Milankovich had returned home after the war with his manuscript, and he published it in 1920 under the title Théorie mathématique des phénomènes thermiques produits par la radiation solaire.19 By his own account, the work had received “a lukewarm response,” but his Paris publisher had sent out copies to a variety of scientific institutions and “to some known scientists” to promote it.20 One of these scientists was Wladimir Köppen.
Milankovich’s theory concerns the variation in solar radiation reaching the surface of Earth: day and night, season to season, year to year. To explain long-term variation in solar radiation, he had settled, after a number of tries, on three controlling astronomical elements, all known to fluctuate on different timescales above 10,000 years. These were the precession of the equinoxes (whereby Earth’s axis spins like a top completing one revolution every 26,000 years), then a much longer (96,000-year) cycle of the changing ellipticity of Earth’s orbit, and finally a 41,000-year cycle in the tilt of Earth’s axis. These three parameters each caused a different variation in solar radiation arriving on Earth, and thus they could either reinforce each other or cancel each other out. Milankovich had the astronomical data for these changes, which had been published by Ludwig Pilgrim (1849–1927), but he had to work out the thermal consequences of the interactions and develop the mathematics to do this successfully in a general way. The result of his calculations was a theory of Quaternary ice ages based on these fluctuating parameters over the past 130,000 years.
In September 1922, Milankovich received a letter from Köppen, who told him that he had read the book with interest and that he believed his theory to be correct and able to explain the main features of the climate changes in the Quaternary period. Köppen also informed Milankovich that he was now working with his son-in-law, Alfred Wegener, on a major project in paleoclimatology. Köppen had long been an advocate of astronomical “forcing” (as we now say) of meteorological phenomena, and he had worked long and hard on various ways to correlate weather with the eleven-year sunspot cycle known as the “Maunder Minimum.”21
Köppen went much further than to simply praise Milankovich for his theory. He asked Milankovich if he would be willing to collaborate with them. Köppen wanted him to calculate the intensity of sunlight and its long-term changes in the Northern Hemisphere for the past 650,000 years. He would not have to take the effect of Earth’s atmosphere into consideration, but Köppen wanted separate calculations for latitudes 55°, 60°, and 65° north. Köppen also told him that in his calculations he should concentrate on summer rather than winter sunlight, as he and Wegener were convinced (as are scientists today) that the key to ice ages is not colder winters but cooler summers that allow the persistence of ice and snow cover from year to year and accelerate the spread of ice by progressively reflecting more incident sunlight back into space.22 Köppen asked him to produce a graph of the summer intensity of sunlight, as well as the amplitude and frequency of its variation, for sixty-five 10,000-year intervals: Köppen and Wegener would publish the resulting graph and attendant data in their planned book, and they wanted Milankovich to publish his mathematical apparatus and calculations separately, so that they might reference his publication.23
Milankovich agreed and went to work immediately to develop the attendant mathematics. By 13 November he had developed the equations to determine the fluctuations in solar radiation at the specified latitudes. He and Köppen exchanged ideas about the best way to graph this material so that it would clearly demonstrate the timescale, the amplitude, and the frequency. Milankovich’s calculations showed that the oscillations in summer sunlight over long periods of time are not regular in their frequency or their amplitude, and he concluded that this irregularity was indeed caused by his three astronomical parameters, all of which were the result of the gravitational interaction of the Sun, the Moon, and the large planets. He then computed the values; the calculations took 100 days, and he finished them sometime in late February 1923.24
Enlisting Milankovich in this way emphasized the collaborative character of the planned work and the clear partition of authorship. Köppen and Wegener had decided that Wegener was to be responsible for assessing climate change from the Carboniferous and Permian through the Tertiary; Köppen would be in charge of the material on the Quaternary. Milankovich’s results and theory would be restricted to Köppen’s portion of the book.
Not only would the assignments differ, but also the approach. Wegener’s aim, or the aim of Wegener’s chapters, would be to establish the “Climate Belts” (Klimagürtel) for each of the major divisions of geological time, based on the geological and paleontological evidence, and then prepare a map of the results for each period. The map projection would be the oval “Hammer” view of Earth which Wegener had used in the third edition, and the conventions would be the same: the parallels of latitude and meridian of longitude would be those of the modern world, with Africa in its current position. The maps for successive geological periods would show the continuous dispersal of continents relative to Africa, and (very much to the point) the map would also show, as curved lines, the positions (relative to the present) of the equator, the lines of 30° and 60° latitude, and also the poles—the Klimagürtel for each period.
Köppen’s aim, or the aim of his long chapter on the Quaternary, would be more detailed, in keeping with the greater volume and variety of evidence for this recent period. Rather than establishing “Climate Belts,” he would produce a summary of the climate evidence and a variety of maps of the extent of the ice as it oscillated back and forth over the past half-million years. He would correlate this evidence with the theory of Milankovich and map the displacement of the poles throughout the Quaternary. Milankovich’s work would take up perhaps one-fourth of the part of the book allotted to Köppen, and it would be restricted to the Northern Hemisphere. The chapter title, “Die Klimate der Quartärs” (plural), indicates the concentration on the frequent climate oscillations within the framework of the ice age.
