Frozen Earth: The Once and Future Story of Ice Ages

by Doug Macdougall

  Milankovitch was single-minded about his work. Just as he did not want to end up as an “ordinary” engineer in a provincial city, he also did not want to take on an “everyday problem” for his thesis. Original science was what interested him. And he also decided not to follow the usual course of doctoral students, which was to ask the most distinguished professors to suggest a thesis project. They would, he commented, give him only “the crumbs from their tables.” Instead, he decided to find his own research topic, one that was both interesting and significant.

  The subject he eventually settled on doesn’t, at first glance, seem very glamorous, nor does it have any conceivable connection with ice ages. Milankovitch decided to work on . . . concrete. Although cement and concrete had been around as building materials at least since the Romans, concrete—especially the reinforced variety—was just beginning to see significant use in large structures. Whether or not Milankovitch realized that it would become the most important building material ever, he did feel that this was an area where he could both learn something new and also apply the principles of higher mathematics. Concrete was a material with known properties. He could calculate stresses and pressures and determine the size and shape limits that would be safe for actual structures. Not only would the mathematics be interesting, but the results would also have practical applications. Milankovitch dove into the problem with enthusiasm. It was his first real experience of independent scientific work, and he took to it like a duck to water. At the end of his project, when he had completed his thesis, he wrote in his memoirs: “I would be the happiest man in the world if I could live my whole life as I have this last year.”

  Milankovitch had been right about the value of his investment in education. The expertise in the properties of a new building material that he acquired during his doctoral studies gained him employment with a large engineering firm in Vienna. He was immediately given tasks that involved the design of large concrete structures throughout Europe. In less than half a year, he was appointed chief engineer, and soon he was patenting his ideas and publishing theoretical papers that dealt with problems in concrete construction. He had a comfortable life in a vibrant city, and a successful career as an engineer seemed to stretch out in front of him far into the future.

  However, political events of the early twentieth century intervened. The “Bosnian crisis”—an earlier Bosnian crisis than the one of recent memory—cast long shadows across Austria. The Hapsburg Empire, after secret negotiations with Russia, annexed Bosnia and Herzegovina in 1908. This political maneuvering incited violent reactions from Serbia, which in turn provoked a backlash against people of Slavic origin in Vienna. Although he was not directly affected, the general atmosphere made Milankovitch uncomfortable. In the midst of this crisis, he was offered a position at the University of Belgrade, and, in spite of entreaties from friends and colleagues in Vienna—and the prospect of a drastic drop in salary—he decided to accept. Not only would he be going home, but he would also be able to devote himself to research in an academic setting. In 1909, he arrived in Belgrade to take up the chair of applied mathematics and begin a new career. It was quite different from anything he had done previously—his first task was to set up a complete three-year program for undergraduate students that would cover the various aspects of applied mathematics. Characteristically, he wanted to do things his own way rather than simply repeat the lessons given by his predecessor. He immediately and systematically set about drafting hundreds of pages of lecture notes. Although he had no experience as a teacher, Milankovitch was a good and popular lecturer. “I explained my mathematical formulae slowly and wrote them on the blackboard precisely,” he said. He also interspersed his discussions of formulae with stories about famous scientists and the discoveries they had made, “which the students appreciated because they were not required to be learnt for the exams.” Some things never change.

  In spite of the fact that he had to live much more modestly in Belgrade than he had in Vienna, life as a university professor suited Milankovitch. He gave up his Austro-Hungarian citizenship and became a Serbian citizen. Once he had prepared his lecture notes, he turned to the activity he liked best: solving scientific problems mathematically. This is what he had done as an engineer; now he was free of practical constraints and could focus his talents on most any sphere of science. And although in most respects he was a modest and low-key man, in this sphere he was very ambitious. He was determined to make lasting contributions, and he sought out problems that, as far as he knew, were unsolved. Early on, one experience brought him up short. He had eagerly followed the scientific discussions surrounding Einstein’s then-new theory of relativity, and he submitted for publication a short paper dealing with an aspect of the theory that he could treat mathematically. A short while later, he got the paper back with a note from the editor explaining that three separate papers covering the same ground had already been published by American scientists. Milankovitch had been completely unaware of them. It made him realize that to make an impact in a field, one had to be at the center of it and know the literature intimately. The University of Belgrade, while a good institution with some excellent faculty members, was clearly not a center of cutting-edge science. Milankovitch resolved to find an interdisciplinary area where he could use his expertise in mathematics to address important problems, “an arable field which I could cultivate with my mathematical tools,” as he put it. Meteorology, he recognized, was just such a field. It was interdisciplinary, and nearly all the work being done was empirical. Like Croll before him, Milankovitch was intent on discovering the underlying principles, principles that he could investigate mathematically. In short order, his initial foray into the meteorology literature led him to climate and the problem of ice ages. He aimed very high: his goal was no less than to derive a mathematical theory of the Earth’s climate.

