The collapse of the Amazon is part of the full reversal of the carbon cycle that is projected to happen at around 3 degrees of warming - a view confirmed by a range of researchers using carbon-coupled climate models. A 3-degree warming would see significantly large areas of the Earth's terrestrial environment rendered uninhabitable by drought and heat. Rainfall in Mexico and Central America is projected to fall by 50 per cent. Southern Africa would be exposed to perennial drought, and a huge expanse of land centred on Botswana could see a remobilisation of sand dunes, as is predicted to happen in the western US even sooner. The Rockies would be snowless, and water flows in the Colorado River would fail one year in two. Drought intensity in Australia could triple, and World Heritage ecosystems would severely degrade or die, while hurricanes could increase in power by half a category above today's top-level Category Five.

With such extreme weather conditions, world food supplies will be devastated. This could mean billions of refugees moving towards the mid-lat.i.tudes from areas of famine and drought in the sub-tropics. As rising temperatures cause the Himalayan ice sheet to melt, long-term water-flows into Asia's great rivers and breadbasket valleys - including the Indus, Ganges, Brahmaputra, Mekong, Yangtse, and Yellow rivers - will fall dramatically. It has been predicted that if global temperatures rise by 3 degrees, which is becoming the unofficial target for some governments of richer nations, water flow in the Indus would drop by 90 per cent by 2100. But the loss of the Himalayan ice-sheet now looks likely to occur at well less than a 3-degree rise. Recent estimates are that the Himalayas may be completely ice-free before 2050, or even sooner. The lives of two billion people are at stake.

For all this, 3 degrees is the cap effectively being advocated by Australia's Labor government. In its 2007 pre-election policy, Labor advocated a 60 per cent reduction by 2050 in Australian emissions from 2000 levels. Environment Minister Penny Wong then reaffirmed this, in February 2008, when she tersely rejected suggestions from the Garnaut Review that the cut might need to be 90 per cent.

The goal of a 60 per cent reduction in emissions by 2050 (known as '60/2050') for fully developed nations was first formally articulated by a major organisation in 2000, when it was recommended by the UK Royal Commission on Environmental Pollution. The Rudd government's policy statement makes reference to this; however, the core idea - to make a 60 per cent cut in carbon dioxide emissions compared to 1990 levels - had been given prominence a decade earlier, in the first science a.s.sessment of the IPCC. This was not presented as a goal, as such - it was provided by the scientists to help policy-makers take in the scale of the challenge.

The immediate source of inspiration for Labor's 60/2050 target appears to be the Stern Review, which advocated a 3-degree target. In the report, Stern stated that constraining greenhousegas levels to 450 parts per million carbon dioxide equivalent 'means around a fifty-fifty chance of keeping global increases below 2 degrees above pre-industrial [and it] is unlikely that increases will exceed 3 degrees'; but, he said, keeping to this is 'already nearly out of reach', because it means 'peaking in the next five years or so and dropping fast'. It would require immediate and strong action, which Stern judged to be neither politically likely nor economically desirable.

Instead, Stern pragmatically said the data 'strongly suggests that we should aim somewhere between 450 and 550 parts per million carbon dioxide equivalent'. However, his policy proposals demonstrate that he has the higher figure in mind as a practical goal: 'It is clear that stabilising at 550 parts per million or below involves strong action ... but such stabilisation is feasible'. Stern's policy framework is focused on constraining the increase to 550 parts per million, at which level, he argued, 'there is around a fiftyfifty chance of keeping increases below 3 degrees [and it is] unlikely that increases would exceed four degrees'.

The link between a 60 per cent emissions reduction by 2050 and a 3-degree cap was reiterated during Stern's March 2007 visit to Australia. During an address to the National Press Club in Canberra, he said it would be 'a very good idea if all rich countries, including Australia, set themselves a target for 2050 of at least 60 per cent emissions reductions'. This would leave the planet with about 550 parts per million of carbon dioxide equivalent by 2050, and would leave us with 'roughly a fifty-fifty chance of being either side of 3 degrees above pre-industrial times'.

But according to a draft paper released in 2008 by James Hansen and seven other climate scientists, long-term climate sensitivity of 6 degrees and a doubling of pre-industrial carbon dioxide levels to 550 parts per million would produce a very different planet. Hansen reminds us that 'the last time the planet was five degrees warmer, just prior to the glaciation of Antarctica about 35 million years ago, there were no large ice sheets on the planet. Given today's ocean basins, if the ice sheets melt entirely, sea level will rise about seventy meters'. This would be the likely outcome of Stern's policy and, seemingly, also that of Australian government policy.

