Never mind the weather, climate change could tear apart the very fabric of the Earth, says Bill McGuire

WITHIN days of the Indian Ocean tsunami in December 2004, the internet was alive with theories about how climate change was to blame for the disaster. Some of the explanations were far-fetched to say the least. One I particularly liked proposed that, because of global warming, "magma in the Earth's core [sic] is heating up, raising the Earth's temperature and causing eruptions and earthquakes."

This is clearly nonsense, but the idea that climate change is linked to extreme geological events is not as far-fetched as it might sound. All over the world evidence is stacking up that changes in global climate can and do affect the frequencies of earthquakes, volcanic eruptions and catastrophic sea-floor landslides. Not only has this happened several times throughout Earth's history, the evidence suggests that it is starting to happen again. While no serious scientist is suggesting that the Sumatran earthquake was triggered by global warming, there is a growing consensus that if climate change continues unchecked, we can expect not only a warmer future, but a more geologically turbulent one too.

The climate interacts with the Earth's crust via the changing mass of water and ice that is shifted around the planet. The pressure of water and ice on the crust is considerable: 1 cubic metre of water weighs 1 tonne, while the same volume of ice weighs slightly less, up to 0.9 tonnes. With this in mind, it shouldn't come as a surprise that the loading and unloading of the Earth's crust by ice or water can trigger seismic and volcanic activity and even "landslides. Dumping the weight of a kilometre-thick ice sheet onto a continent or removing a deep column of water from the ocean floor will inevitably affect the stresses and strains on the underlying rock.

Although these forces on the Earth's crust are subtly changing all the time, their effects are most obvious at times of major or sudden climate change, such as at the beginning and end of an ice age or during the period of climate change we are expected to experience over the coming centuries. As the balance changes between the stresses acting on the crust and the strains held within it, the result can be an increase in volcanic eruptions and earthquakes.

The Earth has seen this pattern many times before. In the past 650,000 years alone, the polar ice caps have expanded far beyond their current limits on seven occasions, locking up huge volumes of water in frozen oceans and vast continental ice sheets before retreating again to higher latitudes. These huge reorganisations of the Earth's water resulted in dramatic and repeated swings in sea level, with falls as far as 130 metres below today's level followed by equally spectacular rises. They also led to shifting loads on volcanoes and geological faults. As ice sheets that had pinned down volcanoes and active faults melted away, the Earth's crust bounced back in a process known as isostatic rebound. As it did so, faults were reactivated and seismic activity increased sharply.

In the early 1970s John Chappell of the Australian National University in Canberra was the first to make the link between glacial advances and retreats and the rate of global volcanism. We now know that the warming that heralded the start of the current interglacial period around 10,000 years ago brought forth a burst of volcanic activity in Iceland, as melting ice caps reduced pressures on the magma chambers below. Allen Glazner of the University of North Carolina at Chapel Hill identified a similar pattern in eastern California over the past 800,000 years. Increased levels of volcanic activity are also recorded at mid-latitude ice-covered volcanoes in the Cascades Range of the US and in the Andes.

Earthquake activity shows a similar pattern. Nils-Axel Morner of Stockholm University in Sweden first noted that this isostatic rebound triggered earthquakes in post-ice-age Scandinavia, and the effect has since been noted in Scotland and North America. Patrick Wu of the University of Calgary in Alberta, Canada, and Paul Johnston of the University of Western Australia in Perth have even gone so far as to suggest that the effect may still be playing a role today. In particular, they have speculated that the continuing rebound of the North American continent may have contributed to the great New Madrid earthquakes that shook the central Mississippi valley in 1811 and 1812.

Yet while we may still be feeling the effects of the last ice age, the impact of today's warming trend might already be making itself felt. In 2004 NASA geophysicist Jeanne Sauber and geologist Bruce Molnia of the US Geological Survey linked unloading of the crust as a result of the rapid glacial melting in south-west Alaska to a magnitude 7.2 earthquake in 1979, and warned that more could be on the way. "In areas like Alaska, where earthquakes occur and glaciers are changing, their relationship must be considered to better assess earthquake hazard," says Sauber. This has implications for all those parts of the world where glaciers and active faults coincide, including the Alps, Himalayas, Rocky Mountains, Andes and the Southern Alps in New Zealand.

