Michael Crichton argues that long-range weather prediction is impossible because of the chaotic† mathematics of weather systems. Most professional meteorologists would agree with him, but he is quite wrong when he says that the same is true of climate prediction.
Future climates are much more predictable than is future weather. We know that there is no way to predict if it will, or will not, rain on 2 November 2010 in Berlin. But we can with near certainty say that it will be colder in January in that city than it was in the previous July. Climate change is amenable to prediction, and this is why so many scientists are tolerably sure that a rise of carbon dioxide to 500 ppm, which is now almost inevitable, will be accompanied by profound climate change. Their confidence comes from knowledge of the past history of the many glacial and interglacial events of the past two million years. The record drawn from the analysis of Antarctic ice cores clearly shows a strong correlation between global temperature, carbon dioxide and methane abundance.
If any one of us wants to know the social conditions of Victorian England we go to Dickens, Trollope and the other fiction writers of that time. More than this, we speak about their writings as if they were the true historical account. This is why I take Michael Crichton’s opinions seriously, not because they are true, but because he is such a good storyteller; indeed, he is among my favourite authors of a good yarn (his mix of medieval history and quantum theory in his book Time Line, for example, made it the best of science fiction). The public is much more likely to be influenced by writers like Michael Crichton than they are by scientists. Fiction writers and film producers should ask themselves if they are sure that what they say is true before succumbing to the overriding imperative of the storyline; this is more important than ever before, now that we face deadly change.
The authoritative source of information and prediction on the climate of the coming century is the Intergovernmental Panel on Climate Change (IPCC). The IPCC issued its third assessment report in 2001, and the next is due in 2007. Sir John Houghton, formerly the director of the UK Meteorological Office, was one of the joint chairmen of the IPCC, and his book Global Warming, with its third edition published in 2004, provides the most up-to-date and readable account of our understanding of this fast-changing field of science. It is revealing to look back at the climate forecasts made in the late 1980s. Here, from Stephen Schneider’s 1989 book Global Warming, is a chart that illustrates the thoughts of climate scientists at a conference in 1987 (Figure 2). From the limited knowledge then available they did their best to predict the future climate and showed their guesses as dotted lines on the graph. The upper dotted line is of a scenario they thought almost science fictional in its extremity. The cross I have added to the chart shows where we are now: we are already close to the extreme temperature change that made those pioneers so anxious.
Figure 2. Climate forecasts made in 1988.
Future climate predictions are mostly based on mathematical models of the Earth that were first used to try to predict weather a day or so ahead. These weather models divided the whole atmosphere into small parcels and calculated separately and in combination the changes likely in each parcel. To do this fast and well needs a fairly powerful computer; interestingly, so advanced are home computers now that yours may be powerful enough for a modest model of this kind. When it comes to climate prediction it is not enough to consider just the physics of the atmosphere. We need to take into account the way that the ocean stores heat and carbon dioxide and the dynamics of its interchanges with the atmosphere; we also need to know the nature of the land surface – whether or not it is covered with snow makes a huge difference, for example. Forests we now know are not passive areas on a map with fixed climate properties but are live actors in the climate system; the same is true of the ocean surface and the organisms that live in it. The clouds and the dust particles suspended in the air also have a powerful effect on climate. To take account of all the vast number of variables, we need a large computer. Fortunately, we have at the Hadley Centre in Exeter, UK, and in Japan, at their science city, Tsukuba, the largest climate models in the world, and scientists from the two institutions collaborate. But in spite of the expertise and the powerful computing machinery, our forecasts are provisional and do not include all surprises. Some, like the threshold of irreversible change, we think exist, and we wonder if the circulation of warm and cold water in the North Atlantic may be poised for sudden change. But we are not much better at dealing with the unexpected than were Columbus and his sailors when they set sail westwards for the East Indies. Their model of a round Earth was good, but the real planet had a huge and unpredicted surprise, the existence of the North American continent. We would be wise to expect that instead of temperature and sea level rising smoothly as the years go by, as in the IPCC predictions, there will be sudden and wholly unpredicted discontinuities.*