A number of others have followed Stern's lead. These include the former head of the Australian Bureau of Agriculture and Resource Economics, and Australia's lead delegate to the May 2007 IPCC meeting, Brian Fisher. He says the 2-degree target, with emissions peaking by 2015, 'is exceedingly unlikely to occur ... global emissions are growing very strongly ... On the current trajectories you would have to say plus 3 degrees is looking more likely'.

When Labor announced its 60/2050 target, it made a number of confusing and conflicting claims based on Stern's findings and various CSIRO reports. These included limiting future increases in atmospheric carbon dioxide to 550 parts per million, and setting a target range of 450550 parts per million carbon dioxide. It also claimed to be concerned about the impacts of a 3-degree increase, and warned of 'the need for reductions in annual GHG [greenhouse-gas] emissions of 60 to 90 per cent from 1990 or 2000 levels by 2050 for countries listed under Annex 1 in the Kyoto Protocol'.

Labor also drew on the 2000 UK Royal Commission's report on Environmental Pollution, which set a cap of 550 parts per million carbon dioxide. This is odd, because in the world of climate-change science, and politics, that report is now very old - it relied on an IPCC report now 12 years out of date. Since 2000 there have been two more IPCC reports, the research has rapidly moved on, and the British government has reduced its emissions target to 450 parts per million carbon dioxide; but Labor does not refer to more recent and relevant European research. It does not mention, for example, Meinshausen's contribution to the Stern Review, which says that if greenhouse gases reach 550 parts per million carbon dioxide equivalent, there is a 6399 per cent chance that global warming will exceed 2 degrees.

Nor did the Labor policy statement address the UK's own recognition of error, as George Monbiot noted in the Guardian: 'The British government has been aware that it has set the wrong target for at least four years. In 2003 the environment department found that "with an atmospheric carbon dioxide stabilisation concentration of 550 parts per million, temperatures are expected to rise by between two and five degrees". In March 2006 it admitted "a limit closer to 450 parts per million or even lower, might be more appropriate to meet a 2-degree stabilisation limit".'

What Australian Labor did was establish a target of 3 degrees and 550 parts per million, but they dressed it up as if it was aiming for something lower. Their pre-election statement says: In 2006, CSIRO's Climate Change Impacts on Australia and the Benefits of Early Action to Reduce Global Greenhouse Gas Emissions concluded that: 'Limiting future increases in atmospheric carbon dioxide to 550 parts per million, though not a panacea for global warming, would reduce 21st century global warming to an estimated 1.52.9 degrees, effectively avoiding the more extreme climate changes'.

This is misleading. The report referred to actually says: As mentioned previously, some nations view 60 per cent reductions by 2050 as consistent with placing the world on a path to achieving a 550 parts per million carbon dioxide stabilisation level. According to climate model results ... this level of mitigation would limit 21st century global warming to 1.52.9 degrees, with an additional 0.30.9 degree of warming in subsequent centuries [our emphasis].

Throughout the CSIRO doc.u.ment, temperature increases are taken from a 1990 baseline (0.6 degrees at 1990), so that the phrase '21st century global warming to 1.52.9 degrees' means a total rise over pre-industrial levels of 2.13.5 degrees by 2100. Add in the 'additional 0.30.9 degree of warming in subsequent centuries', and the full temperature rise range becomes 2.44.4 degrees for 550 parts per million. This would clearly const.i.tute dangerous climate change, according to anyone's measure.

This slipping and sliding of parameters, and the shift in the pragmatic goal from 2 degrees to 3 degrees, is also evident in the 2007 IPCC report on adaptation and mitigation. Of the 177 research scenarios a.s.sessed for future emissions profiles, none dealt with a target of less than 2 degrees, and only six dealt with limiting the rise to the range of 22.4 degrees. By contrast, 118 scenarios covered the range of 3.24 degrees, which suggests that the IPCC scientists, following the lead of the politicians, have also largely shifted the focus away from targets of less than 2 degrees.

The effect that carbon levels and temperature increase have on ocean algae introduces another perspective to the dialogue: what if a target of 550 parts per million were to result in the destruction of the Earth's greatest carbon sink? James Lovelock, environment scientist and proposer of the Gaia hypothesis, claims that as the ocean surface temperature warms to over 12 degrees, 'a stable layer of warm water forms on the surface that stays unmixed with the cooler, nutrient-rich waters below'. This purely physical property of ocean water, he says, 'denies nutrients to the life in the warm layer, and soon the upper sunlit ocean water becomes a desert'.

