by Jean Jouzel
Further back in time, most of the information is provided by clues that are currently found on the continents. Although specialists go to Namibia to look for evidence of the great glaciations from 700 million years ago, the continental archives are also very rich in information regarding those of the Quaternary and of more recent periods. The evolution of temperatures, precipitation, and drought in certain regions, of the intensity of monsoons in others, of the vegetation, of the extent of the great polar ice sheets and the glaciers, of the atmospheric circulation that was more active in glacial periods than it is today—everything, or almost everything, is accessible to whomever looks at the continental archives and knows how to decipher their complexity. These archives also hold a very important place in our knowledge, discussed in the next chapter, of the climate that has prevailed during the Holocene for approximately 10,000 years. And it is again to these archives that we turn to learn of the climate of our regions during the recent centuries and to see the imprint of human activity on it over the last several decades. For example, by looking primarily at the long series of temperatures, at the historical data, and at data obtained from tree rings, a recent study has shown that the summer of 2003 was by far the hottest in Europe in five hundred years.17
CHAPTER 5
Glacial Archives
In this extremely rich context, one might think that the glacial archives from those very distant polar regions are of only marginal, almost anecdotal interest. This isn’t true. Although it is true that reconstructing the temperatures in Greenland or in Antarctica adds only a few more sites, the possibility of looking in detail at the evolution of the climate through the years, and over very long periods, is unequaled. Above all, glacial archives are unique because of their ability to trap the atmosphere of the past and, thanks to their extreme purity, to retain traces of the slightest effect, whether of continental, oceanic, volcanic, or extraterrestrial origin, or of that linked to human activity. Before describing the methods used by glaciologists to get a piece of ice to unveil all its secrets, we will briefly tell the long story of a snowflake from the moment it is formed to when it disappears in the ocean.
The Long Story of a Snowflake
The life expectancy of snowflakes that feed the surface of glaciers and ice sheets depends largely on climatic conditions. In cold areas without summer melting, such as the central regions of the ice sheets and the very high-altitude glaciers, they have a very long life, punctuated by several stages. Their youth varies from a few dozen to several thousand years. During this period the snow crystal rapidly loses its splendid needles, rounds out, and becomes granular, then, from the effect of more layers of snow, the grains pile up and are joined together. Nature has put into play on the surface of the white planet, well before the inventive genius of humans and on a gigantic scale, the process that powder metallurgists have perfected to create matter with satisfactory properties of cohesion and rigidity. The adolescence of a snowflake is thus the phase of transformation, called névé, which leads to its adult state: ice, an impermeable, coherent, and more rigid material. Let’s note that on the surface of temperate glaciers summer melting activates this stage of transformation and prematurely creates layers of ice as a result of the refreezing of the melting water. These layers of transparent refrozen crystals that also form in marginal and lower regions of the polar ice sheets are rather easy to identify, and their presence in greater or lesser amounts is likely to serve as a climatic indicator there.
So glaciers and ice sheets are fed through time with the snow deposited on their surfaces. The masses of ice thus formed have a tendency to grow, but their growth is regulated by different mechanisms. As we already mentioned, except at very high altitude, glaciers, as well as the more marginal and lower-altitude parts of ice sheets, are subjected to summer melting. Furthermore, ice can be deformed under the weight of the layers that accumulate over time. This plasticity of the ice is at the origin of the relatively rapid movement of glaciers and of the much slower movement of the ice of Antarctica or Greenland. The displacement of the ice is all the more pronounced when the temperature is close to the point of fusion (0°C on the surface, which lowers to −2°C under 4,000 meters of ice).
In the case of mountain glaciers, the flow follows the direction of the slope and the ablation of the ice due to summer melting increases with the loss of altitude. In some extreme conditions, the glacier can even break apart and spread into the valley, which can cause massive damage (we call this a glacier surge). The final destiny of the ice of large ice sheets is to melt either at the base under the effect of the heat emitted by the Earth at the level of the bedrock that is trapped at the interface with the ice under the effect of thermal isolation caused by the great thickness of the ice, or after floating for some time on the ocean in the form of an iceberg. The large ice sheets are calved by emissary glaciers, some of which flow straight into the ocean. In other cases, these emissary glaciers, true rivers of ice, feed the ice shelves some hundreds of meters thick and which in the ocean disintegrate, forming huge icebergs whose size, from time to time, inspires media attention. In either case, it takes a snowflake born in the center of Antarctica about a million years or more before it disappears into the ocean. In fact, the only ones that escape that ultimate fate are those that are taken by glaciologists in core samples, whose story we will now tell. But let’s begin by explaining how we make this piece of ice “talk” when it arrives in our laboratories and how its age is determined.
The Ice and Its Isotopes: A Paleothermometer
The colder the temperature, the weaker the isotopic content of the snow. There is only one step between this observation, which, as we have mentioned, can indeed be verified when one travels on the surface of Antarctica and Greenland, and the interpretation of analyses of deuterium or oxygen 18 done on samples taken from an ice core at great depth. We have good reasons to think that this relationship between temperature and isotopic content is also true at a given site and that it suffices to analyze the deuterium or oxygen 18 content along an ice core to deduce the temperature at the moment when the snow was formed.