  By 1912, when Milankovitch first began to work on this problem, James Croll’s astronomical theory for the origin of ice ages had been dismissed by most scientists because—as we have already seen—the timing he had predicted for glaciation did not match the estimated ages of glacial deposits very well, especially for the most recent glaciation, about which the most was known. Milankovitch read Croll’s work, which he later called “remarkable.” But initially he approached the problem somewhat differently, not so much from the perspective of finding the cause of ice ages, but rather of understanding climate in general from first principles. He started with the question, Is it possible to calculate, theoretically, what the temperature should be at any place on the Earth’s surface? It was well known that the sun supplied the heat energy responsible for warming the surface, but calculating the actual temperatures would only be possible if he knew the values of all the parameters involved: the amount of energy from the sun that arrives at the top of the atmosphere, how this energy is transmitted through the atmosphere and spread over the Earth’s surface, how much is reflected back into space, and the details of the Earth’s position relative to the sun. The astronomical parameters—the Earth’s position relative to the sun—were the same ones that Croll had investigated: the tilt and wobble of the Earth’s axis of rotation, and its distance from the sun, which is controlled by the eccentricity of its orbit. As far as Milankovitch could tell from an extensive search of the literature, the detailed calculations he envisioned had not been done before. Croll had plotted the variations in the eccentricity of the Earth’s orbit back several million years, and had attempted to show how the Earth’s temperature might have changed in response to these changes, combined with the wobble of the rotational axis, but his calculations had necessarily been somewhat qualitative. He did not have accurate information about the amount of energy received from the sun, nor did he consider how heat energy is transmitted through the atmosphere. Milankovitch had the advantage of all that had been learned about these matters in the half century since Croll had first written on the subject. During that time, the amount of solar energy arriving at the top of the
Earth’s atmosphere had been determined quite precisely, and more accurate and detailed calculations of the variations in the tilt and wobble of the Earth’s axis, as well as the eccentricity of its orbit around the sun, had become available for the period covering the past 1 million years of Earth history. Most of the important parameters that Milankovitch needed for his work were known or could be estimated quite precisely. The stage was set for him to begin his investigation.

  The first major paper Milankovitch published on this subject appeared in 1913. In it, he gave a general account of his calculated theoretical relationship between heat supplied by the sun’s radiation and the temperatures at various points on the Earth, taking into account the effects of the Earth’s tilt and its annual orbit around the sun. It was the first real attempt at a theoretical portrayal of global climate, but it was only a beginning. Milankovitch knew that he had much more to do if he were to produce an accurate mathematical description of the Earth’s climate. But in the summer of 1914, his life—and the lives of many in Europe—was thrown into chaos. Archduke Franz Ferdinand, the heir to the Hapsburg throne, was assassinated in Sarajevo by a Bosnian nationalist, and within a month, war was declared. Milankovitch, as a Serbian citizen, soon found himself in a prisoner-of-war camp in Hungary. He had recently married—the wedding had been just weeks before the assassination—and his new wife tried desperately to have him freed, initially with little success. Finally, she spoke to one of his professors from his days in Vienna, a well-known scientist who had powerful connections in the Austrian government. Milankovitch in prison was regrettable, but when his friend learned that imprisonment prevented him from doing science, he was truly dismayed. That was intolerable. The government was soon persuaded to release Milankovitch, although he had to agree to remain in Budapest and report weekly to the authorities. It was a very civilized arrangement, given the violence of the war, and it allowed Milankovitch to continue his work. He was welcomed at the Hungarian Academy of Sciences, where he toiled away on a massive treatise on climate. The manuscript was completed in Budapest; it was not until 1919, well after the war was over, that he was permitted to return home to Belgrade.

  With a semblance of normalcy restored in Belgrade, Milankovitch could set about publishing his manuscript. It was an expansion of his earlier short papers on climate, a detailed and comprehensive treatment of how solar radiation affects the Earth’s surface temperature at different latitudes and in different seasons. Because his theory didn’t take into account the transport of heat around the globe by the oceans and the atmosphere, his calculated values didn’t match the measured temperatures exactly—they were too low at high latitudes and too high nearer the equator. But averaged over the surface, they were in remarkably good agreement with the true, observed average temperature. This convinced him that his approach was the correct one—and equally important, it persuaded others too.