There are several reasons to think that our journey into the future will not be plain sailing and that one or more thresholds or tipping points do exist. Jonathon Gregory and his colleagues at Reading University reported in 2004 that if global temperatures rise by more than 2.7°C the Greenland glacier will no longer be stable and it will continue melting until most of it has gone, even if the temperatures fall below the threshold temperature. Because temperature and carbon-dioxide abundance appear to be closely correlated, the threshold can be expressed in terms of either of these quantities. The Hadley Centre scientists Richard Betts and Peter Cox conclude that a rise in temperature globally of 4°C is enough to destabilize the tropical rain forests and cause them, like the Greenland ice, to melt away and be replaced by scrub or desert. Once this happens the Earth loses another cooling mechanism, and the rate of temperature rise accelerates again. In Chapter 1 I describe a simple model where the sensitive part of the Earth system is the ocean; as it warms, so the area of sea that can support the growth of algae grows smaller as it is driven ever closer to the poles, until algal growth ceases. The discontinuity comes because algae in the ocean both pump down carbon dioxide and produce clouds. (Algae floating in the ocean actively remove carbon dioxide from the air and use it for growth; we call the process ‘pumping down’ to distinguish it from the passive and reversible removal of carbon dioxide as it dissolves in rain or sea water.) The threshold for the failure of the algae is about 500 parts per million (ppm) of carbon dioxide, about the same as it is for Greenland’s unstoppable melting. At our present rates of growth we will reach 500 ppm in about forty years. The monitoring now in progress of all these crucial parts of the Earth system – Greenland, Antarctica, the Amazon forests and the Atlantic and Pacific oceans – shows a trend towards what on our timescale could be irreversible and deadly change. Indeed, the science editor of the Independent newspaper, Steve Connor, reported on 16 September 2005 the statements of several climatologists who had found the melting of Arctic ice to be so rapid that we may already have passed a tipping point.
Deadly it may be, but when we pass the threshold of climate change there may be nothing perceptible to mark this crucial step, nothing to warn that there is no returning. It is somewhat like the descriptions some physicists give of the imagined experience of an astronaut unlucky enough to fall into a massive black hole. The threshold of no return from a black hole is called the event horizon; once this distance from the centre of the hole is passed gravity is so strong not even light can escape. The remarkable thing is that the astronaut passing through would be unaware; there is no rite of passage for those passing thresholds or event horizons.
For several years now I have had on the wall above my desk that amazing graph of the temperature of the northern hemisphere from the year 1000 to the year 2000. It was produced by the American scientist Michael Mann from a mass of data from tree rings, ice cores and coral. Part of the version in the 2001 IPCC report is reproduced below. It is called in America, mostly by sceptics, the ‘hockey stick’ graph. This is because it looks like a hockey stick lying flat with its striking end pointing upwards. I keep it in view to reinforce my arguments with sceptics of global heating and also as a reminder of how severe it will be. The graph shows the natural fluctuations of temperature, and for the first 800 years of the past millennium there is a slight but perceptible downward trend, which, if projected, points to an ice age in about 10,000 years. Then, at the start of the industrial period in about 1850, it slowly begins to rise, and with ever-increasing acceleration
Figure 3. The ‘hockey stick’ graph.
it climbs to reach temperatures nearly 1°C above the long-term average. A single degree rise in temperature may seem trivial, but remember we are looking at an average for half the world, the northern hemisphere. The difference between the long-term average of the graph and the ice age, 12,000 years ago, is just over 3°C. The IPCC 2001 report suggests that the line of the hockey stick graph might rise a further 5°C during this century. This is about twice as much as the temperature change from the ice age to pre-industrial times.