Their focus on the Quaternary on the one hand and the importance of the Permo-Carboniferous glaciation for Wegener’s displacement theory on the other, along with the possibility of even earlier ice ages in the Algonkian (Achaean), led Köppen and Wegener to believe they could safely assume that in all periods of Earth history “the same climate zones as today’s have existed, namely, an equatorial rain zone, two arid zones, two rain-zones in temperate latitudes,” in turn flanked by polar climates beyond them.25 They also assumed that just as on the present Earth, the poles would lie 90° away from the equatorial rain zone and 60° away from the nearest arid zone. Their work would also reflect certain other conventions of Köppen’s climatology concerning atmosphere and ocean circulation, as well as the tendency of dry zones to be interrupted on the eastern margin of continents by monsoon rains.26
If their climate belt scheme were true for the past as well as the present, then they should be able to map climate zones in each period of Earth history by having, for each zone, a uniquely distinguishing marker: something that unambiguously said “polar,” or “arid,” or “temperate rain,” or “tropical.” Most paleogeography and paleoclimate books had used fossil animals and fossil plants as such indicators, but, as Wegener pointed out in the paleoclimate chapter of his third edition, these were “general indicators” requiring large aggregations of data for a distinct signature. Further, the adaptive plasticity of both plants and animals made it difficult to rely on them alone. He suggested instead a greater reliance on geological indicators: desert deposits, coal, and glacial deposits, along with some particular plants (trees with rings, and the famous Glossopteris flora of the Southern Hemisphere). He had drawn a conjectural map using such geological makers and their distributions for the Carboniferous and Permian to show how these helped locate the equator and the poles.27
These classes of geological deposits (and a few fossils) would be for Köppen and Wegener what we now call “proxy” data that would “stand in for” a certain climate. This was a
common practice in paleontology, where the occurrence of a certain fossil species was so reliably an indicator of a specific period of the past that it became an “index fossil.” The same is true today in ecology, where some animal or plant becomes an “indicator species” for a certain ecological assemblage. There existed, in 1922, a huge, painstakingly assembled geological literature of where (on the current surface of Earth) geologists had found and mapped glacial deposits, arid formations, coal, massive limestones, and reef corals. Each of these layers in any given locale was assigned to a time in the past. Geologists (minus a few outliers) agreed that these geological markers unambiguously indicated not only certain kinds of climates but climates associated today with specific latitude zones between the pole and the equator. The key to a narrative history of the climate of the past would be to pull apart the geology books that tell us which rocks are in what place now and rewrite the story to say where they were then. “The climate history of any place on earth is, to a first approximation, the history of its position between the pole and the equator.”28
The question was not whether such a reassembly of the data was possible but the security of the markers. This turns out not to have been a serious issue for most of the categories. There was in the 1920s, as there is now, agreement that salt deposits and gypsum are “evaporites” that can form only when there is an excess of evaporation over precipitation. They are evidence of arid conditions at the time in which they were deposited. There was (and is) agreement on so-called aeolian sandstones: that such formations were once windblown dunes, in which the bedding is so distinctive and characteristic that they may be mapped to show the direction of the prevailing winds at the time they were deposited. There was (and is) agreement that massive limestones from diatoms and reef limestones indicate tropical seas. There was no dispute that the rocks called “tillite,” or “boulder clay,” or Blocklehme—rough, poorly sorted material of every grain size from boulders down to individual plates of clay minerals—were remnants of glacier action.
It is a standard geological problem to identify such sediments and to distinguish them from their imitators. For instance, landslide deposits called “turbidites,” which occur where the landslide takes place underwater—on the continental shelves—give convincing imitations of glacial deposits. Similar pitfalls await the casual reconnaissance of the other kinds of deposits as well. But by 1920 detailed geological mapping of the continental surfaces had already exceeded about 20 percent of the total area, and while a deposit here or there might be misidentified, this would not be likely across dozens or hundreds of locations. Moreover, the figure of 20 percent is misleadingly small: salt, gypsum, limestone, and coal are all economically important minerals for which prospectors (today called “economic geologists”) competed avidly. Investigators were not limited to surface exposures either, as salt, limestone, coal, gypsum, and sandstone could be identified by drill cores in the many thousands of prospecting wells drilled in the first twenty years of the century. The data set was large enough that an occasional error would not falsify the entire picture.
The one climate marker that in 1920 still needed discussion and defense was coal. Almost everyone agreed that coal was the result of the burial of vegetation and imagined that the greatest coal beds, those laid down in the eponymous “Carboniferous,” were the accumulation of huge quantities of vegetable matter in freshwater basins—coal bogs and swamps—where, through deep burial, compression, and heat, they were turned into various forms of coal.