We must be cautious, however, and the best way to do this is to try to understand how water isotopes, those HDO and H218O molecules, behave between the time they evaporate on the surface of the ocean and when they arrive on the surface of polar ice sheets. To do this, we are aided by what we call isotopic models. Some are simple; they describe what occurs in a given mass of air during its travels from the ocean to polar ice caps.1 Others are much more complex since they involve following these isotopic molecules in models of general circulation of the atmosphere, the very ones that are used to predict the temperature we will have in a few days or what the climate will be in a few decades.2
These models tell us that the intuition of the glaciologist, which consists of applying what he observes on the surface snow to deep ice, is justified but only under certain conditions; we will point out the two most important ones. First, it is necessary that the temperature and humidity conditions in the ocean regions from which the polar snow comes not be modified throughout time, quite simply because the conditions that rule in those “source” regions influence the amount of deuterium and oxygen 18 of the precipitation throughout the trajectory of the air masses. Second, if the proportion between the summer snow (rich in isotopes) and winter snow (poorer in isotopes) is modified in comparison to the current climate, the “isotopic thermometer” is skewed: if the proportion of summer snow has increased, it indicates a mean temperature that is warmer than it actually was.
This skewing appears to have only a marginal influence in Antarctica because the proportion between summer snows and winter snows, what we call the seasonality of precipitations, has probably not varied between the Last Glacial Maximum and the current climate, quite simply because the geography and the topography of this ice sheet have not been greatly modified between these two periods. This is not the case for Greenland, which, 20,000 years ago, was flanked by the enormou
s Laurentide ice sheet, whose volume was greater than that of Antarctica. The result: the seasonality of the precipitations was completely modified. It is impossible then to apply this notion of isotopic paleothermometer without complementary information; we will see that the measurement of the temperature in the bore hole and the isotopic analysis of air bubbles provide further information.
For that which concerns any distortion connected to the origin of precipitations we are also fortunate, since we know water has two isotopes with slightly different behaviors. Both can indifferently serve as paleothermometers, but in comparing them minutely we can evaluate how the “source” regions have changed throughout time and achieve a correct evaluation of the temperature at the coring site.3 A correction, as we will see, above all necessary for Greenland.
Impurities with Multiple Sources
Apart from a few layers of ash corresponding to very intense volcanic eruptions, the observation of polar ice, to the naked eye, reveals no presence of impurities. Ice is one of the purest types of matter we know, in general on the order of a gram of impurity per ton of ice in Antarctica, more in Greenland, which is closer to continental sources. And yet the atmosphere carries dust, marine salt, and various chemical components that mix together in the snow and end up trapped in the ice. These impurities can be introduced directly into the atmosphere or be produced in it by phenomena of oxidation implied in the cycles of sulfur, nitrogen, and carbon, and those of halogen, chloride, fluoride, bromide, and iodine compounds. With their multiple sources, the impurities present in the ice, which we now know how to detect at extremely weak levels of concentration, are a unique mine of information. We will return to this, but let’s first point out what we can identify in the polar ice.
Volcanic eruptions can also produce large quantities of gaseous compounds, compounds of sulfur in particular, which are transformed into sulfuric acid in the atmosphere, transported over long distances, integrated into precipitation, and, in polar regions, produce layers of more “acidic” snow, then ice. They are easily found through their electric properties, which can be systematically and continuously measured along the deep ice cores. The analysis of their chemical composition enables a description of the corresponding eruptions, some of which produce large quantities of hydrochloric or hydrofluoric acid.
In desert or semidesert regions the wind, especially when it is violent, raises dust, the finest of which can be transported as far as Antarctica or Greenland. The isotopic analysis of elements such as rubidium and neodymium makes it possible to detect their origins because the great deserts have their own unique isotopic signatures. Great forest fires also provide quantities of aerosols and chemical elements, formate, and ammonium acetate, traces of which we find in the ice of Greenland where certain pollens can also be transported, which is not the case for Antarctica. The ocean is also a great purveyor of impurities for polar ice, marine salt foremost but also sulfur compounds. These compounds are also produced by continental surfaces and by volcanoes, and here, too, isotopic analysis can be a precious help in pinpointing their origins.
These various sources—volcanic, continental, and oceanic—contribute to the complexity of the chemistry of the atmosphere, a true factory that transforms these elements but also produces them, as in the case of storms (nitrogen compounds). The chemist of ice, a glaciochemist, is at the end of the chain, which is an advantageous position for deciphering the atmospheric chemistry in its entirety and the way in which it has been affected by climatic changes. But he or she must first understand well the chemical processes that take place in the névé, which, in a certain number of cases—such as that in which oxygenated water is present on the surface but disappears in the depths—cause the chemical composition of the ice not to reflect exactly that of the polar atmosphere.