Every 25,000 years or so, the position and inclination of the Earth to the sun changes so that there is a small increase in the total flux of warmth the Earth receives. On every third of these successive pulses of extra heat Gaia has reached its lowest temperature and lowest carbon-dioxide abundance; this is a sensitive state where the extra heat is more than can be managed and regulation fails. Gaia then enters an unstable state called an interglacial, much like a fever in one of us. This is the state of the Earth now.
We either forget or never knew how different the climate was in the last ice age. Most of the United Kingdom, and north-western Europe including Scandinavia, was buried beneath 3,000 metres of ice, a glacier as thick as that on Greenland now. North America was similarly glaciated as far south as St Louis, latitude 35°N. Despite all this ice it was probably a healthier world than now and more vegetation grew, both on land and in the sea. We think this because the abundance of carbon dioxide in the air was then below 200 parts per million. It takes a lot of life to pump it down that low.
The sea level was 120 metres lower than now, and land equal in area to the continent of Africa which is now below water was then above it. Much of this extra land was in South East Asia, which may explain why Australia was reached by humans during the ice age: the distance was short enough to be made on rafts or simple boats. Imagine there was a civilization 12,000 years ago with cities on the coast of that extended southern Asian continent. Who among them would have believed an early climate forecaster who claimed that soon they would be 120 metres beneath an ocean?
The changes likely in the world to come will, in their different ways, be as great as or greater than this. True, the sea cannot rise more than another eighty metres, the amount of extra water which would be released if the ice of Greenland and Antarctica melted. But the worldwide torrid conditions would reduce the productivity of the remaining land and sea, and the loss of vegetation would slow the rate of removal of carbon dioxide and so sustain the hotter age for 100,000 years or more. The greatest observable changes so far are in the Arctic, as was predicted in the first IPCC report in 1990. Below are satellite views of the Arctic basin in 1987, 2003 and an estimate view for some time between 2030 and 2050.
Figure 4. The progressive summertime decline in the area of floating ice.
The floating ice of the Arctic covers an area equal to that of the United States and serves as the home of polar bears and other animals; it is also the destination of the brave explorers who travel on foot to the North Pole. But, much more than this, it serves us all as a white reflector of the summer sunlight that falls upon it and helps to keep the world cool. When that ice melts, as soon it may, you will be able to reach the North Pole in a sailing boat, but we will have lost the air-conditioning capacity of the Arctic ice; the dark sea that replaces it will absorb the sun’s heat and, as it warms, accelerate the melting of the Greenland ice.
While Gaia may suffer from the unfreezing of the Arctic basin and Greenland, these areas may become the future centres of an appropriately diminished civilization, and already shipping companies are beginning to prospect new polar routes. The Northwest passage that for so long has been barred by ice will soon open; the tundra wastelands of Siberia and northern Canada that remain above sea level will be rich with vegetation, and the enlarged Arctic Ocean, filled with algae, may become the fishing grounds of the future.
Another likely change often discussed by climatologists concerns the path of the great ocean conveyor belt that moves the waters of the world’s oceans. The distinguished American Earth scientist Wally Broecker first warned us that the North Atlantic part of this conveyor depended upon the presence of Arctic conditions near Greenland. The waters that flow north on the surface of the Atlantic are warm and lose water by evaporation and so become saltier; salt water is denser than fresh water and it would sink were it not that the cold waters beneath are denser still. When this warm dense salt water is cooled by contact with the Arctic ice it sinks to the bottom of the ocean; the sinking provides the force that drives the conveyor and keeps moving the warmer salt water that drifts north-eastwards across the Atlantic, what we call the Gulf Stream. Broecker warned that if the down flow of cooled salt water ceased, northern Europe would no longer receive the benefit of this flow of warm water. Sensational fiction often portrays this as the return of Arctic conditions to northern Europe and the east coast of North America. But of course by the time it happens the Arctic ice will be well on the way to disappearance. I can’t help wondering if the climate of the British Isles and the western part of northern Europe, which is now 8°C warmer than the same latitudes in other parts of the world, may be largely unchanged by global heating, because the 8°C lost when the Gulf Stream fails is just about equal to the predicted rise of temperature from global heating. Perhaps this is no more than wishful thinking, and we will certainly have to pay through the loss of land as the ocean rises to repossess it.
