Can geoengineering avert climate chaos?

• New Scientist 24 September 2012 by Stephen Battersby
• Magazine issue 2883.

http://www.newscientist.com/article/mg21528831.700-can-geoengineering-avert-climate-chaos.html?full=true

From sunshades to algal blooms, there are plenty of ideas for cooling the planet. But are there hidden dangers?

See more in our gallery: "The pros and cons of geoengineering"

OOPS. We really didn't mean to, but we seem to have broken the planet. Is there anything we can do to make it better?

Climate change is already upon us, melting ice, killing forests and making floods and heatwaves more intense. Meanwhile, global emissions of carbon dioxide and other greenhouse gases continue to increase, promising far worse to come. Even if we stopped all emissions tomorrow, temperatures would keep rising for decades, with potentially catastrophic consequences ranging from famines to rapid sea-level rise.

So perhaps it is time to get serious about the audacious idea of geoengineering. The hope is that by deliberately tinkering with our planet's climate machine, we might be able to fix our gargantuan blunder, or at least avoid some of the most serious consequences, or just buy ourselves a bit more time to cut emissions.

Dozens of schemes have been devised to cool the planet. We could launch a vast fleet of ships to whiten the clouds by spraying salt mist, or squirt sulphuric acid into the stratosphere to reflect the sun. Send a swarm of mirrors into deep space. Engineer paler crops. Fertilise the oceans. Cover the world's deserts in shiny mylar. Spread cloud-seeding bacteria. Release a global flock of microballoons.

These schemes are ingenious, but would any of them work? Or would they just make things worse and hasten catastrophe? Short of taking the biggest gamble imaginable and actually trying one out, the best that we can do is try to explore each idea with detailed calculations and computer models. As the results of such studies mount up, we're starting to get an idea of what geoengineering might - or might not - be able to achieve.

Some ideas can be dismissed with relative ease. Covering deserts in reflective plastic, for example, could reflect a lot of sunlight and cool the planet somewhat, but it probably is as crazy as it sounds. It would devastate ecosystems, alter regional climate patterns and require an immense army of cleaners to keep it going.

Others are beyond our powers today. To shade Earth with a swarm of space parasols would require an estimated 20 million rocket launches. Without some radical new technology, that would be astronomically expensive and fatally polluting. "This is complete science fiction," says Tim Lenton of the University of Exeter, UK. "We ought to stop talking about it."

Many other schemes, such as painting roofs white, are certainly feasible - but can they actually fix the climate? The basic problem, of course, is that rising levels of greenhouse gases in the atmosphere are acting like a blanket around Earth, trapping heat. Some time this century we are likely to have doubled the concentration of CO₂ in the atmosphere, reducing heat loss by about 3.7 watts per square metre, averaged across the planet. To stop Earth warming, any geoengineering scheme either has to block as much incoming heat from the sun or increase heat loss from the top of the atmosphere by as much.

We have other prerequisites for our global refrigerator (see diagram). It needs to work without drastically altering regional climates, while also preventing sea level from rising. Ideally we want to stop the oceans becoming so acidic that coral reefs vanish, too.

But the first test is potency. In 2008, Lenton and Nem Vaughan of the University of East Anglia in Norwich, UK, combined various model results with their own calculations to assess the potential cooling power of a couple of dozen proposals. "It was born of frustration," says Lenton. "I had been at one too many workshops where people were advocating their pet technologies and arm-waving about 'was this more effective than that'."

They found that many schemes would make little difference. Take the idea of making roofs and roads whiter to reflect more sunlight. Even with optimistic assumptions, this could only reflect about 0.15 watts per square metre - at best a minor contribution to restoring Earth's heat balance.

A seemingly more promising plan is to fertilise the seas. Plankton consume CO₂ as they grow, and sometimes their dead bodies sink to the sea floor and get buried, locking this carbon away. Adding nutrients that are in short supply, such as iron, could boost plankton growth. By the end of the century, this could improve the radiation balance by as much as 0.2 watts per square metre, Lenton and Vaughan calculated. Handy, but not a game-changer - and again that's the top-end estimate, which could fall considerably as we learn more about this process.

Many of the other proposals, such as encouraging downwelling in polar regions to speed up the transport of carbon into the ocean depths, are even more limited. But two schemes stand out as being both highly potent and relatively feasible. Both involve some form of sunshade.

One idea is to whiten marine clouds - specifically the low, flat stratus clouds that cover a large swathe of sky. Ships scattered across the world's oceans would send plumes of fine salt spray up into the air. By acting as nucleation sites, the salt particles should encourage droplets of water to form in clouds. With more droplets per cubic metre, these clouds would be whiter than normal, and reflect more sunlight. Potentially, this could offset the entire warming from a doubling in CO₂.

Cloud-whitening has its upsides, such as not involving any hazardous chemicals. But cloud nucleation is not well understood, so it might not work as well as its proponents suggest, and cooling only the oceans could disrupt local climate. A study published this year found that seeding clouds over the Pacific might alter rainfall patterns in a similar way to the highly disruptive La Niñaweather phenomenon, for instance.

