Frozen Earth: The Once and Future Story of Ice Ages

by Doug Macdougall

  In chapter 7, I briefly discussed the crucial role that data from deep-sea cores played in confirming the connection between ice age climate and variations in the Earth’s orbit around the sun, as originally proposed by Croll and Milankovitch. Coring the ocean floor is one of the most effective ways to unravel the Earth’s past environment, and it is worth making a slight diversion here to explore the way in which this important technology has developed.

  The concept is an old one. James Croll was probably the first person to suggest that clues to the history of ice age climates might be neatly stored away in the sediments of the ocean; he thought that the remains of land plants and animals would be washed into the sea and preserved, layer by layer, and that the climate changes of the ice ages would be reflected in the types of plants and animals buried in the sediments. Very little was known about ocean sediments when Croll wrote about ice age climates in the 1860s, but that was soon to change. A decade later a major expedition devoted to scientific exploration of the oceans was conducted aboard the vessel H.M.S. Challenger. The ship left England just before Christmas in 1872, embarking on a three-year voyage that traversed the world’s oceans and made landfall on all continents, including Antarctica. Although the Challenger was part of the British Admiralty’s fleet, and the voyage was subsidized by the Navy, the expedition itself was conceived by Charles Wyville Thompson, professor of natural history at the University of Edinburgh, and facilitated by the Royal Society of London. And while the British Navy undoubtedly benefited from its investment—the voyage revolutionized our understanding of the oceans—it was also a wonderful example of the creative use of military resources. The voyage of the Challenger is generally acknowledged to have been the first truly scientific oceanographic expedition, and in many ways, it marked the beginning of the modern science of oceanography.

  The scientists on board the Challenger made meteorological observations, measured ocean currents, recorded seawater temperatures, collected biological specimens, and took water samples for chemical analysis. They also lowered dredges to the seafloor at regular intervals in order to collect bottom sediments—with the aid of some 270 kilometers of Italian hemp rope taken along for the purpose. One of the multitude of important discoveries they made was that in much of the deep ocean, far from the influence of continents, the sediment on the seafloor is a very fine grained ooze made up primarily of the shells and skeletons of tiny, surface-dwelling organisms—the plankton. Croll’s idea that deep-ocean sediments would contain the fossils of plants and animals washed in from the continents was wrong. But the actual situation was even better, although this realization would come only gradually. Many different species comprise the plankton. Their life cycles are typically measured in days, and both the mix of species and the chemical composition of their shells respond rapidly to changes in the properties of the surface water. The continuous rain of dead plankton down to the seafloor, and their accumulation there, provides an unparalleled record of conditions at the ocean surface, stored in the layers of sediment. However, the dredges used on the Challenger expedition simply scooped up the muddy ooze from the bottom, mixing the layers together in the process and destroying the most important aspect of these sediments—the dimension of time that is preserved in them. The challenge for oceanographers would be to devise a way to sample the layers of deep-sea sediments without disturbing the original sequence of deposition.

  As the science of oceanography evolved, the pressure to find a way to collect bottom sediments without distorting their vertical structure increased. On land it was routine to examine sequences of sedimentary rock layers in order to learn about the geologic history of a region. But geologists couldn’t (at least in those early days) visit the deep seafloor for direct observations; sediment samples would have to be brought to the surface in their original configuration. Early attempts were predictably crude. Simply dropping an open pipe into the seafloor ooze would sometimes work, but there were multiple problems: How do you ensure that the pipe goes into the sediment vertically? How can you prevent the sediment from simply falling out of the pipe as you haul it up to the surface? In deep water, how do you know when the pipe is nearing the seafloor? How can you drive the pipe further into the sediments to obtain a longer core? With trial and error, and characteristic ingenuity, oceanographers gradually improved the design and use of coring apparatus. Fins and other stabilizing devices were added to the pipes to keep them upright as they fell through the sea like guided missiles. “Core catchers” were added, to prevent the sediment from falling out of the corers. Lead weights and sharpened, tapered nosecones were fitted to help the core barrels penetrate more deeply.

  Most of the early devices were capable of retrieving cores that were only one or two meters long. Although sediments accumulate slowly over much of the ocean floor, often at rates of only a few centimeters every thousand years, two meters of core still doesn’t represent a very large interval of geologic time—generally less than 100,000 years. Nevertheless, the early cores provided valuable insights. In the 1930s, a series of short cores recovered from the Atlantic showed that the mix of plankton species changes systematically with depth in the sediments. Based on what was then known about the conditions under which the different species grow, it was concluded that the changes probably reflected changes in surface water temperature—alternating periods of warmth and cold. This insight was a tantalizing foretaste of how sediment cores might shed light on ice age climates.

