Snowballs and Greenhouses
Imagine a time when carbon dioxide levels in the atmosphere have doubled. Methane, released from warmed sea floor in the oceans as well as from melted permafrost on land, has also accumulated in the atmosphere and enhanced the global warming so that the Earth’s average temperature has risen by 8°C. The resulting disaster is particularly acute inside the Arctic and Antarctic circles where temperatures have risen by 20°C and the ice caps have completely melted away. Our planet had been in the grip of an ice age in the geologically recent past but now has no sea ice at all. Catastrophe has come to the tropics as well as to the poles. There had been deserts to the north and south of the equator before climate change began but these harsh, sparsely inhabited zones were separated by a wide, lush band of tropical vegetation straddling the equator. That verdant, rich band of life has been squeezed out of existence by greatly expanded deserts that have coalesced. Half the entire surface area of the continents is now arid and bare. Matters are, if anything, even worse in the seas than on land. Currents, normally driven by the sinking of cold, dense polar seawater, have stagnated so that life in the deep oceans is dying from lack of oxygen. Animals and plants experience an unprecedented mass extinction as species encounter extremes of heat and drought on land and oxygen-starvation in the oceans. Ninety-five per cent of marine life and 70 per cent of terrestrial species become extinct. Armageddon comes because too much carbon dioxide has been dumped into the air.
This horrific vision is not from 250 years into the future. This is an ancient global warming catastrophe from 250 million years in our past. The carbon dioxide was released by huge volcanic eruptions that covered much of present-day Siberia in a layer of solidified lava and volcanic ash several kilometres thick, and the resulting temperature rise was enough to wipe out nearly all multi-celled species. It would be hard to find clearer evidence that complex organisms are vulnerable to even quite moderate climate change. Perhaps the best known animals to disappear at this time were trilobites, distant relatives of modern horseshoe crabs. Trilobites had successfully dominated the seas for nearly 300 million years but nothing had prepared them for rapid global warming and they were completely exterminated. The environmental changes were simply too fast to allow adaptation by these relatively large, slow-breeding organisms. The species that vanished were eventually replaced by many entirely new ones but this took a long time and Earth’s biodiversity was measurably reduced for 10 million years after the catastrophe. The dramatic change in dominant life-forms revealed in the fossils from this time has led geologists to assign different names to the Periods either side of this event. The preceding 50 million years is the Permian Period and the succeeding 40 million years is the Triassic Period; the disaster is therefore known as the Permo-Triassic mass extinction event.
The pattern of death and, eventually, new life seen in the Permo-Triassic mass extinction was repeated in each of the other, similar catastrophes that have hit our planet five times in the past half-billion years. In each disaster a substantial fraction of existing species died out to be replaced over the next 5 to 10 million years by new animals and plants that evolved from the survivors. Extinctions on each occasion were caused largely by bad weather, although the exact causes of the climate change varied from one catastrophe to the next. The best known example is the mass extinction 65 million years ago at the end of the Cretaceous Period. In that case there seems to have been a volcanically induced gradual deterioration in the climate followed by a coup de grâce of a major meteorite impact that would have had additional severe climatic effects. The result was the extinction of the dinosaurs, along with many other life-forms, but the subsequent opportunities opened up for new species allowed mammals to become the dominant large animals of the present day. Occasional disasters therefore encourage evolutionary innovation, but you can have too much of a good thing. If major climate catastrophes happened every few million years there would be insufficient recovery time between the successive mass extinctions and biodiversity would be permanently reduced. Fortunately, on Earth, the average gap between the really big disasters has been about 100 million years and life has been able to recover and thrive after each one.