When we talk of climate change we often think more about the temperature and less about changes in the other qualities of the physical environment. Kangsheng Wu has pointed his research to the fresh-water balance of the world and reported a persistent increase in the flux of fresh water to the oceans particularly in the North Polar basin. The freshening of these northern waters might alter the course of the Gulf Stream. In a similar way, increasing warmth may expand the Hadley cells (see pp. 102–4) and so cause a migration of the trade winds and the westerly winds to zones nearer the poles. Changes in these other properties of climate will surely happen as the Earth heats. The planners of large schemes for renewable energy using wind and water power need to keep in mind the likelihood that they may become expensive mistakes.
While we cannot go back to the achingly beautiful world of 1800, when there were only one billion of us, we may not be incapable of lessening the consequences of global heating. If there is a threshold and we pass it, the nations of the world could limit the damage by stopping carbon dioxide and methane emissions; the temperature rise would then be slower, as would the rise of sea level, and it would take longer to reach the final steady hot state than it would if we continued business as usual. Even so, enormous damage would still have been done. In a later chapter I will discuss proposals to use either terrestrial or space-mounted sunshades to cool ourselves back to pre-industrial temperatures. But even if we succeeded we would find ourselves saddled with the appalling responsibility of managing the Earth’s climate, something that previously was provided free by Gaia, and we would still need to remove carbon dioxide from the air to prevent the poisoning of contemporary ocean life.
Recently the BBC broadcast in their Horizon series of science programmes an account of ‘global dimming’; in it climate scientists, among them V. Ramanathan and Peter Cox, voiced their concern that we have already, in a sense, passed the point of no return in global heating. The science behind this programme appeared in a Nature article in 2005 which included as an author the distinguished German scientist M. O. Andreae. Industrial civilization has released into the atmosphere, in addition to greenhouse gases, a huge quantity of aerosol particles, and these tiny floating motes reflect incoming sunlight back to space and cause global cooling. On large areas of the Earth’s surface the aerosol haze reflects sunlight back to space sufficiently to offset global warming. By themselves they cause a global cooling of 2 to 3°C. Back in the 1960s, when we knew much less about the Earth and its atmosphere, a few scientists even speculated that continued economic growth would increase the density of the aerosol and lead to global cooling and even precipitate the next glaciation.
The present extent of aerosol cooling is real and seriously worrying. It may have allowed us to continue our business as usual, not noticing how much we had changed the Earth nor realizing that we would have to pay back the borrowed time. Aerosol particles stay only a brief time in the atmosphere: within weeks they settle to the ground. This means that any large economic downturn, or a planned reduction in fossil-fuel usage, or unwise legislation to stop sulphur emissions, as the Europeans are now enacting to stop acid rain, will allow the immediate expression of greenhouse warming. It has been suggested that part of the excessive heat of the 2003 summer in Europe was caused by the European Union’s efforts to remove the aerosol which is the source of acid rain. Peter Cox reminded us that because the aerosol was not fully included, climate modellers may have underestimated the sensitivity of their models to greenhouse-gas abundance and failed to notice that we may already be beyond the point of no return.
Predictions of climate change do not depend only on theoretical models in the form of computer simulations of the Earth. There is now a vast array of monitoring activities sustained globally. Air and sea temperatures are continuously measured, as are the gases of the atmosphere, the cloud cover, the floating ice and the glaciers and the health of the ecosystems in the ocean and on the land. The truth of the models is therefore continuously tested against the observations coming in from the real world. Satellites orbiting the Earth monitor its ever-changing scene. The more subtle instruments aboard these spacecraft monitor temperatures at different levels in the air and many different atmospheric gases; they also check the health of ecosystems. I often think that the great unsung wonder of the space programme is the way it has revealed so much about the Earth.
