Music of the Spheres
The stars and planets were much more familiar to our ancestors than they are to most people today. The absence of bright street lights, until very recently, partly accounts for this greater familiarity but it was also due to the fact that their survival as farmers, hunters or gatherers relied on understanding the passing of the seasons as they are reflected in the changing constellations visible through the year. Simple observations of the night sky told farmers when to plant or harvest their crops and told hunter-gatherers when to look for migrating prey. One of the best known examples of this comes from ancient Egypt, where the first appearance of the brightest star in the sky, Sirius, heralded the annual flooding of the Nile and was anxiously awaited every year. Thus, regularities in the heavens were both useful and reassuring. Given this background, it’s not surprising that all early civilisations were confused and intrigued by the planets (from the ancient Greek word planetoi meaning wanderers): their positions change, making the timing of their reappearance each year much harder to predict than that of the fixed stars. It seemed reasonable to assume that planets, like the fixed stars, foretold something useful but that the gods had made the message more cryptic and readable only by an elite priesthood. Thus was astrology born.
Astrologers have therefore long held that planets affect our destiny. They are, of course, completely correct. Gradual changes in planets’ orbits have climatic effects and, if this book has a theme, it is that climate is destiny (or at least a precondition for destiny). The orientations and shapes of planetary orbits slowly oscillate over tens of thousands of years and these mild orbital vibrations, along with moderate wobbling of the Earth’s axis, drive the ebb and flow of ice ages. My tour of past climate change is therefore incomplete until I have described the gentle celestial dance of our solar system and how this has driven glaciations, especially over the last 2.5 million years, during which we have been living through one of the Earth’s rare ice ages.
The discovery of the ice age, repeated episodes of significantly colder climate in the geologically recent past, is one of the great achievements of 19th-century science. The key moment was a meeting of the Swiss Natural History Society in the town of Neuchâtel in 1837. Neuchâtel had been chosen to host this prestigious meeting because of the irreproachable reputation of the curator of the town’s museum, the biologist Louis Agassiz. Expectations were high that, in his welcoming speech, Agassiz would reveal some of his latest findings on fossil fish; work for which he had rightly garnered international respect. Instead, Agassiz described evidence that Switzerland had once been the location of a massive ice cap to rival that of modern Greenland. His audience was shocked, and none more so than the people who had shown him the evidence, his friends Jean de Charpentier and Ignatz Venetz. They’d only been trying to convince Agassiz that Switzerland had once been marginally cooler; Agassiz’s reinterpretation that vast swathes of the entire world had been covered in ice took things much further than they were ready to believe. Agassiz’s proposed ice age seemed a regressive step at a time when the young science of geology was still struggling to establish that the rocks of the Earth showed the cumulative working of natural processes over immense periods of time rather than the effects of biblical catastrophism.
Agassiz therefore needed to work hard to strengthen support for his ideas, and he visited Scotland in 1840 specifically looking for evidence that a country that now has no glaciers at all had been the centre of vast ice sheets in the past. He found what he was looking for and it convinced him that much of the northern hemisphere had once been covered by thick ice extending from the pole down to the latitude of the Mediterranean. The evidence was there for all to see. Moraines, for example, chaotic piles of boulders and sediment bulldozed into ridges typically tens of metres high by glaciers, were soon found in Scotland by Agassiz and others. Other features too can be found throughout northern Europe, such as erratic boulders the size of houses sitting in fields after being carried by ice many kilometres from their place of origin. The Scottish lochs and Norwegian fjords are also signatures of recent glaciation: straight, steep-sided channels are characteristic of valleys carved by ice. These valleys became flooded after the glaciers retreated, forming the characteristic long, narrow water bodies. Within a few decades, unmistakable glacial hallmarks were also found in other parts of Europe and in North America and Asia; these showed conclusively that Agassiz’s idea of a widespread ice age was accurate.
In fact, there have been many such ice ages or, more accurately, many phases of glaciation within a single ice age that has unfolded over the last 2.5 million years. Agriculture, civilisation and the whole of written human history have all occurred since the current ‘interglacial’ period of warmer weather began just over 11,000 years ago. We can’t be sure when this brief interval of more clement weather will end, but interglacials tend to be much shorter than the periods of truly cold climate that they punctuate. What drives this ebb and flow of ice caps? The idea that it was due to changes in the Earth’s orbit was suggested almost immediately, although this remained a controversial idea until very recently.
