by Jean Jouzel
Others, quickly convinced of the existence of such a glacial period, turned into true detectives. They traveled the world in search of random boulders. Over the course of approximately forty years they were able to pinpoint the regions that in the past had been covered with ice. And they were huge regions: a large part of the North American continent, the northern United States and Canada, part of northern Europe, the British Isles, Scandinavia, northern France, and part of Siberia. A study of the mountain massifs located within very defined perimeters enabled scientists to estimate the maximum thickness of the ice; some massifs had rounded shapes that had been completely covered; others, as is seen by the presence above a certain altitude of sharp and very irregular peaks, had been only partially covered. And the corresponding volumes of ice are impressive; on the continents of the Northern Hemisphere, there were, we now know rather precisely, as much as 50 million km3 of ice, almost double the quantity that currently covers the Antarctic continent. Consequently, as evaluated correctly in 1868, the sea level was then some hundred meters lower than its current level. The geography of the coastal regions was thus rather different in certain regions from what we know today.
There were abundant theories. John and Katherine Imbrie discuss eight of them, developed between the middle of the nineteenth century and the 1960s. One of the earliest theories was based on the idea that this enormous quantity of ice was present because the Sun at that time shined much less intensely. Three of the theories, including that one, were immediately rejected, and the remaining five could not be tested, at least until recently. Among the latter there was the suggestion of a connection between glaciation and the concentration of carbon dioxide in the atmosphere which, we will later see, was advanced by the Swede Svante Arrhenius in 1896. It is, moreover, rather interesting to note that in 1986, the date of the publication of their book, John and Katherine Imbrie placed that theory in the category of ones that could not yet be tested. Thanks to the great ice core drillings in Antarctica, whose history we will look at in a later chapter, we now know much more about the connection between the climate and the greenhouse effect in the past.
It was only five years after Louis Agassiz had proposed the idea of a large advance of ice in the past that the theory (which was definitively established in the twentieth century) of a connection between the position of the Earth on its orbit around the Sun and this expanse of ice, henceforth known by the name of astronomic theory, was born. In 1842, the French scientist Joseph Alphonse Adhémar proposed this idea, then the Scottish scientist James Croll formalized it in a book that was published in 1875, and the Serbian mathematician Milutin Milankovitch gave it his seal of approval some fifty years later. We will soon look at the evolution of this theory whose validity was confirmed in 1976 from the analysis of marine sediment.
Let us emphasize that in Croll’s time astronomic theory had been rather favorably received because it foresaw that in the rhythm of our planet’s path around the Sun, there must have been not one glaciation but a succession of glacial periods interrupted by warmer periods. This is in fact what geologists had proven, noting the presence of different layers of matter, clearly of glacial origin, separated by soil in which one could identify pollen or vegetal matter undeniably appearing at times when the climate was milder. In 1909 the German geographers Albrecht Penck and Edouard Bruckner named the last four great glaciations that occurred in Europe after the names of German valleys: the Wurm, the Riss, the Mindel, and the Gunz. By examining the thickness of the sediments of the valleys, these two scientists estimated that the last glacial period occurred 20,000 years ago, which has proven to be correct. But at the same time that Croll was developing the bases of astronomic theory, other geologists were interested in the deeper rocks with the objective of describing the history of our planet as completely as possible. Between 1830 and 1865 the American Charles Lyell introduced a series of names to divide this history into eras and periods with lengths that were then unknown but were rightly thought to have been very long. Another mystery: glacial periods had taken place in the very distant past since traces of them were found in rocks from the Precambrian, the first period in the history of the Earth. The adventure of this ice of long ago, that present before the Quaternary, which covers the last two million years, is now relatively well-known but stills holds some surprises.
Ice of Long Ago
At least two ingredients are necessary for ice sheets to form on our planet: cold and enough precipitation in the form of snow in regions where there are continents. The temperature curve of the Earth’s surface since its birth 4.6 billion years ago (Figure 3.1) becomes more precise as we get closer to the current era. It shows that the Earth’s climate has been colder than it is today and thus conducive to a covering of ice greater than that of today—but only about 20% of the time. The appearance, evolution, and disappearance of glaciations throughout the history of our planet have been conditioned by the drifting of continents and by the evolution of the parameters that influence the energy available to heat our atmosphere to a greater or lesser degree: the luminosity of the Sun, the distance between the Sun and the Earth depending on the Earth’s orbit, and changes that have occurred in the composition of the atmosphere (greenhouse gases, aerosols).
Since our Earth was born, the luminosity of the Sun has increased by 25–30%. The brilliant American astronomer Carl Sagan (1934–96) played an essential role in the development of the American space program and in popularizing science. In an article published in 1972 with George Mullen he stressed the paradox of the pale Sun. Its weak solar luminosity would have caused the Earth to resemble a ball of ice during its first two billion years of existence. However, the geological and paleontological evidence tells us that running water and living matter had already appeared on the surface of the planet. Other factors thus influenced the primitive climate, and we suspect that large quantities of greenhouse gas, water vapor tied to the presence of oceans, and carbon dioxide exhaled by the Earth’s crust were present in the atmosphere and compensated for the solar deficit.