This chlorophyll-deprived, azure-blue water is currently found predominantly in the tropics, which lacks the richness of the marine life of the darker, cooler oceans. In this nutrient-deprived water, ocean life cannot prosper and, according to Lovelock, soon 'the surface layer is empty of all but a limited ... population of algae'. Algae, which const.i.tute most of the ocean's plant life, are the world's greatest carbon sinks, devouring carbon dioxide while releasing dimethyl sulphide (DMS), which is transformed into an aerosol that contributes to greater cloud formation and, hence, affects weather patterns. The warmer seas and fewer algae that Lovelock predicts are likely to reduce cloud formation and further enhance positive climate feedbacks.

This process should be distinguished from the phenomenon of green, red, or brown algal blooms, which can occur in fresh and marine environments when phytoplankton a.s.sume very dense concentrations due to an excess of nutrients in the water. The dead organic material becomes food for bacteria, which can deprive the water of oxygen, destroying the local marine life and creating a dead zone.

Because algae thrive in ocean water below ten degrees, the algae population reduces as the climate warms. Lovelock says that severe disruption of the algaeDMS relation would signal spiralling climate change. Lovelock and k.u.mp's modelling of climate warming and regulation published in Nature in 1994 supported this: [A]s the carbon dioxide abundance approached 500 parts per million, regulation began to fail and there was a sudden upward jump in temperature. The cause was the failure of the ocean ecosystem. As the world grew warmer, the algae were denied nutrients by the expanding warm surface of the oceans, until eventually they became extinct. As the area of ocean covered by algae grew smaller, their cooling effect diminished and the temperature surged upwards.

According to Lovelock, the end-result was a temperature rise of 8 degrees above pre-industrial levels, which would result in the planet being habitable only from Melbourne to the South Pole (going south), and from northern Europe, Asia, and Canada to the North Pole (going north).

The likelihood of dramatic effects is beginning to be recognised more widely. Stern is now saying that his 2006 Review substantially underestimated the growth rate of greenhouse gases; the impact of greenhouse gases on the levels of warming; and the degree of damage, and the risks, of climate change.

He says it also overestimated the capacity of the carbon sinks to absorb carbon dioxide, and he now warns that a 5-degree temperature increase would, most likely, transform the physical and human geography of the planet, leading to ma.s.sive human migration and large-scale conflict. Yet Stern still formally advocates the 550-parts-per-million target that he proposed in 2006, even though the data that he now uses in presentations shows that this target carries a 41 per cent chance of exceeding 5 degrees!

One possible future is that the world will fail to recognise the danger posed by a temperature rise of 3 degrees or more, and will let greenhouse-gas levels rise to 550 parts per million. In that case, it would take a long time, even under a crash emissions-reduction program, to draw down the excess carbon dioxide. As temperatures rose, driven by positive feedbacks, declining carbon sinks, and non-linear events, the climate system would have so much momentum that we would be unable, effectively, to apply the brakes at the 3-degree signpost.

In 2004, Tom Athanasiou and Paul Baer, co-founders of the climate-change social justice group EcoEquity, encapsulated the absurdity of the dilemma: We'd all vote to stop climate change immediately, if we only believed that doing so would be so cheap that no country or bloc of countries could effectively object. But we do not so believe. Thus we're forced to start trading away lives and species in order to advocate a 'reasonable' definition of 'dangerous' ... So it's no surprise that ... the advocates of precautionary temperature targets strain to soft-pedal their messages, typically by linking 2 C of warming to carbon dioxide concentration targets that can be straight-forwardly shown to actually imply a larger, and sometimes much larger, probable warming... Climate activists soft-pedal the truth because they think it will help, and perhaps they are even right. Who are we to know? Nevertheless, we also believe that the waffling is becoming dangerous, that it threatens, if continued, to critically undermine the coherence of our emerging understanding. That it delays difficult, but necessary, conclusions.

CHAPTER 12.

Planning the Alternative.

What should we do once we acknowledge that 2-degree and 3-degree targets are too high, and we know that the climate system seems likely to reach more than 2 degrees of warming?

Efforts to tackle climate change, so far, have been aimed at creating a 'less bad' outcome. Society seems to be preparing simply to head into the catastrophe more slowly, which does not seem to be a very practical strategy. The alternative would be to aim for the future we want: a safe climate.