Mankind’s quest to understand planetary motions probably began when we first noticed these wanderers in the sky, perhaps 200,000 or more years ago. We can only speculate about what early watchers of the sky thought of the complex movements of the planets, but a desire to find an underlying beauty and simplicity is clear by the time of the semi-legendary Pythagoras who was born on the Greek island of Samos in about 570 BC. Pythagoras is best remembered today for his eponymous theorem; the one about the ‘square on the hypotenuse’ that links the side-lengths of a right-angled triangle. This early philosopher is a rather shadowy figure. We can’t be sure which achievements were truly his own and which came from his followers and successors. Even his theorem may have been discovered centuries earlier by Egyptian, Indian or Babylonian mathematicians, although this suggestion is controversial and unproven. Nevertheless, even if Pythagoras and his followers didn’t discover it themselves, ‘the theorem of Pythagoras’ was just what they were looking for, because they believed that links between geometry, numbers and the natural world were spiritually significant. In a search for similar relationships to bolster these beliefs, the Pythagoreans stumbled on another equally interesting connection, one between geometry and musical harmony. They discovered that two strings, identical in every way except for one being twice the length of the other, produce notes separated by exactly one octave. More than this, they found that all harmonious pairs of musical notes have string lengths with simple ratios. Strings with lengths in the ratio of two to three, for example, are aurally separated by what musicians call a ‘fifth’ and they sound pleasing together. This discovery reinforced the Pythagoreans’ belief in the interconnectedness of all nature and encouraged them to look for similar relationships in the heavens. They began to search for a ‘music of the spheres’: harmonious relationships linking the planets to one another in a similar way to the links they had found between harmonious notes.
Two thousand years later this rather vague idea was taken up by Johannes Kepler, a German mathematician born in 1572, who was much more precise about what exactly celestial harmony should look like. Kepler was an early supporter of Copernicus’ heliocentric ideas. He appreciated the simplicity of Copernicus’ Sun-centred scheme, compared to the complex apparatus needed to explain planetary motions if the Earth lay at the centre of the Universe, and he wanted to reinforce this elegance. More specifically, he wanted to understand the architecture of the solar system; the way that the worlds are distributed through space. It was already known that the planets are crammed relatively closely together in the inner solar system but are more spaced out as we move away from the Sun. Kepler believed that this pattern was important and that the increasing separations of successive worlds should obey some simple but mathematically beautiful rule. Along with most astronomers of this pre-Newtonian time, he imagined each planet to be embedded in the surface of a crystal sphere centred on the Sun, so that the solar system consisted of a set of nested spheres. He then suggested that five very special shapes, the Platonic solids, could be inserted precisely into the five gaps between these six spheres.
Platonic solids are a set of particularly simple three-dimensional shapes that were first investigated by the ancient Greeks. The best known of these is the cube, a shape made from six identical square faces. The tetrahedron, sometimes known as the triangular pyramid, is another shape that can be made from identical faces; in this case four equilateral triangles. All the Platonic solids have this property of being constructed from identical faces that, in turn, have sides of identical lengths. The Greeks discovered that there are only five such shapes, the other three being the eight-faced octahedron, twelve-faced dodecahedron and twenty-faced icosahedron. Kepler was impressed that the number of gaps between the six known planets was exactly equal to the number of Platonic solids and so he wanted this to be a law of Nature – an expression of the world’s beauty and harmony. In Kepler’s scheme a cube, for example, just fits inside Saturn’s sphere while perfectly enclosing the sphere of Jupiter. Other pairs of orbits are similarly filled by the remaining Platonic solids.