Human activity has added to the complexity of the chemistry of the atmosphere. We will return to this point. Traces of such activity are well noted in Greenland, which is close to anthropogenic sources of lead and other heavy metals, sulfates, nitrates, and many other substances that bear witness to human activity, and we can see in the ice that these substances have increased the oxidation capacity of the atmosphere. The Antarctic snow is not exempt from these traces of pollution: we have even found traces of radioactive elements produced during atmospheric nuclear testing in the Northern Hemisphere.
Ultimately the ice caps and ice sheets are not sheltered from what falls onto them from the sky. The interaction of cosmic rays with the upper layers of the atmosphere forms elements, which we call cosmogenic isotopes, other than carbon 14. Thus in ice we find beryllium 10, attached to aerosols, and chloride 36, in the form of hydrochloric acid; the applications based on an analysis of these elements are numerous. Antarctica and Greenland are also favored places of hunters of micrometeorites, even if they must melt enormous quantities of ice to extract usable information from it; zones with very weak accumulation are a priori the most favorable for this hunt, but nature does things well, since in certain coastal zones, where there is so-called blue ice, there is fusion then refreezing and thus a natural concentration of micrometeorites.
Air Bubbles in the Ice: A Very Beautiful Story
The initial meters of the firn are very porous, and the air circulates by convection under the effect of surface winds. Deeper down, at the bottom of the layers of firn, the air is completely isolated from exchanges with the exterior and is then imprisoned in the ice in the form of bubbles (Figure 5.1). This process takes a few dozen years for relatively warm sites with high accumulation and as many as several thousand years for those in the central regions of Antarctica. It is within this range of time that the air bubbles are younger than the ice that encloses them.
After the bubbles are enclosed they become smaller and the air pressure increases with depth. As we go deeper, the bubbles gradually disappear, to the point of no longer being observable under several hundred meters of ice: air molecules are then imprisoned inside the ice matrix in the form of air hydrates (also called clathrates). The ice, without its bubbles, becomes perfectly translucent, even though the volume and the composition of the air it encloses are unchanged in relation to the initial bubbles.
It is an old idea, perhaps first expressed by Per Scholander in the 1950s, to try to extract the air from the past from these little bubbles that escape by the thousands when one plunges a “polar” ice cube into one’s whiskey (an observation that was at the origin of our quest for fossil air that began in the 1960s!). In the 1960s and 1970s two teams, one from France, directed by Claude Lorius, and the other from Switzerland, directed at the time by Hans Oeschger, developed analytical methods to reconstitute the past variations in the concentration of carbon dioxide in the atmosphere from air bubbles trapped in polar ice. The teams’ objectives were to extend back in time the carbon dioxide record initiated by Charles Keeling a few years earlier from atmospheric measurements and to explore the hypothesis proposed seventy years earlier by the Swedish scientist Svante Arrhenius regarding the origin of glacial periods. This story deserves to be told.
Figure 5.1. Air bubbles trapped in the ice. © CNRS Photothèque/Volodia Lipenkov.
In April 1896 Arrhenius, a physics professor on the faculty of Stockholm University, a private scientific institution, published an article titled “On the Influence of Carbonic Acid in the Air upon the Temperature of the Ground” in the London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science. This work was truly important, but it was overlooked for a long time in the study of what today is known as the “greenhouse effect.” It was the Frenchman Joseph Fourier who, in his “Remarques générales sur les températures du globe terrestre et des espaces planétaires,” presented in 1824 in the annals of the Académie royale des sciences of the Institut de France, was the pioneer of this concept, which constitutes a major challenge for the future of the environment of our planet. He spoke at that time of the luminous heat received from the Sun by the surface of the Earth and of the obscure heat reflected
by the Earth, which, in going back through the atmosphere with more difficulty, contributes to global warming. But it was Arrhenius who was the first to evaluate the influence of a change of atmospheric content in carbon dioxide on the surface temperature of the Earth. How can we not admire him when we know that the calculations carried out a bit more than a century ago by that great scientist—he received the Nobel Prize in 1903, but for his work in chemistry—led to an estimate of the warming of the Earth’s surface similar to what the community of researchers has provided at the beginning of the twenty-first century with the help of computers, which are among the most powerful available! Luck, genius….how can we explain this coincidence?
Arrhenius had a career as a physicist-chemist behind him when, at the beginning of the 1890s, he founded the Stockholm Physics Society, which quickly became a mecca for interdisciplinary exchanges. It was there that in 1893 he attended a lecture by Nils Elkohlm from the Swedish Meteorology Office on the origin of glaciations. A discussion took place during which the participants granted only little credit to the influence of the variations of the Earth’s orbit or to that of the displacement of the poles on the surface of the planet. Then, the following year, he followed attended a lecture given by one of his colleagues, Arvid Högbom, on the reasons why the content of atmospheric carbon dioxide could vary throughout geological periods. It was probably through the effect of these two debates that Arrhenius undertook what he called “tedious calculations” regarding the causes of the existence of the ice ages, with the idea of linking climatic changes to those of CO2 over the long term. In one year—a surprisingly short amount of time—he conceptually developed the first approach based on empirical data to what today is called the modeling of the evolution of the climate in response to a change in atmospheric CO2. He thus calculated what must have been the variation in average temperature depending on the season and the latitude for various contents of atmospheric CO2.