A safe climate does not involve losing the Himalayan glaciers, or endangering food production in much of China, Bangladesh, India, and Pakistan. It does not involve losing the West Antarctic ice sheet, or converting the Amazon to dry gra.s.sland. It does not involve releasing ma.s.sive amounts of carbon dioxide and methane into the air by melting permafrost, or degrading nature's carbon sinks; and it does not involve having an ice-free Arctic in summer.

But to reverse or prevent these conditions is clearly a very challenging task. Perhaps it is even an impossible task.

If it had been suggested 50 years ago that humans should set out to remove the Arctic ice cap and warm the entire globe by 12 degrees, people would have said that this was crazy and physically impossible - that it should not and could not be done.

Fifty years on, we are well on the way to 'succeeding' in this project.

Humans now have the most powerful economy of all time. If we choose to apply this economic power to create a safe climate, and we act decisively before uncontrollable natural feedbacks are set fully in motion, we could succeed.

The first task is to understand what a safe climate means and what action is required to achieve it.

The safe-climate zone: When considering climate change, we can identify a range of climate conditions that are safe. Our goal must be to keep environmental conditions within that safe-climate zone, or return to that zone if conditions have, or are likely to, stray beyond it. We also need to be aware of the speed and momentum of changes in the climate system now, as it moves away from the safe zone, and later, when thinking about actions that will move it back to the safe zone.

What is to be protected? If climate goals are to be well formed, their underlying values need to be explicit; for example who or what are we intending to benefit? What level of risk of adverse outcomes is acceptable?

In public discussion about climate change, it is clear that motivations for action include concerns for people in various parts of the world, for other species, and for current and future generations. These concerns can be amalgamated into a concern to protect the welfare of 'all people, all species, and all generations'.

All people: Although the a.s.sumption in international climate negotiations is that policy is designed to benefit all people, in practice, some nations - especially the rich, high-polluting ones - plead for exceptions and special circ.u.mstances, with some success. For the international reality to shift, all the major national players must start negotiating for the global 'common good' as an extension of their self-interest. For political elites in the developed world, the motivating factors need to be both altruistic and self-interested. Compa.s.sion for people in every part of the globe must coincide with enlightened self-interest, because failure to act for all people will result in a global food crisis, ma.s.s flows of environmental refugees, and possible armed conflicts.

All species: In protecting other species for their own sake, we also protect ourselves, for we all belong to interdependent ecological systems. On the other hand, the elite in most countries are so reticent about acknowledging the significance of other species that their proposed climate solutions will likely take little, if any, account of the need to protect them. What we need is an ethic of cross-species compa.s.sion, together with the realisation that our own interest also requires the protection of all species.

All generations: In modern societies, we tend to act on the a.s.sumption that if we look after our own best interests, the future will look after itself; but this approach will not work with climate change. We need to consider the needs of future generations - both people and other species. The work of the IPCC is an interesting example. Its models and projections are run to the year 2100, a vastly longer timeline than is considered in most political processes; however, when it comes to climate change, this is not long enough. When climate impacts are suddenly found to be running 100 years ahead of schedule, as is the case in the Arctic, we are shocked, because we have to respond to events that were barely on the radar. One solution would be to follow climate models through until a stable, long-term state is reached, so that we understand the implications for all generations, rather than just for those who will live through the next 100 years. If tackled this way, we would quickly realise, for example, that the politically 'tough' target of 450 parts per million carbon dioxide would likely result in an ice-free world and a real chance of an 87-metre sea-level rise if climate sensitivity is on the high side.

Finally, of course, there is the overarching need to maintain human civilisation, in the best sense of the acc.u.mulated knowledge and wisdom that, among other benefits, informs our capacity to feed and protect the health of large numbers of people now alive.

Risk: As we strive to protect all people, species, and generations, we need to know how much risk people and species will be exposed to, and what humanity would consider acceptable. When approving new pharmaceuticals, and designing aircraft, bridges, and large buildings, strict risk-standards are applied: a widely used rule of thumb is to keep the risk of mortality to less than one in a million. The Apollo Program, for example, aimed to keep the risk of Saturn rockets plunging into population centres to less than one in a million. When it comes to climate change and the viability of the whole planet, it makes no sense to apply a lesser standard of risk aversion. We should aim, for example, to have less than a one-in-a-million chance of losing the Greenland and West Antarctic ice sheets, or of failing to recover the full extent of the Arctic summer sea-ice.

So far, however, governments have been accepting much higher risks in setting global-warming targets; take Stern's promotion in January 2008, for example, of a 550 parts per million carbon dioxide target which, by his own admission, means accepting devastating species loss, as well as coral reef destruction, ice-sheet disintegration, and economic damage 'on a scale similar to [that] a.s.sociated with the great wars and the economic depression of the first half of the 20th century'.