This recipe for constructing a solar system sounds hopelessly mystical to the modern mind and turned out to be completely wrong. For a start, this solution to the Pythagorean search for celestial harmony doesn’t even give the right architecture for the solar system. A cube between the spheres of Jupiter and Saturn places Saturn 86 million kilometres closer to the Sun than it really is. This became obvious within Kepler’s lifetime, but even if the theory had predicted orbital sizes with sufficient accuracy to be more credible, it would have been dealt a mortal blow by the discovery of Uranus in 1781. There were no Platonic solids left to place between Saturn and this new world. However, enlargement of the solar system’s retinue of planets still lay almost 200 years into the future and so Kepler initially persisted with his ideas. This was one of the most productive mistakes in the history of science. While trying to prove a link between Platonic solids and planetary spacing, Kepler stumbled instead on the first mathematically accurate descriptions of planetary orbits: Kepler’s laws of planetary motion. Kepler’s first law simply states that orbits are not circular, as everyone had believed, but are instead elliptical. His second law describes how the orbital speed of planets increases and decreases as their elliptical orbits bring them closer to, or further from, the Sun. Finally, in 1619 and ten years after publication of the first two laws, Kepler revealed a third relationship that showed how the duration of an orbit increases with distance from the Sun. This third law explains why Mars’s year is twice as long as our own, for example.
Kepler’s laws are still used today and they were among the first-ever examples of mathematical formulas that accurately describe a facet of our Universe. If Kepler hadn’t discovered them it is debatable whether the mathematical laws of mechanics would have been formulated by Galileo in the 1630s and whether Isaac Newton’s law of gravity could have been published in 1687. Kepler’s somewhat misguided search for celestial harmony was therefore a fruitful one since it played a major role in the emergence of modern science. In any case, it should be seen as part of a long tradition that continues to this day of looking for beauty in our descriptions of nature.
The Pythagoreans’ search for a link between musical harmony and solar system architecture therefore ultimately failed, but another aspect of how strings vibrate does turn out to be useful for understanding planetary systems. The Pythagoreans investigated the note produced by a string but, in reality, no string produces a pure note. Instead, the dominant pitch of a string is contaminated by fainter contributions from other notes, overtones as they are known, and they give the sound of a string its unique character. In exactly the same way, planetary orbits do not wobble at a single frequency, but instead undergo a combination of many different vibrations.
To explain this more clearly, I’ll stick with musical character for a little while longer before coming back to planets. All musical instruments produce notes that include overtones but it’s the different combinations of these that give them their distinctively different sounds. If all instruments produced a perfectly pure note they would sound identical and music would be far less beautiful. Stringed instruments, trumpets and glockenspiels vibrate in unique ways that combine many tones, and so they sound quite different to one another. Any instrument playing a middle C will vibrate about 260 times per second, but superimposed upon this fundamental note will be fainter vibrations at other frequencies. For example, most instruments also vibrate at twice this rate, 520 times per second, but the exact size of this additional contribution to the sound will vary from one case to another. Brass instruments do not produce very much of this so-called first harmonic and, instead, vibrate strongly three times faster than the fundamental note at 780 times per second. It’s the near-absence of every other harmonic that gives that distinctive ‘brassy’ sound. A piano string, on the other hand, produces a gradual dropping off in intensity with each increasing harmonic. Overtones of many instruments also include vibrations at frequencies that are not simple multiples of the fundamental note, and with these too the exact contribution to the overall sound varies from one instrument to another.
This idea that the characteristic sound of an instrument results from adding together a fundamental note and an instrument-dependent set of overtones applies equally well to the behaviour of planets. The planets in our solar system have orbits that vibrate at a fundamental period with oscillations at other periods, overtones if you like, added in to give a unique character to the wobbling of each world’s orbit. Imagine each orbit as a hoop that, instead of lying flat, has been tilted a little. The direction and size of the tilt slowly changes through time, and each hoop’s shape also oscillates between being slightly oval and being circular. Planetary orbits, unlike musical instruments, therefore sing two songs simultaneously: one is the song of orbital orientation and the other is the song of orbital shape. However, there is also an important similarity. The solar system is like an orchestra tuning up before a performance. The individual instruments, the planetary orbits, all play the same notes but they exhibit distinct combinations of the overtones so that each planet, like each instrument in an orchestra pit, sounds a little different.