  As the technology for retrieving sediment cores improved, so did the length of the cores, and therefore the timespan they represented. In the 1940s, a Swedish oceanographer named Börje Kullenberg made a breakthrough. He put a piston in the traditional core pipe and designed it so that the piston would be drawn upward as the pipe penetrated into the sediments, effectively pulling the sediment into the core barrel. Although there have been some modifications, his basic design is still in use today. The original version of the Kullenberg piston corer was deployed at regular intervals during the Swedish Deep Sea Expedition of 1947–49, which, like the Challenger voyage three-quarters of a century earlier, circumnavigated the globe. Kullenberg and the other scientists aboard the ship Albatross were ecstatic when the new device recovered cores up to 15 meters in length. In all they collected some 200 sediment cores; lined up end to end, they amounted to a ribbon of sediment one and a half kilometers long. Fulfilling Croll’s dream, some of the cores they retrieved represented nearly a million years of sediment accumulation.

  Aboard the Albatross in 1947 and 1948 was Gustaf Arrhenius, a young geologist from a prominent Swedish family. His grandfather, the chemist Svante Arrhenius, had been the first Swedish recipient of the Nobel Prize. Svante Arrhenius’s interests were wide-ranging, and among his many accomplishments, he was the first to suggest that changes in the CO2 content of the atmosphere could affect climate through carbon dioxide’s ability to trap solar heat. Fittingly enough, it was his grandson Gustaf who discovered features in the long sediment cores taken with the Kullenberg corer that are directly linked to CO2 changes in the ice age atmosphere. Examining cores from the Pacific Ocean, Gustaf Arrhenius found that the sediment layers were alternately rich and poor in calcium carbonate. The changes were quite regular, and Arrhenius concluded that they were somehow connected to glacial-interglacial climate cycles. Subsequent work has proven he was right about this—dating has shown that the timing of the sediment cycles corresponds well with glacier advance and retreat on land—although his interpretation, that they were due to variations in the intensity of ocean water circulation in the Pacific, turned out to be incorrect. The sediment cycles were later linked to large changes in atmospheric CO2 that accompanied the Pleistocene glacial-interglacial changes. So interdependent are the atmosphere and ocean that changes in one invariably cause changes in the other, and this is particularly true of CO2. As is the case with all gases, the amount of carbon dioxide that dissolves in seawater depends on its concentration in the atmosphere. And the amount of calciu
m carbonate that accumulates in deep-sea sediments in turn depends on just how much dissolved carbon dioxide there is in the oceans. Alternating layers of carbonate-rich and carbonate-poor deep-sea sediments would be expected if CO2 in the atmosphere rose and fell through the glacial cycles. As we shall see, gas bubbles trapped in Greenland and Antarctic ice show that this is exactly what occurred.

  The ten-meter cores collected during the Swedish Deep Sea Expedition were at the cutting edge of technology in their day, and they provided much information about ice age climate fluctuations, but they are a far cry from the several-kilometers-long cores that can be retrieved with present-day techniques. Today, drilling is the preferred method for obtaining long cores. Each metal core barrel that penetrates the seafloor has a plastic liner that fills with sediment as the drilling proceeds; when the cores are pulled up to the surface, the liners can be quickly and efficiently pulled from the barrel without distorting the layers. Typically, the cores are sliced in half lengthwise, and one portion put away as an archive, to be used, if at all, only when all other material is gone or new methods of analysis are developed.

  The most ambitious and successful ocean sediment drilling program is the Ocean Drilling Project (ODP), the current incarnation of a program that has operated continuously since 1968. It utilizes a converted petroleum drillship, the JOIDES Resolution, which circles the globe taking cores of the seabed for scientific research. A massive drilling tower the height of a twenty-story building sits astride the Resolution, over a large hole—the “moon pool”—that penetrates the middle of the ship’s hull, through which the drill string can be lowered directly into the sea. The Resolution is jammed with labs and computers and fitted with a bevy of thrusters, computer-controlled propellers that can keep it stationary at a drilling site even in heavy weather. Lowering the drilling rig to a target on the seafloor several kilometers below the ship is no mean feat, but together the Resolution and its predecessor in the project, the Glomar Challenger, have drilled into the ocean bottom at many hundreds of sites, in all the world’s oceans, since the program began. Originally solely an American endeavor and later expanded to include international partners, the drilling program has been a notable success story among large-scale scientific projects. Drilling goes on around the clock. A typical “leg” of the Resolution’s continuing expedition lasts about two months and is usually focused on a particular scientific question in a specific part of the ocean. Between legs, there are short port calls for refueling, restocking, and exchange of scientific crews, which are usually international in character and eclectic in scientific background and expertise. One of the important goals of the drilling project is to unravel the details of the Pleistocene Ice Age through the sedimentary record of glacial-interglacial cycles. The availability of cores that extend back well beyond the inception of the current ice age also makes it possible to examine climate changes at even earlier times. Although the seafloor is much younger than most parts of the continents, its oldest sediments date back to about 200 million years (older seafloor than this has been destroyed in the great plate tectonic cycle of seafloor creation at the ocean ridges and consumption at ocean trenches). Having a 200-million-year window into climate and environmental change means that a considerable slice of Earth history can be examined in great detail.