Of particular concern is the continental shelf around Greenland. Here, the unloading and uplift that would follow catastrophic melting of the ice sheet might trigger earthquakes strong enough to dislodge the huge piles of sediment that have accumulated around the edges of the land. The resulting underwater landslides could generate tsunamis on a scale comparable to those that followed the Storegga slide 8000 years ago off the west coast of Norway. The Storegga collapse is thought to have been caused by an underwater earthquake that led to three huge sediment slips. The result was a tsunami more than 20 metres high in the Shetland Isles off the north coast of Scotland and up to 6 metres high along the east coast of the Scottish mainland. This region is now stable, but similar piles of sediment near Greenland are ripe for collapse.

Changes in water distribution don't have to be as dramatic as a tsunami to trigger geological activity. The rising sea levels that followed glacial retreat at the end of the last ice age also spawned a burst of volcanic activity in areas far removed from the direct effects of ice unloading. In 1997 my own research group at University College London used data from sediment cores taken from the sea floor of the Mediterranean to show that the intensity of volcanic activity there has been directly linked to the rate of sea-level change over the past 80,000 years. The most pronounced effect was over the past 15,000 years or so, a period of rapid sea-level rise. It doesn't seem to be a localised effect. Greg Zielinski of the University of New Hampshire and colleagues have picked up evidence for bursts of volcanic activity from the same time in ice cores taken from deep within the Greenland ice sheet.

More to come

Melting ice and sea-level rise also mean that previously exposed continental margins become inundated with water. At the end of the last ice age, the extra load was more than enough to reactivate faults and trigger earthquakes around the rims of all the major ocean basins, some of which are thought to have set off giant landslides on the sea floor. Evidence of 27 such slides has been uncovered in the North Atlantic basin, and many of these were indeed triggered by rising sea levels over the past 15,000 years.

So could all this happen again? Sauber and Molnia's studies suggest that it has already begun. Ultimately the amount of destruction we see will depend on the scale of the environmental changes that global warming brings and exactly how sensitive our planet's crust is to these. There are, however, worrying signs that the answer to the former is "big" and to the latter, "very".

The recently recognised acceleration in melting of the Greenland ice sheet and the potential collapse of the West Antarctic ice sheet, which Chris Rapley of the British Antarctic Survey refers to as "an awakening giant", mean it is possible that sea levels will rise by several metres over the next few centuries. This would match the fastest rates of the immediate post-glacial period, and may be the recipe for a fiery and shaky future, as well as a warm one.

Even a more conservative estimate, such as that offered by the Intergovernmental Panel on Climate Change in its 2001 report, could be enough to make the difference. The IPCC predicted a rise in sea level of up to 88 centimetres by the end of the century - not quite the 130 metres seen 10,000 years ago, but still enough to add a huge load of water onto an unsuspecting volcano or fault line. Far smaller volumes of water have been known to trigger earthquakes in the recent past. The filling of the Vajont reservoir in north-east Italy has been charged with triggering earthquakes that contributed to the catastrophic collapse of the adjacent Mount Toc into the reservoir in 1963. This created a gigantic wall of water that overtopped the dam and obliterated the town of Longarone, killing more than 2000 people. Four years later, a large earthquake that killed 180 people in the Indian state of Maharashtra was blamed on the filling of the reservoir behind the Koyna dam, while today, concerns about the potential for China's Three Gorges reservoir to trigger earthquakes has led the government to establish a dedicated monitoring system.

As for volcanoes, 57 per cent of the world's 600 or so active volcanoes form islands or occupy coastal sites, while 38 per cent are located within 250 kilometres of the coast. Many, or all, could be susceptible to the stress changes in the crust that will accompany sea-level rise if wholesale melting of the great polar ice sheets really gets going. Evidence from the end of the last ice age suggests that an increase in continental-margin earthquakes must also be expected, making it more likely that sea-floor landslides will be triggered in areas that are unstable, including the east coast of the US, the northern Caribbean and offshore California.

A particular worry is that this in turn will contribute to large-scale releases of methane gas from the solid gas hydrate deposits that are trapped in marine sediments. Gas hydrates have been identified around the margins of all the ocean basins, and outbursts of gas may occur as sea temperatures climb or as rising sea levels trigger underwater quakes in the vicinity.

Being a highly effective greenhouse gas, these additional methane "burps" could bump up rising global temperatures even further, although the hope is that the extra load of higher sea levels might actually help to keep the gas hydrates solid and stable.