The other leading contender is an old one: fill the atmosphere with a haze of fine particles. In fact, we are doing this already. Sulphur dioxide pollution forms fine droplets of sulphuric acid that already reflect an estimated 0.4 watts per square metre [http://www.tellusb.net/index.php/tellusb/article/view/16899]. But SO₂ from fires and factories doesn't remain in the atmosphere for long, so its effects are limited. If sulphate gets as high as the stratosphere, however, it can linger for years, so its cooling effect is much greater. The proof comes from volcanic eruptions large enough to inject SO₂ into the stratosphere. The 1991 eruption of Mount Pinatubo in the Philippines cooled the planet by up to 0.5 °C over the following couple of years.

Bargain price

To balance the warming effect of a doubling in CO₂, we would need to pump up to 5 million tonnes a year of SO₂ into the stratosphere. According to Justin McClellan of Aurora Flight Sciences in Cambridge, Massachusetts, whose team evaluated several ways to deliver the sulphates, this would cost about $10 billion per year. Compared with the stupendous costs and consequences of global warming, this is an absolute bargain. Sea level rise alone will swallow up many trillions of dollars' worth of cities and farmland.

Unfortunately, our sulphur spray may barely slow the seas' advance. Sulphur droplets do not linger in polar regions as long as they do in the tropics, making them less effective polar coolants. So even if aerosol injection brought the average global temperature back down to that of the 1800s, the poles would not be as cold as they were and the ice caps would keep melting. This might not be enough to avert catastrophes such as the collapse of the West Antarctic ice sheet, which would raise sea level more than 3 metres.

It is not clear whether a different kind of reflector, such as solid metallic particles or tiny, shiny balloons, would be any better. Pumping out a gas is so much simpler and cheaper, so most studies have concentrated on sulphates.

While coastal plains and cities drown, the rest of the planet might dry out. With any kind of sunshade, less sunlight will reach the sea surface, reducing evaporation. So far, the effect of sulphur pollution has been outweighed by warming, which increases evaporation. But if we reduced the temperature to the preindustrial level this way, there would be a dramatic decline in rainfall. That might be avoided by not reducing the temperature as much - but then the ice sheets would melt faster.

Sunshades could also have disastrous regional effects, according to climate models. If they disrupted the monsoons, they could bring permanent famine to billions. "Or say you changed the circulation patterns that feed moisture to the Amazon rainforest," says Tim Palmer of the University of Oxford. "You might turn the Amazon to desert."

In 2010, Myles Allen of the University of Oxford and his colleagues looked at the effect of varying amounts of sunscreen in the stratosphere using a detailed climate model. They found that there is no solution that works for everyone. An amount of aerosol that would take China close to comfortable preindustrial temperature and rainfall might cool India far too much.

Or it could be the other way around. Climate models agree fairly well on the global effects of sunshade schemes, but produce different patterns of regional climate change.

This may be because of the different assumptions and values used in different studies. Or it may be due to the limitations of existing climate models. As they improve, their regional projections may start to agree with each other, which would give us some degree of confidence in them.

Some of the factors affecting regional climates are inherently unpredictable, though. How much of the rainforests will be left standing in 100 years' time? How much will emissions fall, if at all? How will ecosystems respond? As a result, we can never be 100-per-cent certain that any particular scheme will have the desired result.

This makes any sunshade highly risky. If it turned out to have some terrible consequence and we suddenly stopped replenishing sulphates or whitening clouds, the planet would warm very rapidly over the next few years. Such a sudden transition would be even more damaging than a gradual warming to the same level, giving no time for people and wildlife to adapt. "You are upping the stakes," says Lenton. And if we reach for the sulphates, we might need another type of geoengineering, such as cirrus seeding (see "You cannot be cirrus") to cool the poles, prescribing not just one but two dangerous drugs for the planet.

So instead of blocking sunlight, maybe we should get at the actual cause of the problem and actively scrub CO₂ from the air. The concentrated gas could then be pumped into underground reservoirs such as depleted gas and oil fields. But no one has devised an efficient method for doing this. "The problem is you're trying to capture a very dilute gas, which is inherently costly compared with capture from a concentrated source like a power station," says Lenton.

With existing technology, there is no realistic prospect of mopping up all the extra CO₂ we are adding to the atmosphere in time to prevent further climate change. Even an industrial effort on a vast scale could take centuries, and the longer CO₂ emissions keep rising, the greater the challenge will be.

Instead of covering the planet in carbon-eating machinery, how about speeding up the reaction of CO₂ with silicate rocks? Over millions of years, this process,called weathering, soaks up vast amounts of CO₂, which is eventually returned to the atmosphere by volcanoes. But to deal with just a single year's worth of emissions, we'd need to grind up at least 7 cubic kilometres of rock and spread it so thinly that it would cover several per cent of Earth's land surface. So this process cannot save us either.