These orbital vibrations are generated by the gravitational attraction between the planets. A simple solar system consisting of a single planet orbiting the Sun is rather boring. The planet goes round and round the Sun for ever in an elliptical orbit that never changes its shape, never changes its orientation and never does anything interesting at all. The picture gets a little more complex if the planet, or the Sun, is not a perfect sphere but I’ll wait until later to describe some of the trouble that difficulty causes. For now, I want to look at a rather different complication. In the late 18th and early 19th centuries the French genius Pierre-Simon Laplace and the equally brilliant Italian, Joseph-Louis Lagrange, accurately showed for the first time how adding more worlds to a one-planet system made the behaviour of the system much more interesting. The orbits wobbled and changed their shapes over tens and hundreds of thousands of years to generate a genuine ‘music of the spheres’ in which every planet sang the same notes but in their own distinctive voices.
Of course, we can’t really hear the planetary orbits singing, because sound does not travel through the vacuum of space – and in any case the fundamental vibration of the solar system, a change in the orbital tilts over a period of 50,000 years, produces a note 50 octaves below middle C and far too deep to be detectable by human ears. Nevertheless, it would be fascinating to hear Jupiter’s hum turned to sound and speeded up 400 trillion times to upscale the fundamental frequency into a middle C. We’d start with silence when Jupiter was alone, but bringing in Saturn would generate the expected pure note near middle C as the orientation of Jupiter’s orbit started to wobble. We’d also hear that second song, another deeper note around F sharp, generated by the vibrating shape of Jupiter’s orbit. Then, as we added, say, Uranus with its particular mass and orbital radius, new overtones would be brought to the songs and the fundamental notes would shift very slightly. The richness of Jupiter’s song would increase further as each additional world was placed in its orbit until Jupiter was generating a complex hum. We could also listen to the songs of the other planets and there would be a family resemblance between them, because they would all be playing the same fundamental notes, but the different combinations of available overtones would create distinctly different songs.
Like going from a pure middle C to the sound of a violin string, adding planets generates beauty through complex overtones, overtones that ultimately control the rhythms of our ice ages. The full family of eight planets produces oscillations with periods ranging from 46,000 years up to 4 million years. Many of these control changes in the orientations of the planetary orbits. The Earth’s orbital wobbles, for example, are dominated by four or five vibrations the most important of which produces a 69,000-year variation in the tilt and tilt-direction of her orbit. Other overtones control the rate at which orbits change their shapes. The resulting main period of oscillation for the Earth’s orbital ellipticity is about 400,000 years with a number of other, slightly less important, periods clustering near to 100,000 years.
So Pythagoras and Kepler were, in a sense, correct after all. The solar system does behave like a musical instrument and its notes are controlled by the size of the orbits, together with the planetary masses, in a way reminiscent of how string pitches are controlled by their lengths and tensions. There really is a music of the spheres but it emerges only as a result of a complex analysis using 18th-century mathematical tools undreamt of 200 years earlier. Indeed, the full complexity is only now being unravelled, as a result of 21st-century computer power.
There is yet another set of wobbles in the solar system that need to be included if we are to understand how orbital oscillations affect ice ages. These occur because planets are not perfect spheres. All spinning worlds bulge a little around their equators as centrifugal forces generate the planetary equivalent of middle-aged spread. The Earth, for example, is 43 kilometres fatter when measured across the equator than it is between the poles. These equatorial bulges cause the individual planets to sing solo, by wobbling on their own spin axes, at the same time as they participate in the collective hum of the solar system. The planetary, as opposed to orbital, wobble is exactly the same as that of a child’s spinning top. I am watching a ‘Thomas the Tank Engine’ spinning top wobble in front of me on my desk right now as I write this paragraph. The toy is spinning round several times a second, and as it does so, its handle is slowly drawing a small circle in space every few seconds. The gradually slowing spinning top is precessing, as this motion is called, more and more rapidly and the wobble has just become so violent that the edge of the toy has touched the desk and stopped its movement. Planetary rotation axes precess in an identical manner to my spinning top but in the case of the Earth it takes thousands of years, rather than seconds, for each wobble. Like the handle of the spinning top, the Earth’s North Pole draws a giant circle in the sky, returning to its original position only after 26,000 years. At present, anyone standing at the North Pole and looking straight up on a clear night will stare almost directly at the star Alpha Ursae Minoris, better known as the pole star. This is a temporary, if convenient, fact. Precession of the Earth’s axis slowly alters the orientation of our planet so that, a few thousand years into the past or the future, the North Pole no longer points at the pole star.