Insurance Australia Group actuary Tony Coleman says insurers are familiar with managing risks to our community that are potentially catastrophic, yet, he says, when it comes to climate risk, we seem to have different parameters: '... Australia is tolerating a level of climate change risk that would be unthinkable if the nation was held to the same standards that we apply to safeguard the survival of the insurers, banks and superannuation funds that we all depend upon in our daily lives.' These levels of risk, which are less than one in 200, says Coleman, 'are completely dwarfed by the risk levels to our way of life that are now reliably attributable to potentially catastrophic climate change impacts, unless we act with urgency to rapidly reduce greenhouse emissions'.

We insist on standards of safety for individuals that are many times higher than the standards we apply to all humans and to the ecological life support systems that we, and other species, depend on.

We should not accept actions that could trigger an irreversible chain of climate-change events or produce dangerous impacts. We cannot gamble on how far we can push the system before it breaks. As is the case for civil aviation, climate-change safety policy must allow for less than a one-in-a-million chance of catastrophic failure.

Speed of transition: The speed of our transition to a safe-climate zone is also a critical issue. The risks posed by allowing the world to stay outside the safe-climate zone need to be a.s.sessed, along with the impacts that would be generated, ecologically and socially, by speeding up the transition. We know, for example, that species losses increase the faster that temperatures change, and we need to weigh up those sorts of impacts alongside the risks a.s.sociated with a slower transition.

Dangerous climate change versus the safe-climate zone: How does the achievement of a safe climate relate to the more widely held goal of avoiding dangerous climate change?

The core objective of the UN Framework Convention on Climate Change, which also governs the Kyoto Protocol, is to achieve: stabilization of greenhouse gas concentrations in the atmosphere at a level that would prevent dangerous anthropogenic interference with the climate system. Such a level should be achieved within a time-frame sufficient to allow ecosystems to adapt naturally to climate change, to ensure that food production is not threatened and to enable economic development to proceed in a sustainable manner.

It is self-evident that the world should act to avoid 'dangerous anthropogenic [human] interference with the climate system', but making that our primary goal has created problems.

The theory of the greenhouse effect was developed more than a hundred years ago by the Swedish chemist Svante Arrhenius. In the 1960s, scientific interest in the theory began to intensify when the American scientist Charles Keeling published his findings, which showed that carbon dioxide levels in the atmosphere were systematically rising from the 315 parts per million that he had first observed in 1958. Thirty years later, scientific knowledge was strong enough to lead to the formation of the IPCC, by which time the atmospheric carbon dioxide level was 350 parts per million.

Although concern about global warming was now growing strongly, it was also clear that changing human behaviour enough to stop, or slow, greenhouse-gas emissions would require very significant changes to the economy, and that these changes would be resisted strongly. So, scientists focused their message on the need to avoid dangerous climate change - a message that proffered threats big enough to grab the attention of the political elite and, perhaps, convince them that matters other than short-term economic gain should be considered. Corporate elites, however, maintained widespread resistance to their message, so scientists opted to focus on only the strongest concerns about climate change: how could we avoid the loss of the Greenland and the West Antarctic ice sheets and the metres of sea-level rise that would occur as a result? How could widespread ecosystem collapse be avoided?

The result of this tactic was that scientists began to focus on action to avoid outright climate catastrophe. In response, the politicalcorporate elite set targets for change just a fraction under the levels that the scientists identified as having catastrophic consequences. Once these targets were articulated - for example, an upper-warming limit of 2 degrees, or 550 parts per million atmospheric carbon dioxide - policy inertia tended to lock them in, regardless of later changes in scientific knowledge. Scientists' concerns about identifying dangerous climate change, and about the measures necessary to avert it, were transformed into a process for avoiding catastrophe, or apocalypse, in some far-distant future. As a result, unrealistic targets have been set that, even if achieved, would see civilisation-destroying climate change.

An alternative approach would identify climate conditions that are known to be safe, and then make it the goal of public policy to get back into this safe-climate zone and avoid leaving it again. Instead of scientists being asked to identify what elevated greenhouse-gas levels might be bearable (should we stabilise at 450, 550, or 650 parts per million?), the safe-climate approach would be to ask what actions are necessary to get back to the zone in which greenhouse-gas levels are known to be safe.