  Milankovitch had written his manuscript in German, but in the end it had to be published in French. It was not a language in which he was fluent—he had to secure the help of a colleague from the University of Belgrade for the translation—but eventually, in 1920, his Théorie math-ématique des phénomènes thermiques produits par la radiation solaire (A Mathematical Theory of the Thermal Phenomena Produced by Solar Radiation) appeared. Most of Milankovitch’s earlier publications on climate and solar radiation had been short papers written in Serbian, and they had not been widely distributed. To some extent, his book—a thorough treatment of the subject in a language that most scientists in Europe and many in North America could read—corrected this, although it was clearly not a best-seller. Milankovitch himself characterized its reception as “polite but lukewarm.” He was not perturbed, noting that many important discoveries had gone unappreciated for years. If his work was significant, he said philosophically, it “would find its way without any help, recommendation or praise.”

  The book did attract considerable interest among meteorologists, because it was a departure from the traditional empirical approach to temperature and weather. For those who were willing to consider it seriously, it provided a theoretical framework for understanding meteorological observations. But in spite of the fact that it contained calculations of surface temperatures back to 130,000 years before the present, the book did not at first have much of an impact among geologists. A mathematical theory of temperature variations seemed remote from their immediate concerns. In fact, more attention was paid to Milankovitch’s results for temperatures on Mars than to those for the Earth (he had carried out calculations for Venus, Mars, and the moon, in addition to the Earth). For several decades, the popular press had been full of stories of “canals” and sentient life on Mars, but Milankovitch’s calculations predicted temperatures on the red planet far too low for liquid water or any life to exist.

  Eventually, however, geologists and others interested in the Earth’s climate history, especially its ice ages, came to realize the importance of Milankovitch’s work. That appreciation arose through another of those serendipitous circumstances that are common in science and history. One of the meteorologists who read Milankovitch’s book and was much impressed by his theoretical approach was the prominent German scientist Wladimir Köppen, who immediately saw the relevance of the calculations for understanding climate changes. It so happened that Köppen’s daughter had married the geophysicist and arctic explorer Alfred Wegener, best known today as the first person to propose a comprehensive theory of continental drift. When Köppen read Milankovitch’s book, he and Wegener were in the process of preparing a book manuscript on the subject of past climates. Köppen noticed that there was a direct correspondence between the timing of glaciation in the Alps (as far as it was then known) and the Northern Hemisphere temperatures calculated by Milankovitch. He immediately wrote to Milankovitch asking if he would be willing to collaborate with him and Wegener. Thus began a long association, which lasted until Köppen’s death in 1940 and ensured that Milankovitch’s calculations would be taken seriously. It was a fortunate and productive collaboration, because Milankovitch had the mathematical tools, Köppen had the international stature that guaranteed widespread attention to their ideas, and between them Wegener and Köppen had a good understanding of the geological evidence for ice ages.

  Milankovitch’s theory attempted to show how temperatures on the Earth would fluctuate over time in response to astronomical parameters. Like Croll before him, he considered the eccentricity of the Earth’s orbit around the sun, and the wobble of its rotational axis, but he also introduced a new factor—he showed that the amount of tilt of the rotation axis is a far more important factor than had previously been believed. The angle of tilt doesn’t change very much—it varies through a total range of only about two degrees, from roughly one degree less to one degree more than today’s 23.5 degrees—but it markedly affects the contrast between winter and summer temperatures. This is not difficult to understand if you think about it. When the tilt is greater, the sun rises later and sets earlier in the winter hemisphere, and the converse is true for the summer hemisphere. So compared to the present, the summer hemisphere is heated more, the winter hemisphere less. Like the other astronomical parameters, changes in the amount of tilt are regular— they go through a complete cycle every 41,000 years. Milankovitch’s work also revealed another aspect of the solar radiation effects that had not been realized previously. Because he made calculations for a range of latitudes rather than just for the Northern and Southern Hemispheres, as had been done before, he discovered that the various astronomical factors have quite different effects at different latitudes.

  Figure 14.The graph that Milutin Milankovitch sent to Wladimir Köppen shows the “equivalent latitude” for 65° N in summer, based on the amount of solar energy received, for the past 600,000 years. Milankovitch believed that glacial intervals would occur at times when the equivalent latitude was significantly higher than today, indicated by the shaded peaks on this graph.

  Köppen was especially
interested in these nuances in Milankovitch’s calculations. He argued that low summer temperatures in northern regions were the most critical factor for initiating permanent ice cover in Europe and North America—if summers were cool, more of the winter snow would persist without melting, the additional snow cover would cause more sunlight to be reflected back into space, and temperatures would drop further as a result, eventually leading to widespread glaciation. This was a radical departure from the conventional wisdom, the exact opposite of what most workers up to this time—including Croll and many of his supporters—had assumed: that cold winters were the key to the start of glaciation. According to Milankovitch’s calculations, the same orbital conditions that produced cool summers resulted in winters that were slightly warmer than normal. But this did not bother Köppen—the higher winter temperatures would promote evaporation, and the temperatures would still be low enough that the resulting increased precipitation would fall as snow on the growing ice sheets.