What about modifying land use and agriculture to capture more carbon? Simply planting forests remains a good thing, although geography limits its potential to about 0.5 watts, and all that carbon could end up back in the atmosphere if forests die or burn as the planet warms.

Locking away carbon

One way to lock away the carbon stored by plants is to turn them into charcoal - biochar - and bury it. Another is to burn crops in power plants fitted with carbon-capture technology. These ideas need land, so they will compete with food production. This year, Lenton calculated that the total benefit could be a useful 0.3 watts by 2050 - but only if we increase farming efficiency and eat less land-hungry, methane-belching meat. At present, meat consumption is rising while crop production is already being hit by extreme weather and water shortages, so this looks optimistic barring some breakthrough, such as genetically altering plants to enable them to capture more of the sun's energy.

Carbon-capture schemes, then, can at best slow the pace of warming over the coming century. If they are implemented as alternatives to cutting emissions - for instance, to earn carbon credits that can be sold to those who want to emit CO₂ - they won't achieve even this.

They will also be of no use if we are nearing a tipping point such as the widespread dying of forests, the massive release of methane from thawing permafrost or the collapse of the West Antarctic ice sheet. So perhaps we should keep the potent but risky schemes such as sulphur sprays in reserve for the direst circumstances? Perhaps. But Lenton, who helped to define the notion of tipping points in a paper in 2008, is sceptical. "People say that is why we need solar reflection in our back pocket, but they haven't proved you could get early warning of a tipping point, or deploy in time, or that these schemes would not cause other tipping points," he says.

If we wait until the last possible moment, then, it could be too late to avert climate chaos. "You shouldn't think of this as a magic button that you can press if things get out of control - it may turn out to be a bit of a nightmare," Palmer says. And even if we did go for the nuclear option of a sunshade scheme, almost all climate scientists agree we would still need to make aggressive cuts in emissions.

There are a few dissenters. Peter Cox at the University of Exeter points out that higher CO₂ boosts the growth of some kinds of plants and reduces water loss, as plants don't have to keep their pores open as long. So if you could have higher CO₂ without the droughts, floods, storms and growth-impeding heat that global warming will bring, then food production would increase. Maybe we could achieve that with sunshields. "In terms of the things we care about most, it might be a better option than conventional mitigation," says Cox. Such a cool-but-carbonated future carries frightening risks, though, and Cox is only suggesting we consider the notion.

In the end, the greatest obstacle to any drastic form of geoengineering may turn out to be politics. "You can't have competing geoengineering programmes, there has to be just one," says Allen. "So some supranational body would have to decide on the weather."

Achieving agreement may be almost impossible, because different countries will have different priorities. Some are most threatened by sea level rise, others by sheer heat or shifting rainfall. And if the Kyoto protocol is any guide, if any agreement is eventually reached it might be a far cry from what's actually needed.

However, international agreement will be needed only for big sunshield schemes, with their global dangers. There is nothing to stop individuals, institutions or countries going it alone with a bit of biochar or some other carbon-capture scheme. It may seem mundane compared with shiny space mirrors, but for now perhaps the safest tools for engineering the planet are to be found down on the farm.

You cannot be cirrus

The high, wispy cirrus clouds that sometimes grace an otherwise blue summer sky may seem an unlikely enemy, but David Mitchell is making plans to attack them. Destroying cirrus might not only reduce global temperature but also help save the ice caps and curb extreme weather.

Clouds have complex effects on Earth's heat budget, reflecting some incoming sunlight and trapping a lot of outgoing infrared radiation. Lower-altitude clouds such as marine stratus also radiate a lot of heat from their tops out into space, so overall they cool the planet. Icy cirrus clouds radiate much less heat, so their net effect is to warm us up.

In 2009, Mitchell - based at the Desert Research Institute in Reno, Nevada - suggested that we could use aircraft to spread bismuth triiodide, a non-toxic compound that should seed relatively large ice crystals. These would fall from the sky faster than natural cirrus ice, so the clouds would disperse.

Preliminary attempts to model the process, which Mitchell presented at a meeting in July, indicated that this could cool the planet by about 2 watts per square metre - enough to prevent half of the warming from a doubling of CO₂.

Better still, the method ought to work best where it is most needed, at high latitudes. Concentrating efforts here could protect our fragile ice caps. It would also help to restore the temperature difference between tropic and pole. That difference has been eroded by the rapid warming in the Arctic, which is thought to be one reason why we are seeing more extremes of weather.

The modelling is at a very early stage, Mitchell cautions. "Lots of research needs to be done on representing cirrus in global climate models - and not just for geoengineering." He would like to see a cloud-seeding experiment in a small area to see what really happens.

What's more, dispersing cirrus shares many of the risks of sunshade schemes (see main story): it may well have disastrous regional effects, and stopping it abruptly would be dangerous.

Stephen Battersby is a consultant for New Scientist based in London