The Science of Discworld I

28. THE ICEBERG COMETH

THE EARTH HAS BEEN A GIANT SNOWBALL on many occasions. It was a snowball 2.7 billion years ago, 2.2 billion years ago, and 2 billion years ago. It was a really cold snowball 800 million years ago, and this was followed by a series of global cold snaps that lasted until 600 million years ago. It reverted to snowball mode 300 million years ago, and has been that way on and off for most of the last 50 million years. Ice has played a significant part in the story of life. Just how significant a part, we are now beginning to appreci­ate.

We first began to realize this when we found evidence of the most recent snowball. About one and a half million years ago, round about the time that humans began to become the dominant species on Earth, the planet got very cold. The old name for this period was the Ice Age. We don't call it that any more because it wasn't one Age: we talk of 'glacial-interglacial cycles'. Is there a connection? Did the cold climate drive the naked ape to evolve enough intelligence to kill other animals and use their fur to keep warm? To discover and use fire?

This used to be a popular theory. It's possible. Probably not, though: there are too many holes in the logic. But a much earlier, and much more severe, Ice Age very nearly put a stop to the whole of that 'life' nonsense. And, ironically, its failure to do so may have unleashed the full diversity of life as we now know it.

Thanks to the pioneering insights of Louis Agassiz, Victorian sci­entists knew that the Earth had once been a lot colder than it is now, because they could see the evidence all around them, in the form of the shapes of valleys. In many parts of the world today you can find glaciers, huge 'rivers' of ice, which flow, very slowly, under the pressure of new ice forming further uphill. Glaciers carry large quantities of rock, and they gouge and grind their way along, form­ing valleys whose cross-section is shaped like a smooth U. All over Europe, indeed over much of the world, there are identical valleys, but no sign of ice for hundreds or thousands of miles. The Victorian geologists pieced together a picture that was a bit worry­ing in some ways, but reassuring overall. About 1.6 million years back, at the start of the Pleistocene era, the Earth suddenly became colder. The ice caps at the poles advanced, thanks to a rapid build­up of snow, and gouged out those U-shaped valleys. Then the ice retreated again. Four times in all, it was thought, the ice had advanced and retreated, with much of Europe being buried under a layer of ice several miles thick.

Still, there was no need to worry, the geologists said. We seemed to be safe and snug in the middle of a warm period, with no prospect of being buried under miles of ice for quite some time ...

The picture is no longer so comfortable. Indeed, some people think that the greatest threat to humanity is not global warming, but an incipient ice age. How ironic, and how undeserved, if our pollution of the planet cancels out a natural disaster!

As usual, the main reason we now know a lot more is that new kinds of observation became possible, propped up by new theories to explain what it is that they measure and why we can be reason­ably sure that they do. These new methods range from clever methods for dating old rocks to studies of the proportions of dif­ferent isotopes in cores drilled from ancient ice, backed up by ocean-drilling to study the layers of sediment deposited on the sea floor. Warm seas sustain different living creatures, whose death deposits different sediment, so there is a link from sediments to cli­mate.

All of these methods reinforce each other, and lead to very much the same picture. Every so often the Earth begins to cool, becom­ing 10°-15°C colder near the poles and 5°C colder elsewhere. Then it suddenly warms up, possibly becoming 5°C warmer than the cur­rent norm. In between big fluctuations, there are smaller ones: 'mini ice ages'. The typical gap between a decent-sized ice age and the next is around 75,000 years, often less, nothing like the com­fortable 400,000 years of 'interglacial' expected by the Victorians. The most worrying finding of all is that periods of high tempera­tures, that is, like we get now, seldom lasted more than 20,000 years.

The last major glaciation ended 18,000 years ago.

Wrap up well, folks.

What caused the ice ages? It turns out that the Earth isn't quite as nice a planet as we like to think, and its orbit round the sun isn't quite as stable and repetitive as we usually assume. The currently accepted theory was devised in 1920 by a Serbian called Milutin Milankovitch. In broad terms, the Earth goes round the sun in an ellipse, almost a circle, but there are three features of the Earth's motion that change. One is the amount through which the Earth's axis tilts, about 23° at the moment, but varying slightly in a cycle that lasts roughly 41,000 years. Another is a change in the position of Earth's closest approach to the sun, which varies in a 20,000-year cycle. The third is a variation in the eccentricity of the Earth's orbit, how oval it is, whose period is around 100,000 years. Putting all three cycles together, it is possible to calculate the changes in heat received from the sun. These calculations agree with the known variations in the Earth's temperature, and it seems particularly likely that the Earth's warming up after ice ages is due to increased warmth from the sun, thanks to these three astronomical cycles.

It may seem unsurprising that when the Earth receives more heat from the sun, it warms up, and when it doesn't, it cools down, but not all of the heat that reaches the upper atmosphere gets down to the ground. It can be reflected by clouds, and even if it gets to ground level it can be reflected from the oceans and from snow and ice. It is thought that during ice ages, this reflection causes the Earth to lose more heat than it would otherwise do, so ice ages auto­matically make themselves morse. We get kicked out of them when the incoming heat from the sun is so great that the ice starts to melt despite the lost heat. Or maybe the ice gets dirty, or ... It's not so clear that we get kicked into an ice age when less of the sun's warmth reaches the Earth, indeed the slide into an ice age is usu­ally more gradual than the climb back out of it.