However bad it gets, climate change is unlikely to result in anything other than a small fraction of the ice-sheet melting and consequent sea-level rise that followed the end of the last ice age. However, the rate of change we are expecting means that the rate of melting and sea-level rise in the next few centuries will be as great as anything since the glaciers retreated to the poles.

Not every volcanic eruption and earthquake in the years to come will have a climate-change link, whatever you might read on the web. Yet as the century progresses we should not be surprised by more geological disasters as a direct and indirect result of dramatic changes to our environment. The only saving grace is that a significant increase in volcanic activity would pump large volumes of sulphate gases into the stratosphere, which would cool the Earth's surface and slow global warming, at least for a time. It's a hell of a price to pay, though, for ignoring a phenomenon that could be far more easily sorted if we lived more considered and sustainable lives.

[Sidebar]

Pavlof volcano in Alaska is thought to be triggered by seasonal changes in the weather

[Sidebar]

"The amount of destruction will depend on the scale of changes that global warming brings"

[Sidebar]

FEELING THE HEAT

How the Earth's crust will be affected by climate change

VOLCANOES

Eruptions of the Pavlof volcano in the Aleutian Islands, Alaska, are thought to be triggered by low-pressure weather systems which increase the weight of water on the area surrounding the volcano. Since 57 per cent of all volcanoes are located in islands or on the coast, it is possible that changes in sea level or the weather will trigger future eruptions worldwide

SEA-FLOOR LANDSLIDES

As the Greenland ice cap melts, the release of pressure on the Earth's crust could lead to earthquakes. These in turn might trigger a massive undersea landslide and a mega-tsunami similar to that which caused the Storegga slide oil the coast of Norway 8000 years ago

GAS HYDRATES

The effects of global warming and a rise in sea levels on gas hydrates are unclear, while rising sea temperatures might destabilise gas hydrates, a rise in sea level would increase the pressure on the deposits, and hopefully stabilise them

EARTHQUAKES

A dramatic use in sea levels would place plate boundaries under greater pressure and might trigger catastrophic earthquakes in densely populated coastal areas

[Sidebar]

Storm clouds gather

In the Bering Straits, there is a volcano called Pavlof, whose eruptions are controlled by the weather.

Most of Pavlof's eruptions occur in the autumn and winter months. Steve McNutt of the Alaska Volcano Observatory thinks that is because eruptions are triggered by lowpressure weather systems at that time of the year. According to McNutt, a fall in pressure raises the water level around Pavlof by 30 centimetres, and storm winds may pile the water even higher. He speculates that this increased loading on the volcano may squeeze the magma upwards "like toothpaste out of a tube", leading to eruptions so regular that you can set, if not your clock, at least your calendar by them.

Pavlof is not the only volcano to be so sensitive to the environment. Research by Ben Mason and colleagues at the University of Cambridge suggests that most, if not all, volcanoes are highly sensitive to changes in their surroundings. By examining the timings of more than 3000 eruptions between 1700 and 1999, the researchers revealed a broad seasonal trend in volcanic activity across the planet, with more eruptions occurring between November and March than in the rest of the year.

Once again, it looks as if the oceans are to blame, this time as a result of a seasonal shifting of trillions of tonnes of water every year between the hemispheres and from the ocean basins onto the margins of the continents. These enormous redistributions of mass seem to be sufficient to deform the Earth's crust and stress volcanoes in such a way as to encourage eruptions.

Some earthquake-generating faults could also be sensitive enough to respond to changes in the weather. In circumstances analogous to those at Pavlof volcano, seismologists Selwyn Sacks and Alan Linde of Washington DC's Carnegie Institution have been able to link a certain type of "slow earthquake" in eastern Taiwan to the passage of intense low-pressure weather systems - in this case, typhoons.

It looks, then, as if many potentially hazardous geological systems are sensitive to short-term changes in currents, sea level or atmospheric pressure. With a doubling in the number of more powerful tropical cyclones in the last 30 years attributed by some to climate change, and perhaps worse on the way, it may be that a stormier future will have an extra sting in the tail.

[Author Affiliation]

Bill McGuire is professor of geological hazards at University College London, UK, and director of the university's Benfield Hazard Research Centre. His most recent book is Surviving Armageddon: Solutions for a threatened planet (Oxford University Press)