The danger of tipping points: a major concern is the possibility that key elements of the Earth system could go through critical thresholds or tipping points (as discussed in Chapter 10) that lead to a significant increase in warming processes, such as a big jump in greenhouse-gas emissions, or to a major change that severely harms other species or human societies.

The evidence is clear that the Earth's biosphere is already in a state of dangerous climate change. Current impacts - including desertification and water shortages, extreme weather events, severe and frequent bushfires, ecological breakdown, difficulties with food production, and changes to major geophysical systems such as the Arctic - are already causing problems in many parts of the world. Pressures are building in the Earth system that will give way to even bigger changes in temperature and the environment, which means that the problems already causing concern are a relatively mild foretaste of what will come if the economy and the climate system are left to follow current trends.

While climate danger is generally cast as occurring at some time in the future, climate change is already dangerous for some people: the populations of the small nations of the Pacific, who are already abandoning their low-lying island atolls because rising sea levels and storm surges make life there impossible; people in sub-Saharan Africa badly affected by extended drought; and the Inuit people of the Canadian Arctic who can no longer move safely across the sea-ice to hunt, and whose homes are cracking and tipping as the permafrost melts.

The task, now, is to establish the boundaries of the safe-climate zone. Policy and action should be framed: * to protect all people, all species, and all generations; * to accept an even smaller risk of failure than the best-practice safety standards for the protection of people in civil engineering (the one-in-a-million principle) in avoiding dangerous changes to the Earth caused by climate change; and * to keep the Earth in the safe-climate zone, rather than to simply avoid dangerous climate change.

CHAPTER 13.

The Safe-Climate Zone.

For the past 100,000 years, humans and their predecessors have survived and adapted as the Earth's temperature has fluctuated by up to 7 degrees. The current global average temperature is within 1 degree of the maximum temperature known to have occurred during the past million years, but conditions 6 degrees colder were experienced during the depths of the recurring ice ages. At a cold point 20,000 years ago, so much ice was stacked on the land that sea levels were 120 metres lower than they are now. On the other hand, 125,000 years ago, when temperatures were similar to today, the sea level was 56 metres higher.

The past 11,500 years since the last ice age is known as the Holocene - a period that coincides with the establishment of human civilisation. During the Holocene, temperatures have varied within a 1-degree band, although the variation has, for the most part, been considerably less. Sea levels have been almost constant over the last few thousand years of human civilisation and, more significantly, over recent centuries, when most climate-sensitive infrastructure has been built. Coastal cities, including the special case of Venice, shipping facilities, the permanent settlement of river deltas, and other low-lying areas have survived because sea levels have moved very little.

Increasingly, however, human activity is changing the surface of the planet and also, consequently, the climate. Today we see that impact around the globe. Large parts of the land have been taken over by humans for grazing and cropping, and for cities. Wetlands have been drained on a huge scale; rivers have been regulated with dams; and forests have either been cleared, or cut into small patches by roads and clearings. As we've extracted and processed resources, and thrown away our wastes, our natural world has become very fragile, fragmented, and impacted by chemical and physical a.s.saults.

The nation-states and the vast, fixed physical infrastructure of cities and roads that human civilisation has built will make it very hard to adapt and move across the continents if the climate were to become more changeable - if, for example, it began to swing between warm periods and glacial periods, such as those that left much of North America and northern Europe under metres, and sometimes kilometres, of ice, 20,000 years ago. While we might adapt to lower sea levels, higher seas would be catastrophic for whole cities, farming communities, nations, and coastal-wetland species.

Given our sedentary pattern of living, how can we identify a band of environmental conditions that defines a contemporary safe-climate zone? Would the relatively stable climate pattern of the Holocene and its development of agriculture and civilisations be appropriate? Can we tolerate today's temperature, which is at the top end of the Holocene range? Should we accept a summer-ice-free state in the Arctic as a normal part of the range of conditions to be included in the safe-climate zone? To maintain the Earth system's resilience, is it ecologically necessary to cycle through a summer-ice-free state periodically?

Avoiding a summer-ice-free Arctic.

During the past million years, the Arctic has been partially free of summer sea-ice for short periods, but today's circ.u.mstances are very different. In the past, this event represented the gently sloping top of the warming hill; now, however, the level of greenhouse gases, and the upward pressure on temperatures, is substantially higher. What is more, the temperature is charging through this barrier with the human foot still pressing on the emissions accelerator. The real risk is that, rather than mark the natural peak of the temperature cycle between periods of ice ages, a summer-ice-free state in the Arctic will kick the climate system into run-on warming and create an aberrant new climate state many, many degrees hotter. The last time such a warming occurred - many tens of millions of years ago - many plants and animals became extinct around the world.