Figure 3.1. Temperatures throughout geological history. Source: Sylvie Joussaume, Climat d’hier à demain (Paris: CNRS Éditions/CEA, 2000).
As we mentioned earlier, the climatic history of our planet is essentially “warm.” It was interrupted only from time to time by a small number of short (on a geological scale) periods characterized by more or less intense glaciations. The first to have been recorded dates from around 2.3 billion years ago, as proven by the stria that have been observed in rocks dating from that distant time and identified in many places. Indeed, only the friction of the ice could have sculpted such grooves on these rocks.
Then, it seems, there was nothing more of note; there were no known glaciations before 900 million years ago. Since then we encounter a past that was relatively rich with glaciations but also with very hot periods without ice. Thus there is strong evidence that the submerged land was substantially iced over several times between about 750 and 600 million years ago; our planet really did resemble a true snowball. At the dawn of that period, the land was but one supercontinent that was beginning to fragment. Almost all the continents were then gathered in the Southern Hemisphere, abandoning the Northern Hemisphere to a vast ocean. An a priori paradoxical situation, the ice sheets would have on at least two occasions reached the tropical or even equatorial regions.
How can we reach these conclusions for such a distant era? Early on geologists noted that many glacial deposits were preserved in layers formed in tropical latitudes during that era; the same was true for rocks exhibiting the characteristics of debris taken out of the ground by the friction of the moving glacier and then released on the bottom of the ocean by icebergs calved by the glacier. Paleomagneticians then came onto the scene tracing the alignment of the magnetic dust contained in glacial deposits. That alignment is governed by the Earth’s magnetic field, which orients magnetic minerals in its direction. It proves to be an ingenious tracer of the latitude at which geolo
gical deposits were formed. Since the lines of this field go from one pole of the planet to the other, they align the magnetic material parallel to the surface—horizontally—at the equator and perpendicularly—vertically—at the poles. Thus the alignments of the glacial deposits of this era, the Neo-Proterozoic, suggest a formation at very low latitudes, tropical and equatorial. But it still had to be proven that in the course of the past 700 million years the magnetic signature of these rocks was not altered chemically or by any other process. This has been proven since the geological layers record several marked transitions in the latitude of the deposits, which would not have been observed in the case of a remagnetization of the material after its formation.
The paradox of a glaciation that extended as far as the equator motivated theoreticians and global climate modelers as early as in the 1960s. But the idea of “Snowball Earth,” a term invented in 1992 by Joseph Kirschvink,2 a geobiologist at the California Institute of Technology, took root out of observations in low latitudes suggesting not only that there was ice on the continents but that there were also frozen oceans. In fact, the almost permanent covering of the ocean by a layer of ice acted like a huge lid protecting the seawater from any exchange with the atmosphere, and the composition of the ocean was modified because of this; the sediments deposited on the ocean floor have retained the imprint of such modification.
These unusual periods remain marked by a certain mystery. Some scientists think that continents and oceans were covered with a true shield of ice; others believe there were free waters. How could there have been a world so different from the one we know today? And, above all, how did our planet manage to survive such change? The formation of the “snowball” would have been linked to the fact that the luminosity of the Sun, 6% weaker than today, was combined with low levels of carbon dioxide concentrations. Over these very long time periods, the erosion that brought carbonate sediments to the ocean represented a very effective carbon dioxide pump. According to Yannick Donnadieu3 in Saclay and his colleagues, around 750 million years ago the relatively large parceling of the continent called Rodinia would have favored running water, thus erosion. The result: less of a greenhouse effect and a progression of the ice toward the equator that nothing could stop because the ice reflected an increasing amount of solar energy into space.
Greenhouse gases may have entered the scene again and helped the Earth escape from this “snowball” condition. With a layer of ice, there was no erosion of rocks and thus no trapped carbon dioxide being released into the atmosphere; however, volcanoes remained active and continued to emit carbon dioxide, which accumulated in the atmosphere to a point where it overpowered the reflective abilities of an ice-covered planet. Climatic simulations reveal that this required enormous concentrations of carbon dioxide, 350 times more than now, on the order of 10%. These figures are perplexing, and other theories implicate methane, another greenhouse gas. The hypothesis of a variation in the Earth’s axis of incline was also advanced: strong angles of incline would favor colder temperatures in tropical and equatorial regions. This theory seems, however, to have lost its credibility. The debate goes on.
The luminosity of the Sun gradually approached its current levels, thereby eliminating in a certain way this risk of extraordinary glaciations. The climate was characterized by long, hot, humid periods interrupted by two glacial episodes, one at the end of the Ordovician, the other at the beginning of the Permian, around 450 and 300 million years ago, respectively. The ice sheets then developed in the polar regions, even if their geological traces are visible today at lower latitudes. Thus, at the end of the Ordovician, current West Africa was located at the South Pole in the middle of a huge continent, Gondwana, which encompassed Africa, South America, Antarctica, and Australia. So it is in the Sahara that traces of this Ordovician glaciation—extremely well preserved with an extensive ice sheet centered on the south of Hoggar—have been found, though (with a surface area estimated at 8 million km2) it was less extensive than Antarctica’s current ice sheet.