In the case of the spinning top, precession is caused by gravity attempting to pull the toy over. However, it’s not easy to change the orientation of a spinning object, which is why spinning tops and motorbikes generally stay upright (bicycles are more complicated as there’s an additional effect involved). On the other hand, it is relatively easy to make the axis of a spinning object precess. If you get the chance, try it with a bicycle wheel. Hold the axle between your two hands and get a friend to spin the wheel. Then try drawing a circle in the air with the axle. You’ll find you can do it if you move the axle in the same direction as the wheel rotation, but that it is almost impossible if you try to go the opposite way. You’ll also find it very difficult to move one end of the axle in a straight line without it wobbling in a circle instead. So, the spinning top is trying to fall over but, because of gyroscopic effects, it is unable to do so and precesses instead.
Slow wobbling of a planet’s spin axis in space is caused in a similar way. Tidal forces tug on the equatorial bulges in an attempt to bring planets into ‘more upright’ positions. The Earth’s axis, for example, is not perpendicular to its orbit but is tilted by about 23 degrees. This angle is called the obliquity, and tidal forces try to reduce its size as they tug on the equatorial bulges. Solar-generated tides occur because the day-lit side of a planet is very slightly closer to the Sun than the night side. Gravitational attraction by the Sun is therefore marginally stronger on the day side, and this difference in gravity across the planet produces the stretching force responsible for tides. Additional tidal stresses are produced by moons when a planet also possesses these. These forces try to reduce a planet’s obliquity as they tug on equatorial bulges, but the tussle between tides and gyroscopic stability is a fight that neither side wins; precession of the planet’s axis occurs instead. As you might expect from all this, precession speed is controlled by the size of the tidal forces, the rate of spin, the obliquity and the size of the equatorial bulges.
To recap, the picture I have tried to draw here is of a dynamic, almost vibrant, solar system in which the orbits of the planets are constantly changing their orientation and shape while, at the same time, the axes of the planets themselves are experiencing continuous variations in the directions they point in space. Fortunately for us these astronomical oscillations are much smaller in the solar system than they could be but, nevertheless, this music of the spheres is loud enough to affect our climate. In particular, these celestial rhythms are undoubtedly the main factors controlling the ebb and flow of Earth’s ice ages.
Lagrange’s work on the solar system’s oscillations was already 40 years old when Agassiz dropped his bombshell about the existence of ice ages and so an astronomical driver for the growth and decay of the glaciations through time seemed an obvious explanation. However, as I’ve tried to emphasise above, the song of the solar system is quietly sung and its climatic effects should be tiny. At most, the heat received at any given season and place on the Earth’s surface changes by just a few per cent, and averaged over an entire year there are no significant changes in heating at all. Astronomical driving of the ice ages was therefore rejected until an extraordinary Glaswegian entered the scene.
In 1859 the 40 year-old James Croll became caretaker at a school in Glasgow after previous careers as an insurance salesman, millwright, carpenter and farm labourer. At Anderson College this underprivileged but highly intelligent man finally fell on his feet, because the college had an extensive library that Croll was allowed to use. He devoured its contents. Croll developed wide interests in scientific, theological and philosophical matters but he decided that, in particular, he wanted to solve the problem of the cause of the ice ages, and by 1864 he was publishing papers that drew the attention of the scientific establishment. Croll believed that build-up of ice sheets was the key to the problem. He suggested that, when the Earth’s orbit was particularly elliptical, the Earth would spend part of the year significantly further from the Sun, and if this corresponded with winter in one of the hemispheres, ice sheets would grow more extensively during the resultant severe winters and would not melt back completely in the summer. Glaciers would therefore get bigger and cool the planet. Hence, Croll was an early proponent of the ice-albedo positive feedback mechanism that I described in an earlier chapter. Ice-albedo feedback magnifies the climatic effects of orbital changes so that very slight changes in how the Earth is heated lead to relatively large changes in temperature. To test this idea, Croll used Laplace and Lagrange’s theories to calculate how ellipticity has changed through time. This is an extraordinary accomplishment in itself. I’ve done similar calculations using spreadsheets, and with the help of these, the hundreds of calculations necessary can be completed in a few minutes, but Croll had to do all the arithmetic by hand. This tedious and error-prone work must have taken him months. But what he found was encouraging. The best guesses of geologists for the timing of the cold periods coincided with Croll’s calculations for the times of greatest orbital ellipticity.