  Studies of ocean sediment cores, especially the comprehensive collection of materials from the ODP, show that the seafloor is far from the quiet, peaceful place it was once assumed to be—a sheltered library with page upon page of sediments neatly stacked up and waiting to be read. Instead, it has been found that bottom currents sweep through some parts of the seafloor, picking up sediments in one place and dumping them in another; that gravity works just as well at sea as it does on land, sending muddy avalanches down slopes after earthquakes or other disturbances; and that even in the deepest ocean there are living creatures that dig and burrow and churn up the sediments, blurring the layer-by-layer record. All of these processes complicate the interpretation of sediment core properties. But in spite of such difficulties, the amount of information that has been gleaned about the Earth’s past climate is staggering. A good example is the record of ocean temperatures over the past 65 million years shown in figure 20. The data come from analyses of many different sediment cores, most of them from the Ocean Drilling Project. The temperature data are based on measurements of oxygen isotopes in the shells of small animals that live in the deep sea, and the timescale is largely based on interpretation of the magnetic properties of the cores. Both of these approaches were discussed in chapter 7. What is quite remarkable is that we are able to trace the trends in the Earth’s climate over such long time periods with some confidence. The changing temperatures of the deep sea deduced from the sediment cores are believed to be a generalized reflection of changes in the average surface temperatures on our planet. It is clear from the figure that there has been a persistent temperature decrease from about 55 million years ago to the present, and also that there have been a few intervals with very rapid decreases—and some with increases—along the way.

  Figure 20.This qualitative graph, based on oxygen isotope analyses of fossils from deep-sea sediment cores, shows how ocean water temperatures—and presumably also average Earth surface temperatures—have varied over the past 60 million years. It is obvious that there has been a gradual decrease since about 55 million years ago, with especially sharp drops between 40 and 35 million years ago, and again during the past few million years. These are the times when glaciation began in the Antarctic and the Northern Hemisphere respectively.

  Parameters such as the oxygen isotope composition of shells are generally referred to by geochemists as proxy indicators, or simply proxies, because they don’t measure temperature or climate change directly. Rather, a proxy has to be translated into some environmental variable of interest through an understanding of how the proxy itself responds More and more proxies, many of them chemical and most of them measured in ocean sediments, are being added to the geochemists’ arsenals. Nevertheless, oxygen isotopes remain one of the most valuable and informative proxies available to us. Even to a skeptic, the pattern of 100,000-year cycles in oxygen isotope compositions that characterizes the past million years or so, and that shows up again and again in cores from throughout the world’s oceans (figure 16), is striking. It is hard to believe that the correspondence of these variations with the 100,000-year cycles of eccentricity of the Earth’s orbit around the sun is purely coincidental. Nevertheless, as implied in chapter 7, interpretation of these changes is not entirely straightforward, because at least two separate factors influence the oxygen isotope composition of shells. One is the temperature at which the shells grew, something that is reasonably well understood, because it has been calibrated by both laboratory studies and measurements of samples from nature. The second is the oxygen isotopic composition of the seawater, which depends on the amount of glacial ice that existed on the continents when the shells grew. Why this should be so may not seem obvious at first, but the reason is quite straightforward. Evaporation from the oceans preferentially causes one of the oxygen isotopes to be enriched in the water vapor, leaving behind liquid water that is depleted in that same isotope. The oxygen isotopic composition of the ocean is therefore changed by this process. If the evaporated water is precipitated at high latitudes and ends up as glacial ice, the ocean’s isotopic composition remains changed until the ice melts and the water again returns to the sea. The larger the volume of the ice on the continents, the bigger the shift in the ocean.

  Fortunately, both colder temperatures and larger ice volumes change the oxygen isotopes in the same way, so the overall effect is to amplify the glacial-interglacial variability in the oxygen isotope record. This is fine for getting a general picture of the climate variations, but it would also be useful to disentangle the two effects. Recently, some exciting progress has been made on this problem. The approach has been to use a newly developed proxy that is independent of the oxygen iso
tope variations to determine past seawater temperature. With this knowledge, the expected temperature component can be subtracted from the oxygen data, leaving a residual record that should be due only to changes in the volume of continental ice sheets. Two very interesting insights have emerged from this procedure. First, it appears that ocean surface water in the tropical Pacific warmed up by 3–4°C during the most recent deglaciation, a much larger increase than earlier data had suggested (this also means that the tropical ocean had been much cooler during the glacial interval than earlier suspected). Secondly, the data suggest that there was a lag of two to three thousand years between the temperature increase and the decrease in ice volume. This is not very surprising—think about how long it takes for a bag of ice cubes to melt, or an old-fashioned refrigerator to defrost, even when the ambient temperature is far above freezing. But intuitive or not, this information could not have been gleaned from the oxygen isotope data alone—the simultaneous use of more than one proxy was the critical step. The same approach applied to the previous deglaciation, about 120,000 years ago, gives similar results. This new knowledge paints a detailed picture of how the glacial periods ended, not just when, and it suggests that warming in the tropics plays a crucial role in ushering in an interglacial interval.