The great paleobotanist Henry Potonié (1857–1913) had devoted a good part of his life and of his major work, Lehrbuch der Paläobotanik, to the theory of the development of coal. Potonié had demonstrated that coal is formed in a sequence humus → peat → lignite → bituminous coal, through burial, compression, and heat. He did this not just by arguing about it or citing geological literature, but by providing photographic tours of modern locales where peat and humus are being formed and buried. He showed that the kinds of plants that are found in coal are resident today in bogs, moors, reed beds, and swamps, and that the process of humus formation, burial, and transformation into combustible coal products can be observed in every stage today. While many coal beds are filled with the fossils of cycads and ferns, others are produced by wet temperate forests and even from marine plants such as kelp beds buried along the shore.29
Potonié had also insisted, contrary to prevailing textbook opinion, that coals could form in the tropics, because coals form where there are bogs, and bogs are demonstrated to exist in Sumatra, Ceylon, Central Africa, and South America; thus, there must be tropical coals.30 The prejudice against this idea was that tropical heat must speed up the rotting of vegetation to the point where coal could not form. From Potonié’s standpoint, as adopted by Köppen and Wegener, “the real contribution of such accumulation [of coal] to the question of climate lies not in the area of temperature, but of humidity. Before a basin can turn into a bog, it has to be full of fresh water, which only happens in the rainy zones of Earth, not in the arid regions. Neither can coals form in the dry zone of the horse latitudes, but only in the equatorial rain zone and in the two temperate zones.”31
For Köppen and Wegener, coal formed both tropically and in temperate humid climates, even up to the postglacial “pluvial” zones at relatively high latitudes. Where such periglacial rain climates were not available to create bogs, they would form in preexisting topographic depressions. Thus, the Appalachian zone, which they would argue was the Carboniferous equator, was a tropical rain zone, where geologic folding created conditions for freshwater basins and coal formation. They argued that all one might infer from coal formation is abundant rainfall, and the temperature of such regions must be indirectly inferred from the fossil evidence, via systems of floral characters.32
The “big picture” of the history of climate was the latitude and rainfall regime for a region of Earth through time. As Franz Lotze (1903–1971) later pointed out in his classic Steinsalze und Kalisalze (Rock salt and potash) (1938), the worldwide deposits of salt and gypsum look randomly spread on all continents from the equator to very high latitudes. However, when we date the deposits by traditional methods of superposition and sequence—the stratigraphic correlation of known formations as age markers across wide areas—we can see that evaporite deposits for any period of Earth history form distinct zonal bands similar in width to that defining the limits of the evaporates today, that is, areas with a net excess of evaporation over precipitation. Moreover, these bands, when mapped on a globe, migrate through time from current polar latitudes (Carboniferous) to current equatorial latitudes (within 30° of the equator). Thus, either the whole planet was arid and dry in certain periods, allowing evaporites to form at the poles or very near them, or the latitudes have shifted.33 The content of this judgment by Lotze is the more important because he was at the time an ardent opponent of Wegener’s theory of displacements.
As with coal and evaporates, so it was also for sandstones and fossil sand dunes, as well as the history of deserts. Friedrich Solger’s Dünenbuch (1910) was a survey of everything that was known about sand dunes and dune formation, including bedding structure, shape relative to the prevailing wind, and relationship of dunes to the coastline at the time of formation.34 Solger asserted that, assuming that global atmospheric circulation remained constant in terms of the location of polar easterlies and the prevailing westerlies, fossil sand dunes could help to locate the latitude of the deposit when formed.35
Even more detailed support of the relationship between latitude and climate came from Johannes Walther (1860–1937), whose Gesetz der Wüstenbildung (The law of desert formation) (1912) was a major treatise on the character, mode of formation, and history of deserts throughout geologic history. The balance of evidence led him to conclude that relative climate change must take place through the displacement of Earth’s pole—moving the equator and all flanking zones. There might be, he argued, absolute climate change—passage through different
regions of space, variation in solar radiation—but the unambiguous signatures of past climates are left respectively by glacial climates, rainy climates, and desert climates, each of which has characteristic geomorphology, weathering, and sedimentation, including characteristic deposits.36
Of all their proposed markers, best established by 1920 was the association of “glacial till” (Blocklehme) with the presence of large-scale glacier action. Such deposits had been mapped all over Earth, everywhere associated with continental glaciation in both the recent and distant past. Wegener’s 1921 opponent concerning the displacement hypothesis, Albrecht Penck, was the world leader in establishing this particular geological deposit as the signature of an ice age and in establishing a chronology of the European ice ages based on the appearance and disappearance of these deposits in the sedimentary record over the past half-million to one million years.37
All the works cited above constituted “proof of principle” for Köppen and Wegener in the preparation of their book; the actual labor on which they were embarked would be vastly more difficult and required great patience and industry. Beyond the general consensus on climate-based latitude zonation of glaciers, coal, and evaporate deposits, the argument of Köppen and Wegener’s book required specific reference to the original field descriptions and maps—many thousands of locales for each of their markers. The few scattered plots on Wegener’s maps for each period of “E” (Eis), “K” (coal), “G” (gypsum), “S” (salt), and “W” (Wüstensandstein [desert sandstone]) in the published work are an immense reduction of information contained in geological maps and field reports.
Alfred Wegener Page 88