As a direct result of the impact he made through publishing this work, Croll was given a job at the Scottish Geological Survey in 1867, and by 1875 he had published a book setting out all his ideas in a single place. Croll’s astronomical theory of ice ages seemed triumphant, but within a few decades problems began to emerge. For a start, the theory implied that ice ages alternated between hemispheres. Sometimes the northern hemisphere would experience an ice age; at other times, the southern hemisphere would. Unfortunately, evidence was emerging of simultaneous cooling in both hemispheres. The geological dating of glaciations was also being revised and the newer results no longer agreed with Croll’s calculations. The astronomical theory fell out of favour.
However, the fortunes of the astronomical theory of ice ages were as cyclic as the phenomena it set out to explain, and belief in this mechanism came back into fashion again with publication, in 1920, of a paper by Milutin Milankovitch, a Croatian/Serbian professor of applied mathematics who had built his reputation analysing properties of the new wonder building material: reinforced concrete. Milankovitch’s ice age work was summarised in his book of 1941 in which he took Croll’s calculations a stage further and used knowledge of the orbital oscillations to calculate the expected changes in temperature at different seasons and different latitudes. What he discovered was that known glaciations corresponded to times of cooler summers at moderately high northern latitudes. This was almost the opposite of Croll’s assumption. It wasn’t cold winters but cold summers that were important, and this was because they reduced the melt-back of glaciers in the summer months. Milankovitch’s results showed that the Earth becomes colder when summers fail to melt the snow rather than when winters generate more ice. The fit of this new theory to geological data was impressive but, unfortunately, once again misleading. Subsequent, more accurate, dating of glaciations together with the realisation that some of the key deposits were not glacial at all destroyed the beautiful correlation between Milankovitch’s theory and the geological evidence. Scepticism over the astronomical theory grew once again.
The breakthrough finally came in 1976 with publication of a classic paper by James Hays, John Imbrie and Nicholas Shackleton relating astronomical drivers to climate data obtained by drilling in the ocean floor. Analysing sediments on the deep ocean floor has the great advantage that this gives almost perfectly uniform and continuous data extending back hundreds of thousands to millions of years. There has been an internationally funded programme of scientific drilling into the sea floor from 1968 through to the present day and the IODP (Integrated Ocean Drilling Programme), as it is now known, continues to provide some of the best scientific data on the history of our planet available from any source. The cores from this programme can be used to provide climate data such as oxygen isotope analysis (which indicates ice volumes in polar regions, as discussed earlier) along with analyses of planktonic remains that can give an additional indication of the sea surface temperature and salinity at the drilling site. By taking cores from a range of depths below the sea floor it is possible to see how these factors have changed through time. The work of Hays, Imbrie and Shackleton, along with many subsequent studies by other scientists, produced a beautifully simple result. There is a clear, 41,000-year cyclicity in the climate data superimposed on other cycles at 100,000 years and at about 20,000 years. In more detail, the 100,000-year cycles dominate over the last 700,000 years with the 41,000-year cycle dominating before that. Periods of about 20,000, 40,000 and 100,000 years are exactly what the astronomical theory predicts. Let me start with the 41,000-year cycle that results from changes in the tilt of the Earth’s axis.
