The Ocean in the Ice Ages
The ice ages follow the slow changes of the Earth's orbit, and it was two deep-sea cores that proved it in 1976. This unit computes the three orbital cycles and the sunshine they control, reads the spectrum of a core, follows the switch from a 41,000-year to a 100,000-year rhythm, draws the portrait of the glacial ocean at its coldest, and follows the carbon dioxide that the atmosphere lost in every glacial into the deep sea.
The deep-sea cores of Unit 13 showed an ocean that swung between warm and cold dozens of times in the last three million years, with ice sheets that grew and vanished in step. This unit asks what set the beat. The answer, that the slow changes of the Earth’s orbit pace the ice ages, was argued for a century before two cores from the southern Indian Ocean settled it in 1976, and the orbit is the one driver of the climate whose past and future can be calculated exactly, to the year. The unit then draws the portrait of the glacial ocean at its coldest, 21,000 years ago, and follows the carbon dioxide that the atmosphere lost in every glacial into the deep sea, where the search for it is not yet complete.
1 The discovery of the ice ages
Across the lowlands of northern Europe and North America lie boulders the size of houses, made of rock that does not occur for hundreds of kilometres around. Geologists of the early nineteenth century called them erratic boulders and explained them with the biblical flood or with icebergs drifting on it. On 24 July 1837 the Swiss naturalist Louis Agassiz (1807-1873), a professor at Neuchâtel from whom a lecture on fossil fish was expected, opened the annual meeting of the Swiss Society of Natural Sciences with the claim that a sheet of ice had once covered Europe from the North Pole to the Alps, and that the erratic boulders were its luggage. He had taken the idea from two Swiss engineers, Ignaz Venetz and Jean de Charpentier, who had seen the same boulders, polished rock, and ridges of rubble, called moraines, below the living glaciers of the Alps, and his book Études sur les glaciers of 1840 made the ice age a scientific fact within a decade. The question changed from whether there had been an ice age to why there had been one.
The first answer came from astronomy. In 1842 the French mathematician Joseph Adhémar (1797-1862) proposed in his book Révolutions de la mer that the slow turning of the Earth’s axis, which shifts the season of the Earth’s closest approach to the Sun round the calendar, gives one hemisphere the longer winter and an ice sheet with it. The Scottish self-taught scientist James Croll (1821-1890), a janitor at a Glasgow college who learned celestial mechanics from its library, showed in 1864 and in his book Climate and Time of 1875 that this effect grows with the eccentricity, the measure of how far the orbit is stretched from a circle, which Section 2 defines, and he added the feedbacks that are still in every model: snow reflects sunlight and cools the ground further, and a cold hemisphere changes the winds and the ocean currents. Croll put the end of the last ice age at about 80,000 years ago, and when geologists in the 1890s found from the retreat of the Niagara Falls that the ice had left only some 10,000 years earlier, his theory was dropped.
The Serbian engineer and mathematician Milutin Milanković (1879-1958) rebuilt it on firmer ground. Between 1912 and 1941 he computed, by hand, how much sunshine each latitude receives in each season for the last 600,000 years, from the three orbital cycles that Section 2 describes. The German climatologist Wladimir Köppen (1846-1940) supplied the decisive physical argument in 1922: what limits an ice sheet is the summer, because snow that survives a cool summer accumulates, whereas a cold winter adds nothing that a warm summer cannot remove. The quantity to compute was therefore the summer sunshine at the latitudes between 55° and 65° N, where the great ice sheets had their margins, and Köppen and his son-in-law Alfred Wegener (1880-1930) printed Milanković’s first curve in their book Die Klimate der geologischen Vorzeit in 1924. His complete work, the Kanon der Erdbestrahlung, appeared in Belgrade in 1941.
The hypothesis was widely rejected in the 1950s. Geologists had matched the four glaciations known from the Alps to Milanković’s curves, but the curves predicted a dozen ice ages where the land showed four, and the new radiocarbon dates of the moraines did not fit the match either. The land record, not the hypothesis, was the problem, because an advancing glacier erases the moraines of the glacier before it. When Cesare Emiliani (1922-1995) read the complete record from the oxygen isotopes of deep-sea foraminifera in 1955 (Unit 13), he found not four cycles but many, and pointed out that their spacing looked like Milanković’s. Proving it needed cores dated well enough to measure the periods, which took another twenty years.