Another important source of information about the cause of climate change is the long-term geological record. We have learnt an immense amount about the history of the climate and the composition of the Earth’s atmosphere from the analysis of ice taken from the depths of the Greenland and Antarctic glaciers. The snow falling on the glaciers brings air with it in the spaces between the crystals. Each new snowfall buries its predecessor, and so the air becomes trapped in small sealed bubbles made of ice, so that there is a continuous record going back to snow falling one million years ago. The bubbles trapped in the ice of cores bored into the glaciers provide samples of past atmospheres, and from their analysis the composition of these past atmospheres is revealed. From this vast data bank we now have a record, not only of the principal gases, oxygen and nitrogen, but also of the trace gases, carbon dioxide and methane. Indirectly we can calculate the temperature of the Earth when the air was trapped, from the isotopic composition of the oxygen and hydrogen. There are also good ways for ascertaining the date of the air being analysed. In this great store of information we have evidence that gives confidence to our claim that temperature and carbon-dioxide abundance are closely correlated. We know that in the depth of the last glaciation carbon dioxide fell to 180 ppm, rose to 280 ppm after the ice age ended, and has risen now to 380 ppm as a result of our pollution. Already we have made as large a change in the atmosphere as occurred between the ice ages and the interglacials. If it stays at 380 ppm we might expect a comparable rise in temperature, but more probably as we continue to pollute it will rise to 500 ppm or more.*
Going further back in time, there have been hot spells similar to the one we think is now due. The most recent occurred fifty-five million years ago at the beginning of the geological period called the Eocene and is the subject of several papers by Professor Harry Elderfield of Cambridge University. It was in some ways similar to our pollution of the air now and was due to the release of between 0.3 and 3.0 terratons of fossil carbon (a terraton is a million million tons). The source of this huge emission of carbon gases is still under debate: it may have come from the deposits of methane (natural gas) held in an ice-crystal form called a clathrate, which lie on the ocean floor, or it may have been vented from rich carbonaceous deposits in the North Atlantic when heated by a subterranean volcano.*
For comparison, we have already released by fossil-fuel combustion and agriculture about half a terraton of carbon, a quantity within the range estimated for the Eocene hot event. There are differences between the Eocene catastrophe and our present-day pollutions; for example, in the Eocene it was mainly methane that entered the air, not carbon dioxide as now. Professor Elderfield uses the geological record to suggest that fifty-five million years ago the temperatures rose by about 8°C in temperate regions and 5°C in the tropics, and from a world that was somewhat warmer than now, with little in the way of polar ice; the disturbance lasted 200,000 years. The sudden release of methane at the beginning of the hot spell would have rapidly warmed the Earth by its strong infra-red absorption, but it would have oxidized in the air to carbon dioxide and water vapour, and it would have been the carbon dioxide that sustained the heat for so long a period. The removal of carbon dioxide from the air by its chemical reaction with calcium silicate in rocks is called by geologists ‘chemical rock weathering’. It is slow and takes about 100,000 years to remove 63 per cent of the gas. We now know from Gaia Theory that life on the land surface and in the soil actively accelerates rock weathering. The land and ocean surfaces during the hot spell of the Eocene were barren, and that may be why the increased carbon dioxide stayed in the air so long. In addition, the Earth stayed warm because other biological cooling mechanisms that operate on a healthy Earth were disabled during the hot period of the Eocene. If conditions now are equivalent to those of the Eocene emissions, we should be prepared for a hot spell as long as or longer than an ice age. Although the initial conditions of the Eocene event resemble those on Earth now, two important differences are that the sun is now about 0.5 per cent hotter than it was fifty-five million years ago, equivalent to about 0.5°C in global temperature; and we have changed about half the Earth’s land surface from natural forest into farmland, scrub and desert and consequently reduced the capacity of the Earth to regulate itself. There are now, in addition to carbon dioxide and methane, several other greenhouse gases whose presence in the air adds to global heating; these include the CFCs, nitrous oxide and other products of agriculture and industry.