An Arctic free of summer sea-ice cannot, then, be considered part of the safe-climate zone, and urgently restoring its full extent is necessary to avoid significant ecological damage and, possibly, catastrophic greenhouse heating.

In defining the safe-climate zone, it is more important to identify tangible elements of the environment that need to be restored and maintained, rather than just to focus on temperature and carbon dioxide levels.

Some features of a safe-climate policy would include: * retaining the full summer Arctic sea-ice cover, the full extent of the Greenland and Antarctic ice sheets, and the full extent of the mountain glacier systems, including the Himalayas and the Andes; * maintaining the ecological health and resilience of the tropical rainforests and coral reefs, with no loss of area or species; * maintaining the health and effectiveness of the natural carbon sinks, at least, to their level of 50 years ago; and * capping ocean acidity at a level that prevents any risk to organisms.

The appropriate temperature range and climate-system settings compatible with the maintenance of these environmental features can then be determined using the best available climate science, with a risk of loss of less than one in a million. Here are three ways of thinking about this range: The Hansen Arctic threshold: In the draft paper released in April 2008, James Hansen and seven co-authors say that a carbon dioxide level of '300325 parts per million may be needed to restore [Arctic] sea ice to its area of 25 years ago'. In other words, the amount of carbon dioxide in the atmosphere would need to be significantly reduced from the current level of 387 parts per million.

Maintaining Arctic sea-ice thickness: The Arctic sea-ice thinned substantially from about 3.5 metres in the 1960s to about 2.5 metres by the end of the 1980s, which was well before the beginning of the dramatic decline in ice-surface area that became apparent from the mid-1990s onwards. In the late 1980s and early 1990s, shifting wind patterns flushed much of the thick, older sea-ice out of the Arctic Ocean and into the North Atlantic, where it eventually disintegrated, replaced by a thinner layer of young ice that melted more readily in the succeeding summer. Mark Serreze from the University of Colorado says that 'this ice-flushing event could be a small-scale a.n.a.logue of the sort of kick that could invoke rapid collapse, or it could have been the kick itself '. Pulses of warmer water that began entering the Arctic Ocean in the mid-1990s, which promote ice melt and discourage ice growth along the Atlantic ice margin, are 'another one of those potential kicks to the system that could evoke rapid ice decline and send the Arctic into a new state', according to Serreze. In 1989, the global average temperature was about 0.3 degrees cooler than it is currently. To restore this temperature, it would be necessary to drop carbon dioxide levels to 315 parts per million.

An insight from the early Holocene? For part of the period from 60008500 years ago, the Arctic warmed to the point that it was largely free of sea-ice each summer. A Dutch Danish scientific team, using plant-fossil data, estimates that the carbon dioxide level during this time ranged from about 325 parts per million to a less-well-defined lower level that allowed the summer sea-ice to return.

While more research is needed before the boundaries of the safe-climate zone can be set definitively, it is reasonable and prudent to conclude from these three case studies that we should aim, initially, for at least a 0.3 degree cooling to bring the global average temperature-increase above pre-industrial levels to less than 0.5 degrees. To bring the planet within reach of this temperature, the atmospheric carbon dioxide level should be under 325 parts per million - the level that Hansen is arguing is needed to fully restore the Arctic ice.

This would also be a reasonable boundary for avoiding a range of other major climate problems, including the loss of the mountain glaciers, and the Greenland and West Antarctic ice sheets; damage to tropical rainforests; and a decline in the capacity of carbon sinks.

Hansen, discussing the impending loss of the Arctic summer sea-ice in October 2007, noted that the climate system is dominated by positive feedbacks - knock-on effects that exaggerate the current trend of the climate. These feedbacks run in both directions, so if enough of the strong, high-inertia warming feedbacks were stalled, or turned around, and the Earth was cooled for a while, the climate system would then run in the opposite direction. If humans decided to initiate a sufficient cooling, natural feedbacks would complete the job.

Cooling the Earth.

The Earth is already too hot, and there's already too much carbon dioxide and other greenhouse gases in the atmosphere. The first key step to fix this is to stop adding to the heating processes - greenhouse-gas emissions need to be cut to zero. The second step is to remove from the air the excess carbon dioxide that is keeping the planet too hot. The third step, because time is short and there is already so much heat in the system, may be for humans to cool the Earth directly.