There have therefore been quite a few mass extinction events but the Permo-Triassic was the biggest one to hit life in the last 542 million years. This may seem a slightly odd statement. Why not count catastrophes over, say, the last 1 billion years or, more sensibly still, over the entire 4.5 billion-year lifetime of our planet? In fact, 542 million years is not an arbitrary choice. This Eon (geo-jargon for a very long period of time) is called the Phanerozoic, ‘the time of visible life’, because it’s only during this most recent half-billion-year stretch of time that animals have existed with easily fossilised parts such as skeletons, exoskeletons and shells. So, the Permo-Triassic event is simply the biggest mass extinction we have clear evidence for. We have too few fossils from further back in time to be able to tell whether earlier life suffered even worse setbacks. However, it is likely that more severe disasters actually did occur in the distant past because climate variations seem to have been more extreme in those ancient times. The Eon preceding the Phanerozoic is called the Proterozoic, ‘the time of earlier life’, and this Period, which began 2.5 billion years ago, experienced the most savage temperature swings the world has ever seen. Evidence for Proterozoic climate change dwarfing that at the end of the Permian is found at many places around the world such as Canada, Brazil, Namibia, Svalbard, Australia and Oman. I’ve recently been to see these clues for myself in yet another location: Shetland.
The Isles of Shetland are a bleakly beautiful group of islands so far north of Britain that they are closer to the Arctic than they are to London. Shetland’s many peninsulas and interspersed sea lochs give these islands a characteristically intricate coastline that was formed, much like the Norwegian fjords, by the action of glaciers during the Earth’s current ice age. But the story of a much earlier and more extreme ice age is also written into Shetland’s rocks. On the peninsula of Strom Ness I have placed my hand on sediments that show spectacularly rapid and dramatic climate change. In a photograph I couldn’t resist taking, the index finger of my left hand rests on sediments deposited during a time of near-global glaciation while my ring finger sits over carbonate rocks laid down in a world far warmer than our own. The change-over, hidden by my middle finger in the picture, was abrupt but with no sign of any break in deposition. It looks as if the climate switched from bitterly cold to stiflingly hot almost overnight or, at least, over thousands rather than millions of years.
So, what’s the evidence in these rocks for the dramatic climate change I’ve claimed? I’ll come back to the carbonates and how they relate to a very warm world later, but, for now, let me concentrate on the cold-climate sediments preserved at Strom Ness. These are covered in variegated patches of lichens, which, while helping them to stand out from the relatively lichen-free carbonates, makes it hard to see the distinguishing features of deposition in chilly times. However, after rooting about for half an hour or so, I found what I was looking for. On a relatively lichen-free area I could see that the rocks were made from very fine grains arranged in beds typically a millimetre thick or less. This can happen only if the sediments forming the rock were laid down in a quiet environment where even small grains could settle out to rest on the sea floor and where the thin beds thus formed would not be stirred up and disarranged by waves, currents or tides. But, in contrast to this evidence of tranquillity, I could also see a pebble that appeared to have been dropped onto this ancient sea floor from a great height. The beds below the pebble were clearly deformed by the impact and, in any case, the surrounding sediments showed that there were no currents capable of moving such a relatively large object. The pebble seemed to have fallen out of the sky. Fortunately, there is a simple explanation for such an apparently unlikely thing. The pebble probably dropped off the bottom of an iceberg, since there is no other known, natural way of transporting isolated large stones into calm water. Of course, one dropstone on Shetland does not a snowball Earth make, but rocks of similar age in other parts of the world show even more dramatic examples with dropstones sometimes a metre or more across sitting in otherwise fine sediment. Many other types of glacial sediment have also been seen all over the world from rocks of this age, and some of these cold-climate rocks are found in places known to have been near the equator at the time. The Earth must have been a very cold place 635 million years ago and may even have been completely frozen over.
Computer climate models show that an almost completely frozen Earth is not only possible but actually a likely outcome once polar ice caps grow to within about 30 degrees latitude of the equator. A runaway cooling then sets in where the increased ice cover reflects heat into space, causing further cooling and yet more ice until the whole world is white, beautiful and all but dead. In these models the Earth becomes a permanently frozen globe with an average temperature of –50°C. This prediction remains controversial, however, with many geologists arguing for a slushball rather than snowball Earth, but that is just an argument about whether equatorial temperatures were lethally low or merely bitterly cold. From my point of view the important and undeniable message from the rocks is that the Earth had a very different, and much colder, climate 635 million years ago than it does today.