The Earth has recovered after fevers like this, and there are no grounds for thinking that what we are doing will destroy Gaia, but if we continue business as usual, our species may never again enjoy the lush and verdant world we had only a hundred years ago. What is most in danger is civilization; humans are tough enough for breeding pairs to survive, and Gaia is toughest of all. What we are doing weakens her but is unlikely to destroy her. She has survived numerous catastrophes in her three billion years or more of life.
In spite of the heat there will still be places on Earth that are pleasant enough by our standards; the survival of plants and animals through the Eocene confirms it. It is possible that the British Isles, with its oceanic site and high latitude, will be one of these refuges, although it will be more an archipelago than the two main islands it now is. But if these huge changes do occur it seems likely that few of the teeming billions now living will survive.
I think it necessary to repeat that the smoothly rising temperature of the IPCC’s third report shows an estimated average change of the global climate, but what it does not show are unpredicted extremes, including flood events and storms of great severity. We should expect climate changes of a kind never even thought of, one-off events affecting no more than a region. The first of these was the unprecedented European heatwave of 2003 when over 30,000 died of hyperthermia. Swiss meteorologists put the odds against it as no more than an unusually hot spell, at 300,000 to one.
Slower, decadal fluctuations in climate also confound our predictions. In a July 2005 Science article, Reading University scientists Rowan Sutton and Daniel Hodson reported decade-long warming and cooling trends during the twentieth century in the North Atlantic climate and noted that the excessive heat of Europe in 2003 was in one of these warming periods, as was a similar period of warmth in the 1960s and 1970s. The present warm period follows a cooler climate in the 1980s. Variations of this kind are superimposed on the uprising curve of global heating, and we need to guard against an over-interpretation of unexpected warmth and cold as evidence for or against global heating.
In between the sceptics of global heating and those, like me, who are concerned at the possibility of drastic change, are the conservative climatologists who acknowledge global heating but think it unlikely to be severe. Among them are Tom Wigley and G. A. Meehl and his colleagues, who both have articles in the March 2005 issue of Science. These are good and thoughtful papers that forecast a world that will slowly heat by about 2°C and in which sea levels will rise between 10 and 30 centimetres by 2100, and assume rather drastic reductions in emissions. I surely hope that they are right, but I persist in my gloomier view of the future. I do so because several important properties of the Earth system may not have been included in their calculations. These are:
1) The possibility of the disappearance of the present man-made northern hemispheric aerosol. Because of its short residence time, an economic downturn or any of a number of disasters could cause it to decrease in a few weeks, leaving the greenhouse intact.
2) They may be neglecting the extent to which the Earth system is in positive feedback. This would make the sensitivity of their models to increasing greenhouse gases less than they expected.
3) They may not have included the feedbacks from the natural forests and the ocean algal ecosystems. These can make the forest a source of, instead of a sink for, atmospheric carbon dioxide when increasing heat causes the vegetation to die back. The same effect occurs with ocean algae as the seas warm and lessen the rate of pump down of carbon dioxide.
4) It is all too often assumed that the vast changes to the land surface made by agriculture and forestry have had little or no influence on the sensitivity and resilience of the Earth system. I think it probable that the replacement of natural ecosystems with farmland may have altered the dynamics of climate feedback.
In its existence the Earth has experienced many different climate regimes. Soon after life started, Gaia emerged as a regulatory system; we think this led to a profound change in atmospheric composition from one dominated by carbon dioxide to one dominated by methane, which lasted about a billion years until oxygen became the chemically dominant gas, low in abundance at first but ultimately rising to become the air we now breathe.
