by global warming? Will a warmer or a warming world be more unstable, even if it is overall more hospitable to human life?

Research into extreme geological events is certainly some of the most exciting work being undertaken presently: the idea that huge tsunamis would have battered the eastern seaboard of America triggered by sub-marine slumping in the Canary Islands; the possibility of widespread volcanic dust in the western USA and the implications for large-scale volcanism; and of course the realisation that a comet impact probably caused the extinction of the dinosaurs. Grappling with these ideas drives Earth scientists, and it sure beats working on improving seismic data processing techniques for the oil industry.

The notion that global warming will bring fast, extreme and damaging changes to parts of the world has recently been argued by the Chief Scientific Adviser to the British government, Sir Robert May FRS. One of the problems he raised is the increased likelihood of tropical hurricanes in a warmer world. However, the general case for more violent atmospheric circulation globally is not borne out by the historical evidence, again taken from deep ice cores.

The Chief Scientific Adviser also raised the possibility that the Gulf Stream, responsible for bringing warmth to north-west Europe, could be shut off. This is of course scary stuff, for while the world warms Britain would turn into a sub-Arctic wasteland. The argument is based on the idea that increased precipitation in the North Atlantic region, and increased fresh water run-off, would reduce the salinity of surface water. Water will therefore be less dense and would not sink so readily, therby reducing the circulation of warm water from equitorial regions. However the geologic records suggest that the opposite happened in the past - warmer climates meant stronger deep water formation (see M E Raymo, 'The initiation of nor-thern hemisphere glaciation', Annual Review of Earth Planetary Science, 1994). > So Chelsea could play Tromso in northern Norway with no more than a chill in the air.

Given that the evidence does not seem to support many of the dire warnings, why is global warming always seen as such a problem?

The fatalistic tone of the discussion is partly a product of the broader climate of the times, when people tend to be seen as the passive victims of change, unable to adapt, never mind thrive in or modify a changing natural world. This is most evident in the discussion of the Third World.

For instance: 85 per cent of Bangladesh's population depends on the land to survive; the coastal regions of Bangladesh are the most fertile agricultural areas; and those regions could be flooded if warming causes sea levels to rise appreciably. The argument that global warming will be a disaster for Bangladesh is easily made. Yet this argument rests upon the static view that, in say 100 years time, Bangladesh will not have moved beyond the impoverished economy it has today. The solution lies in addressing the root causes of the problems which global warming could bring to the surface - not only by applying the most advanced science and technology to protect agriculture, but by promoting the industrialisation of countries like Bangladesh.

Such development, however, is now seen as a problem, with global warming widely taken as a metaphor for the way we are supposedly messing up the planet. As the American writer Gregg Easterbrook has argued: 'the idea that clumsy tinkering with the environment has done something useful - that we benefit from pollution! - is almost too peculiar for polite discourse.' That phrase - pollution - is indicative of the negative way in which the issue is always discussed: in reality carbon dioxide is no more a pollutant than is fresh water or sunlight.

A futuristic thought to finish: what happens if dumping carbon dioxide into the atmosphere threatens a run-away greenhouse effect, a warming without limits? The obvious answer, according to Professor Andrew Palmer FRS in a recent article in Science and Public Affairs, is to dump it in the oceans instead. Palmer argues that large producers such as power stations should dump their unwanted by-products in carbon dioxide lakes formed in the deep ocean, held there by the huge pressure of the water above. The technology for this already exists. Such a measure would be a step towards the ultimate goal of human control of the climate.

Peter Sammonds is a Royal Society University Research Fellow at University College London

CO2 and hot air

Once it had been established in the mid-nineteenth century that the Earth had experienced periodic and extensive glaciations, covering much of northern Europe and North America, the search was on for the mechanism responsible. After several false starts the Milankovitch-Croll astronomical theory of ice ages finally triumphed in the 1970s (see J Imbrie and K P Imbrie, Ice Ages: Solving the Mystery, 1979). This astronomical theory says that ice ages start when the amount of sunlight on high latitudes in summer is low. Less summer sunlight means more unmelted ice, and expanding ice sheets. The intensity of sunlight can be calculated from astronomical obser-vations of the eccentricity of the Earth's orbit, the variation in axial tilt and the precession of the equinoxes. The Milankovitch theory predicts periodic ice ages - a prediction found to agree with temperature cyclicity recorded in sea-bottom cores and ice-sheet cores. These drastic climate changes are not driven by the concentration of carbon dioxide.

Carbon Dioxide does, of course, play a role in determining the temperature of the planet, as has been known since the late nineteenth century. Clearly the Earth does experience a greenhouse effect; surface temperatures would be 33°C colder without our atmosphere. The physical mechanism is straightforward (see R Buckley (ed), Climate Change, 1997, for a readable review). The Earth's surface temperature depends on the balance between incoming short-wave energy from the sun and outgoing long-wave energy emitted from the Earth's surface. Some gases in the atmosphere, the greenhouse gases such as carbon dioxide, allow short-wave solar radiation to pass through and warm the Earth's surface, but at the same time these gases trap some of the long-wave infra-red radiation re-emitted from the ground.

It is possible to map the historical relationship between concentrations of carbon dioxide and climate change, by studying ice sheets. As well as recording the past temperature, these also contain a high resolution record of past atmospheres, thanks to trapped pockets of air. For instance the Vostock ice core drilled in Antarctica shows a strong Milankovitch temperature cyclicity (see chart). However, the carbon dioxide content of the trapped air-bubbles also follows this cyclicity over a long time-scale.

In other words, for palaeo-climates, the carbon dioxide content of the atmosphere broadly responds to global temperature change. Note the chain of causality: historically, carbon dioxide concentrations have followed climate change, rather than forcing change. This is borne out by studying the big picture on the scale of tens of thousands of years. But it is also important to note that, over shorter intervals of a few thousand years, the temperature record and carbon dioxide record can be moving in opposite directions. This needs to be kept in mind when today's meteorologists are making their observations about global warming over decades, and at best hundreds of years.

All in all, scientists who reject the notion of anthropogenic induced global warming have plenty to get their teeth into. We should be sceptical of any assertion that extreme weather events experienced today are linked to rising carbon dioxide levels.