Changes in this obliquity affect the intensity of our seasons. As a direct consequence of axial tilting, the northern hemisphere gets increased sunlight between April and September, which gives us summer, and reduced sunlight between October and March to produce winter. Of course, when the northern hemisphere is tilted away from the Sun, the southern hemisphere is tilted towards it and so the southern seasons are the reverse of those in the north. As the Earth’s axis and its orbit both wobble in space, the angle of obliquity changes by a degree or two and that alters the intensity of the seasons. When obliquity is small and seasons less intense, there is less melting of the polar ice caps in summer, making the Earth more reflective and cooler. Changes in the Earth’s obliquity occur more slowly than the 26,000-year axis precession because of the additional 69,000-year cycling of the Earth’s orbit itself and it’s the combination of these two wobbles that produces a changing obliquity. The resulting 41,000-year oscillation produces the signal so clearly seen in the ocean sediment data.
Axial precession of the Earth has another climatic effect that results because the Earth’s orbit is not quite a perfect circle. Today we are closest to the Sun in December when the northern hemisphere of the Earth is tilted away and experiencing winter. Conversely, we are furthest from the Sun in the northern summer. This mismatch between season and distance has the effect of reducing the seasonal intensity in the northern hemisphere. However, the southern hemisphere experiences exactly the opposite effect, since we are close to the Sun during the southern summer and furthest in the southern winter. Thus Antarctica, where the bulk of the world’s ice resides, is currently experiencing relatively intense summers that melt back much of the ice cap. However, precession of the Earth’s axis changes this relationship slowly over time. The timescale is again complicated by the fact that the Earth’s orbit is also precessing but the effect is to change the relative intensity of southern versus northern seasons on a timescale of about 20,000 years. This ‘climatic precession’ effect produces the 20,000-year cycles seen in the sea floor data.
Finally, the 100,000-year fluctuations in climate seem to be directly related to changes in the eccentricity of the Earth’s orbit, almost exactly as predicted by Croll except that cooling occurs when eccentricity is low rather than high. The reason is, as stated earlier, that glaciers grow when summers are cool rather than when winters are severe. It remains a puzzle why the effect of an elliptical orbit is so strong when the corresponding changes in seasonal intensity are relatively small. It is also unclear why there was a flip from dominance by 41,000-year cycles to dominance by 100,000-year cycles in the last million years. Despite these remaining uncertainties, astronomical forcing of the ice age fluctuations is now a well-established idea and rarely causes scientific controversy. But it is not the whole story.
Ice ages are actually rather rare in the Earth’s history. We are, as I have said, living through one now, but to find an earlier ice age we have to go back almost 300 million years. There was yet another, probably milder, ice age around 450 million years ago but, prior to that, we have to go back to the Proterozoic snowball Earth episodes to see other occasions when the Earth was at least as cold as today. Through the majority of Earth’s history our planet has been much warmer than today and almost completely free of any sea ice at all. Ice ages therefore clearly need some of the other factors I’ve previously discussed, such as the existence of a continent at the pole, blocking of marine currents or uplift of particularly large mountain ranges. At present we seem to have a triple-whammy of all three going on.
Given all this, you’d probably expect that climate is strongly affected by astronomical cycles only during these rare times of ice age conditions when ice-albedo feedback is able to amplify the small astronomical effects. At other times there is very little ice around, even in the winter, and so there is no opportunity for the ice-albedo effect to magnify the weak influence of the solar system’s song. However, many, but certainly not all, geologists believe that Croll–Milankovitch cycles can be detected throughout the rock record, even at times of unusually warm climate. This may well be correct but it is hard to explain and remains a highly controversial area. Fortunately, it is not an important consideration from the viewpoint of this book.
Much more important is another qualification I should make concerning the contents of this chapter before we move on. I have, perhaps, exaggerated a little the harmonious nature of planetary motions. The solar system’s ‘music’ is actually a rather inharmonious sound; more seagull than nightingale. None of the solar system’s notes would actually sound pretty together because they lack the simple relationships that Pythagoras found for harmonious strings. We’ll see later that this is fortunate. True harmonies in the heavens would lead to catastrophe because the normal, almost clockwork, regularities in the motions of the planets would break down and give way to unpredictable chaos. This is a fate we barely avoided! However, that’s for later. First, I’d like to move from geology and astronomy to biology. It’s time now to consider the possibility that life, itself, is the main guarantor of its own survival.