Because temperature is so important to living organisms, it strongly affects their distribution on the Earth. Photographs of the Earth from space taken to show only the distribution of chlorophyll, the green pigment that vegetable life uses to convert sunlight into organic matter, provide a good way to grasp the effect of temperature on the geographical distribution of life. Chlorophyll is an essential constituent of all the primary producers that use the energy of sunlight to make food from the raw chemicals of the ocean and atmosphere; the distribution of chlorophyll represents that of plants and algae. It also shows where the other forms of life are because they consume it for food directly or indirectly. Figure 5 shows three sketch maps drawn to compare the distributions of plant and ocean algal life on worlds five degrees cooler, as now, and five degrees hotter than now. The centre sketch illustrates how at present the Antarctic continent and much of the North Polar region are almost bereft of life; the greater part of the world’s oceans are also quite barren except for regions close to continents and in the cool waters nearer to the Arctic and Antarctic. The hot and dry deserts of Africa, Asia, North America and Australia are also sparsely populated with life. Abundant life occurs where it is warm and wet on land and where it is quite cool, less than 12°C, in the ocean.
Compare this with the two imaginary sketches: the lower an Earth 5°C hotter than now, roughly that predicted by the IPCC for the end of this century, and the upper 5°C cooler than now, close to the temperature of the last ice age. To judge from the abundance of life, Gaia seems to like it cold, which is why perhaps for most of the last two million years, and maybe much longer, the Earth has been in an ice age. I think it important that we recognize that a hot Earth is a weakened Earth. On the hot planet, ocean life is restricted to the continental edges, and the desert regions of the land are much extended.
What is remarkable is that life on Earth has persisted for nearly four billion years. This is a cosmic lifespan nearly a third as long as that of the universe itself. If life is so fussy about temperature then it implies that the Earth’s temperature cannot have changed much during life’s existence. We are tolerably sure that the sun, like all similar stars, warms up as it ages and is now 25 per cent hotter than when life began. This is equivalent to the temperature of the Earth’s surface rising by 20°C; so that if the Earth was in an ice age when life began at 12°C it would by now be 32°C; or if it was warm, say 25°C, when life began, it might now be 45°C. Both of these are far
Figure 5. The distribution of life now and on a hotter and a colder Earth.
above the 12°C of the glaciation and our present average of 16°C, and well above the 20°C expected from global warming.
This chapter has mostly dealt with climate change, but we should not forget that there are also large and disastrous changes in fresh-water abundance, including floods and droughts. It is still far from clear whether the observed rise of sea level of the past fifty years is due mainly to the expansion of ocean water as it warms or mainly to the melting of glaciers. The 2001 IPCC report suggested that expansion was the principal cause of sea-level rise, but a 2004 Nature paper by scientists at the National Oceanic and Atmospheric Administration in Washington, and an essay by Jim Hansen in Climate Change in 2005, both suggest that the increase of ocean volume is mainly from the melting of land-based ice. If Hansen is right, then rapidly rising sea levels, not foul weather, will be the greater threat. The recent devastation wreaked by hurricanes in the southern United States, especially in New Orleans, reminds us of the damage that temporary flooding can cause; excessive rain and surges of sea water driven by storms can be as disturbing as a permanent slow rise in sea level. All these changes in their turn alter the distribution of forests and deserts and the availability of land on which to grow food. Although we have much to learn, it does appear probable that in a few years, when the carbon-dioxide abundance passes 500 ppm, we will enter the zone where temperatures will rise to a new steady state, perhaps six to eight degrees hotter than now. We do not know if this new regime will be stable in the long term, and if we are foolish enough to continue trying to farm and pollute the air on the remaining habitable parts of the Earth, a final collapse might happen. Nothing in science is certain, but Gaia Theory is now robustly supported by evidence from the Earth and it suggests that we have little time left if we are to avoid the unpleasant changes it forecasts.