To cut greenhouse-gas emissions, we will need to reduce those warming agents that have a short life. Methane, for example, has a relatively short life in the atmosphere of about a decade, so cutting methane emissions would have an effect relatively quickly. Measures to achieve this include stopping coal, oil, and gas mining (to stop methane leakage); re-engineering waste disposal (trapping methane as an energy source); changing irrigation methods and varieties of rice cultivation; and decreasing the commercial herding of ruminant animals, especially cattle.

We must also stop emitting greenhouse gases, including carbon dioxide, and heating agents, such as black soot, urgently. This is essential, because carbon dioxide is acidifying the upper ocean, preventing marine organisms from forming calcified sh.e.l.ls and exoskeletons. If this continues it will lead to major marine animal and plant extinctions in the not-too-distant future. Black soot is a short-lived warming agent that is washed out of the air by rain in a matter of days; cutting its emissions would have an immediate effect. By dirtying ice, black soot also accelerates glacier and ice-sheet melting - particularly in the Himalayas, because one-third of black-carbon emissions come from India and China. Programs to cut black-soot emissions - for example, by ending the use of coal for heating, stopping diesel use, and by providing energy-efficient and smoke-free cookers to rural communities across Asia - would have an immediate and dramatic effect in reducing the heating effect.

Zero-carbon Britain: an alternative energy strategy, published in 2007, is one of many research reports that demonstrate the feasibility of building a post-carbon economy. Many of the practical technologies and solutions are also surveyed in Chapter 20 of this book.

We must also remove excess carbon from the air. We cannot return to a safe climate if we only cut emissions to zero, because carbon dioxide remains in the atmosphere for so long. Estimates by Matthews and Caldeira from Stanford University indicate that around 200 billion tonnes of excess carbon needs to be drawn out of the atmosphere to achieve the 0.3-degree decrease in the global temperature that is necessary.

Techniques for trapping carbon that is already in the atmosphere include boosting the natural terrestrial processes (re-afforestation); and producing agricultural charcoal, known as bio-char, which is sequestered in the ground.

Such large-scale, relatively low environmental-impact methods depend on growing plants that naturally absorb carbon dioxide. Growing these in the extremely large quant.i.ties necessary to draw down substantial amounts of carbon, however, may conflict with land use for nature conservation or food production. This, along with issues such as water availability and social impacts, needs to be considered in planning such schemes.

Although it is necessary to reduce human greenhouse-gas emissions to zero as quickly as possible, there is a critical side effect. Most carbon dioxide generated by human society is produced by deliberately burning fuels such as coal, oil, gas, and wood, or by unintentionally burning plant material in bushfires, but these processes also produce aerosols, including smoke, and small-particle pollution such as soot, dust, and sulphate particles.

If we were to stop burning fossil fuels tomorrow, the aerosols that cool our planet would be rained out of the air in about ten days. Without these aerosols, which mask roughly half the heating effect caused by carbon dioxide, there would be a sudden jump in temperature. Stopping all carbon dioxide emissions could produce a short-term warming of one-half to one degree. Cutting black carbon-soot emissions would offset some of this effect.

Removing aerosols causes steeper warming the more quickly that fossil-fuel combustion is cut. If fossil-fuel combustion were to be cut to zero in two decades then, a.s.suming a mid-range climate sensitivity, the loss of the related aerosols plus the warming already in the system could produce a warming of more than one degree in twenty years. This would be highly destructive to our ecosystems.

Reducing fossil-fuel combustion to zero in 50 years will also produce a rate of warming far beyond the capacity of most ecosystems to cope, because of the aerosol cooling lost. Cutting fossil-fuel combustion much more slowly, to zero in a hundred years, would have the same effect because, particularly in the latter part of the timeline, more carbon would be kicking into the atmosphere from failing natural carbon sinks, exacerbating the long-term trend.

Slowing the rate of reduction of fossil-fuel combustion may then make the warming problem from aerosol reduction less severe in the short term, but worse in the long-term.

These 'd.a.m.ned if you do, and d.a.m.ned if you don't' problems are known in the fields of science, politics, and economics as 'wicked problems', a concept first articulated by design and planning theorist Horst Rittel. A 'wicked problem' describes a complex set of interrelated and circular problems which are resistant to resolution and where any solution is not good or bad, but only better or worse. Each 'wicked problem' is unique and, effectively, offers only one chance to achieve the least-worst resolution, because poorly constructed 'solutions' can compound the problem and, in many cases, there is a limited time horizon for effective action.

Getting to the safe-climate zone will take time. But, as each year slips by, the impact of warming and the problem of positive feedbacks takes us further away from that zone.