So, what of the carbonate rocks at Strom Ness, which tell a dramatically different tale of an extremely warm Earth? Similar ‘cap carbonates’ are found on top of late-Proterozoic glacial deposits at many locations around the world and they record the spectacularly rapid end of global glaciation. A cold world is a very dry world with most water locked up in ice and too little heat to allow much evaporation into the atmosphere. On snowball Earth it never rained and rarely snowed; like C.S. Lewis’s mythical land of Narnia it was always winter but never Christmas. As a consequence of this lack of rainfall, the removal of carbon dioxide from the atmosphere by acid rain, as discussed in the last chapter, ceased; but since volcanoes continued to emit the gas, it built up in the atmosphere. After about 4 million years there was enough greenhouse warming from this to allow equatorial ice to begin melting. The process that drove the Earth into a snowball state then went into violent and rapid reverse. The disappearing ice reduced the reflection of heat into space, giving further warming and even less ice. This became a runaway process with melting driven at an ever-accelerating rate until all the ice had gone.
With the rising temperatures, highly acidic rainfall falling out of the carbon dioxide-rich atmosphere onto volcanic rocks produced the calcium carbonate in the sea that was needed to generate the thick, globally extensive limestone seen in Shetland and elsewhere. Within a few thousand years of the thaw’s beginning, ice completely disappeared from the Earth and our planet entered a super-greenhouse phase in which global mean temperatures may have risen to as much as 50°C above freezing and only slowly returned to more normal levels as the carbon dioxide concentration gradually fell. Thus, in a geological instant, temperatures possibly rose by as much as 100°C. This makes the 8°C change in the Permo-Triassic mass extinction event look like a very minor incident – and this snowball-to-greenhouse upheaval happened not just once but at least four times during the late Proterozoic. Furthermore, a similar phase of repeated snowball Earth episodes also happened in the early Proterozoic around 2 billion years ago. The fine details of all this Proterozoic climate instability remain highly controversial and many experts would disagree with much of what I have said in the preceding few paragraphs but, nevertheless, it remains incontrovertible that our planet was capable in the past of exhibiting very severe climate change; climate change bad enough to wipe out most life-forms and leave behind an unimaginably impoverished biosphere.
In the last half-billion years, temperatures have become more consistent and, for these more recent times, there is less room for debate about how much climate change really occurred because the temperature history is written in the fossil record. That may seem strange. Bones and shells fossilise but softer parts usually do not. The claim that temperature fossilises therefore seems quite outrageous, but, thanks to my clever geochemical colleagues, it really is possible to estimate ancient temperatures. Advances in recent decades, in both the precision with which the chemical composition of ancient rocks can be analysed and in understanding what these analyses can tell us about ancient environments, have been extraordinary. This is the Earth-science breakthrough I alluded to in Chapter 1 and it has provided us with deep new insights into the history of our planet. The chemical fingerprints of ancient seawater temperatures left in the shells of ancient sea creatures are an excellent example of this, and they result because the water incorporated into the carbonate of these shells became heavier when the weather was colder. Let me expand on that.
You may well have heard of heavy water and know that the Nazis attempted to manufacture large quantities of it during the Second World War using cheap Norwegian hydro-electricity. This was a major part of their fortunately unsuccessful nuclear weapons programme. In heavy water, all the hydrogen in the water molecules has been replaced by deuterium, a rare form of hydrogen that is twice the normal weight. This deuterium is an example of what is called an isotope, and elements other than hydrogen have isotopes too. In particular, the heavy water that interests climate scientists is quite different to that used by nuclear engineers because it is made with heavy oxygen rather than heavy hydrogen. Oxygen usually comes in a form called oxygen-16 (which means it has eight protons and eight neutrons in its nucleus), but one atom in every 500 is oxygen-18 (with two extra neutrons). Heavy water containing oxygen-18 occurs quite naturally. In fact, your body holds about 100 grams of it. Glaciers, on the other hand, contain hardly any of this oxygen-heavy water at all. The extra weight of these water molecules that contain oxygen-18 makes them harder to evaporate and also more likely to return quickly to the surface as rain. So, by the time water has been evaporated from the tropical seas and transported in clouds to the poles, it contains little of the heavy oxygen. Ice caps are therefore made almost entirely from what you could call light water. The overall process is quite similar to that used for making strong alcoholic drinks from liquids containing less alcohol. In the same way that evaporation and re-condensation of fermented liquid concentrates the easily evaporated alcohol, so evaporation and condensation of sea-water concentrates the easily evaporated light water. The Earth acts like a large whisky distillery, although sadly the products are not quite as tasty.