After the Permian glaciation, the climate became warm again—indeed, very warm, probably an average of 6°C above the current temperature—around 100–50 million years ago. Fifty-five million years ago there was an abrupt warming, the Paleocene-Eocene Thermal Maximum, which was likely connected to a large increase of either methane, due to the decomposition of clathrates, or carbon dioxide connected to intense volcanic activity. Then the climate gradually became colder; the cold reached Antarctica, which gradually moved farther away from Australia, taking position around the South Pole. This enabled the glaciers, and even isolated ice sheets, to settle at least temporarily on the continent. Of course we cannot observe the signs that they left on the bedrock because today it is covered with a sheath of ice that is several kilometers thick in places. But they calved icebergs into the neighboring seas—icebergs which, in melting, deposited grains of sand and rocks torn from the ground by the ice during its journey to the sea. Paleoceanographers indeed find debris dating from that era in marine sediment that they extract from the bottom of the Southern Ocean.
In the temperature curve in Figure 3.1 we immediately notice the particularly abrupt (on the scale of geological time) cooling that occurred around 34 million years ago. That was the transition between the Eocene and the Oligocene. This was the era when the first ice sheet covering almost the entire Antarctic continent appeared. The glaciation of Antarctica constitutes a major event in the climatic history of our planet, and it has become a favorite subject among paleoclimatologists. Was it the “abrupt” global cooling of the planet that gave birth to that huge inland ice sheet, or did the latter, once formed, change the climate?
Interest in this event goes back to the 1970s. In analyzing microfossils preserved in the deep ocean sediment, two paleoceanographers, Nick Shackleton and James Kennett, revealed the global cooling that had occurred in the course of the last 65 million years and identified the step that corresponded to the Eocene-Oligocene transition. Soon afterward, Kennett associated this event with the separation of Antarctica, Australia, and South America. Begun some tens of millions of years earlier, this separation enabled the formation of a circum-Antarctic marine current suddenly isolating the continent from the flow of heat coming from more tropical latitudes and thus putting into motion the cooling of Antarctica. A plausible explanation, but one that might not be the only one, as proven by climatic simulations published in 2003.4 This work shows that the likely reduction of the atmospheric concentration of carbon dioxide by a factor of two in 10 million years would have had a noticeably greater impact on the cooling of Antarctica and the establishment of the ice sheet than that of the opening of the Drake Passage between South America and Antarctica. Thus the first large ice sheet of Antarctica was born perhaps under the preponderant influence of carbon dioxide, whereas the Southern Ocean was also being formed. By contrast, it was probably only 15 million years ago that a permanent ice sheet would have covered Antarctica with perhaps (here specialists are divided) extreme phases of deglaciation since then.
While Antarctica was separating from Australia and taking its position centered on the South Pole, the continents of North America and Eurasia were getting closer to the North Pole. More and more land was thus assembling in the high latitudes, both in the north and the south, and the ice of Antarctica reflected an increasing amount of energy into space. It thus participated in the cooling of the surface of the planet, likely helped by a decrease in the concentration of atmospheric carbon dioxide. Around 10 million years ago, the glaciers began to appear in several regions of northern high latitude, such as Alaska. But it was only in the Quaternary, the last three million years during which the Earth existed in its current topography, that true ice sheets developed in the northern regions of the Northern Hemisphere, including the Greenland ice sheet, the only large ice sheet that still exists in the north today.
Glaciations of the Quaternary and Astronomic Theory
The northern ice sheets then initiated a
long series of fluctuations, whose rhythm was gradually modified with a dominant periodicity of nearly 40,000 years for up to around 1.2 million years. It then evolved toward a periodicity close to 100,000 years, with very pronounced glacial and interglacial periods for the last 500,000 years or so. The last million years have been the subject of a good deal of research, owing to an impressive wealth of marine sediment cores from the bottom of the oceans. They have enabled us to precisely define the rhythm of oscillations leading to the successive growth and disappearance of the large ice sheets of the Northern Hemisphere and to confirm the existence of a connection between the rhythm of the glaciations during the Quaternary and the position of the Earth on its orbit around the Sun.
Astronomers such as Johann Kepler, who in the seventeenth century demonstrated that this orbit is not a circle but a slightly flattened ellipse, are thus indispensable allies for anyone interested in the earthly climate or in that of the other planets. We must know everything about the trajectory of the Earth to be able to calculate the amount of solar radiation the Earth receives in a given place throughout an entire year (insolation), to understand that it varies from one year to the next, in a very slow but irremediable way, and to precisely determine these variations both past and future. This measure of insolation depends on three so-called astronomical parameters: eccentricity, obliquity (or tilt), and the precession of the equinoxes (Figure 3.2).