All of which makes one wonder whether global warming caused by gases excreted from animals might be partly responsible. When gases such as carbon dioxide and methane build up in the atmos­phere, they cause the famous 'greenhouse effect', trapping more sunlight than usual, hence more heat. Right now, most scientists have become convinced, the Earth's supply of 'greenhouse gases' is growing faster than it would otherwise do thanks to human activi­ties such as farming (burning rainforests to clear land), driving cars, burning coal and oil for electricity, and farming again (cows pro­duce a lot of methane: grass goes in one end and methane emerges at the other). And how could we forget the carbon dioxide breathed out by people? One person is equivalent to half a car, maybe more.

Maybe in the past there were vast civilizations of which we now know nothing, all traces having vanished, except for their effect on the global temperature. Maybe the Earth seethed with vast herds of cattle, buffalo, elephants busily excreting methane. But most scien­tists think that climate change results from variations in five different factors: the sun's output of radiant heat, the Earth's orbit, the composition of the atmosphere, the amount of dust produced by volcanoes, and levels of land and oceans resulting from move­ment of the Earth's crust. We can't yet put together a really coherent picture in which the measurements match the theory as closely as we'd like, but one thing that is becoming clear is that the Earth's climate has more than one 'equilibrium' state. It stays in or near one such state for a while, then switches comparatively rapidly to another, and so on.

The original idea was that one state was a warm climate, like the one we have now, and the other was a cold 'ice age' one. In 1998 Didier Paillard refined this idea to a three-state model: interglacial (warm), mild glacial (coldish), and glacial (very cold). A drop in heat received from the sun below some critical threshold, caused by those astronomical cycles, triggers a switch from warm to coldish. When the resulting ice builds up sufficiently, it reflects so much of the sun's heat that this triggers another switch from coldish to very cold. But when the sun's heat finally builds up again to another threshold value, thanks once more to the three astronomical cycles, then the climate switches back to warm. This model fits observa­tions deduced from the amount of oxygen-18 (a radioactive isotope of oxygen) in geological deposits.

Finally, some drama. About 800 million years ago there was an ice age so severe that it very nearly killed off all of the surface life on Earth. This 'big freeze' lasted for between 10 and 20 million years, the ice reached the equator, and it seems that the seas froze to a depth of half a mile (1 km) or more. According to the 'snowball Earth' theory, ice covered the entire Earth at this time. However, if ice really covered the whole Earth, it should have done more dam­age than the fossil record indicates. So maybe the Earth's axis tilted a lot more than astronomers are willing to concede, and the poles lost their ice while equatorial regions gained it. Or perhaps conti­nental drift was more rapid at that time than we think, and we've mapped out the extent of the ice incorrectly. Whatever the details, though, it was a spectacularly icy world.

Although the big freeze came close to wiping out ail surface life, it may indirectly have created a lot of today's biodiversity. The big shift from single-celled creatures to multi-celled ones also hap­pened 800 million years ago. It is plausible that the big freeze cleared away a lot of the single-celled lifeforms and opened up new possibilities for multi-celled life, culminating in the Cambrian Explosion 540 million years ago. Mass extinctions are typically suc­ceeded by sudden bursts of diversity, in which life reverts from being a 'professional' at the evolutionary game to being an 'ama­teur'. It then takes a while for the less able amateurs to be eliminated, and until they are, all sorts of strange strategies for making a living can temporarily thrive. The succession of icy peri­ods that followed the big freeze could only have assisted this process.

However, it may have been the other way round. The invention of the anus by triplobiasts may have changed the ecology of the seas. Faeces would have dropped to the sea-bed, where bacteria could specialize in breaking them down. Other organisms could then become filter feeders, living on those bacteria, perhaps sending their larvae up into the plankton for dispersal, as modern filter-feeders do. Several new ways of life depended on this primeval composting system. And it's possible that the successful return of phosphorus and nitrogen into the marine cycles led to an explosion of algae, which reduced atmospheric carbon dioxide, cut back on the greenhouse effect, and triggered the big freeze.

Fortunately for us, the big freeze wasn't quite long enough, or cold enough, to kill off everything. (Bacteria in volcanic vents on the ocean floor and in the Earth's crust would have survived no matter what, but evolution would have been set back a long, long way.) So when the Earth warmed, life exploded into a fresh, com­petition-free world. Paradoxically, a major reason why we are here today may be that we very nearly weren't. Our entire evolutionary history is full of these good news-bad news scenarios, where life leaps forward joyously over the bodies of the fallen ...

Rincewind can be forgiven for feeling that Roundworld has it in for him. Life has suffered from many different kinds of natural dis­aster. Here are two more. In the Permian/Triassic extinction of 250 million years ago, 96% of all species died within the space of a few hundred thousand years[40]. William Hobster and Mordeckai Magaritz think this happened because they suffocated. Carbon iso­topes show that a lot of coal and shale oxidized in the run-up to the extinction, probably because of a fall in sea level, which exposed more land. The result was a lot more carbon dioxide and a lot less oxygen, which was reduced to half today's level. Land species were especially badly affected.

Another global extinction, though less severe, occurred 55 mil­lion years ago: the Palaeocene/Eocene boundary. In cores of sediment drilled from the Antarctic, James Kennett and Lowell Stott discovered evidence of the sudden death of a lot of marine species. It seemed that trillions of tons (tonnes) of methane had burst from the ocean, sending temperatures through the roof, methane being a powerful greenhouse gas. Jenny Dickens suggested that the methane was released from deposits of methane hydrates in permafrost and on the seabed. Methane hydrates are a crystal lat­tice of water enclosing methane gas: they are created when bacteria in mud release the gas and it becomes trapped.

Coincidentally, one of the main results of the Palaeocene/Eocene extinction was a burst of evolutionary diversity, leading in particular to the higher primates, and us. Whether something is a disaster depends on your point of view. Rocks may not have a point of view, as Ponder Stibbons pointed out, but we certainly do.