This distillation of seawater can be used to unravel ancient climate because the heavy, oxygen-18 containing, water molecules become a little more concentrated in the sea when there is lots of the light water locked up in ice caps. If this heavier seawater is then incorporated into growing shells, the variations in oxygen-18 content over time become preserved. In fact, the variation is further enhanced since organisms take up oxygen-18 more efficiently when the water is cooler. There are also complications resulting from the more acidic nature of the ancient oceans when carbon dioxide levels were higher, because this acidity also affects the efficiency with which organisms absorb the heavier isotope of oxygen. However, once all these effects have been disentangled, what emerges is an estimate of equatorial temperatures throughout the 542 million-year period that shelly animals have existed on Earth.
Equatorial temperatures vary less than the Earth’s overall mean temperature because polar temperatures always go up and down more strongly than equatorial ones. The results therefore need to be amplified a little to give the global picture, but what emerges is the very robust conclusion that, during the Phanerozoic, temperatures oscillated up and down three or four times with an amplitude of about 10°C. In addition there may have been an overall cooling trend of a few degrees, although this is still being debated. This picture of oscillation around a roughly constant temperature is quite different to what we’d expect based on the climate sensitivity models discussed in the last chapter. Those models predict that temperatures should have climbed during the Phanerozoic by around 10°C as a result of the steadily warming Sun. During the last half-billion years the cooling consequences of geological and biological evolution must therefore have compensated for the warming due to solar evolution. Evidence from further back in time is more contentious but, nevertheless, what there is reinforces the message. When it comes to climate change, Earth evolution-driven cooling has matched, or even outpaced, solar evolution-driven warming.
The other long-term trend that emerges from these analyses is that fluctuations in climate seem to have become less severe in the last half-billion years; we no longer have snowball Earth episodes. But can we be sure that the Earth has outgrown the adolescent behaviour of the Proterozoic and become a more mature, reliable and predictable planet in the Phanerozoic? The modern world is a very different place from the one recorded in the Shetland rocks, and there seems to have been a fundamental change on the Earth 542 million years ago when we entered the present Phanerozoic Eon. Life stepped up a gear at that time in what has been called the ‘Cambrian explosion’, the startlingly sudden appearance of widespread and recognisably animal life. As well as not having much in the way of animal life, the Proterozoic Eon differed from the Phanerozoic in that, even if we discount the snowball Earth events, it experienced much greater environmental change. At the beginning of the Proterozoic, oxygen levels in the atmosphere were a hundred times smaller than they are today and global mean temperatures may have been as high as 70°C. Noon-day temperatures at the equator must have approached boiling point! In contrast, by the end of the Proterozoic, oxygen levels and temperatures were broadly similar to those of today. For some reason the Earth’s climate settled down half a billion years ago and many researchers believe that this is a direct result of the new, wonderfully rich and complex Phanerozoic biosphere controlling the climate through a network of negative feedback processes that did not exist in the simpler Proterozoic world. This idea has been called the Gaia hypothesis, and I’ll look at it in detail later on.
However, it is possible that the Gaia hypothesis has simply confused cause and effect. Perhaps the complex Phanerozoic biosphere is a consequence of climate stability rather than its cause. If the Earth had had a more interesting climatic history – for example, occasional snowball episodes in the last few hundred million years – then our beautiful Phanerozoic biosphere would not have survived and we wouldn’t be here to record the fact. Gaia is therefore not necessary to explain the observations, since the anthropic selection effect I discussed in Chapter 1 accounts just as well for what we see today. This doesn’t mean that Gaia is wrong; just that there is an alternative explanation. To decide between Gaia and anthropic selection, we need to look a little closer at the causes of environmental and climatic change on Earth.