Progress in unravelling ice-sheet evolution can be illustrated by the observation that in the 1960s it was assumed that the Antarctic Ice Sheet was a child of the Pleistocene. Now it is believed that extensive ice built up in Antarctica as long as 34Myr ago at the end of the Oligocene and that in the ensuing 14Myr there was a phase of progressive global cooling culminating in a full-bodied Antarctic Ice Sheet in the middle Miocene around 15 Ma. The period of ice build-up began in cool temperate climates with Croll-Milankovitch cyclic glacier fluctuations coinciding first with beech forests and subsequently with tundra vegetation (Naish et al., 2001; Sugden & Denton, 2004). The maximum ice sheet advanced to the outer continental shelf but subsequently, around 13.6Ma, it withdrew to its present extent with limits in East Antarctica approximately on the present coast. The history of the ice sheet since the mid-Miocene to the present has been the centre of debate. Some have argued for a dynamic history of ice sheet growth separated by episodes of massive deglaciation (Webb et al., 1984). Others have argued for essential stability of the ice sheet for over 13Myr with only coastal glaciers and ice shelves responding to changing global sea levels (Denton et al., 1993). The latter view is consistent with a view that the ice sheet has influenced local climate and ocean circulation to such an extent that the system has achieved a new state of stability (Huybrechts, 1993). Warmer temperatures are unable to increase melting significantly since overall temperatures are too far below the melting point; instead, warmer air brings more snowfall and causes glaciers to thicken. The case of the West Antarctic Ice Sheet may differ, because much of it is grounded below sea level and is thus susceptible to fluctuations in global sea level. Although there is no firm evidence that it has disappeared since it too formed in the early Miocene, there has been progressive thinning throughout the past 10,000yr, apparently in response to sea-level rise following the loss of Northern Hemisphere ice sheets at the end of the last glaciation (Conway et al. 1999; Stone et al., 2003).
In spite of the long history of study since the mid-19th century, the form and behaviour of the Northern Hemisphere ice sheets has been the centre of active debate. A landmark publication was The Last Great Ice Sheets in which a minimum and maximum model were put forwards (Denton & Hughes, 1981). This stimulated a new attack on the problem. It now appears that the Laurentide and Greenland ice sheets were extensive and covered North America on a scale equivalent to that of Antarctica today. The Scandinavian Ice Sheet too was extensive, covering the Barents Sea and extending as ice shelves into the Arctic. Ice in eastern Siberia, however, was limited by a lack of snow and glaciers were centred on mountain massifs. Overall, ice build-up, most of which was in the Northern Hemisphere, was equivalent to a lowering of global sea level by 120 m. Although mountain glaciers have been present in the Northern Hemisphere for millions of years, cycles of massive ice-sheet growth and decay every 100,000yr are characteristic only of the past 500,000yr or so.
Elaboration of this story of ice-sheet evolution over long time-scales has raised many challenging questions. What was the distribution of land masses and ocean currents necessary to initiate ice sheet growth in Antarctica? What were the feed-back relationships between the growing ice sheet, ocean circulation and climate that caused ice to grow to the outer continental shelf? Why did it then retreat from its maximum extent to the present coast? Could it be that it eroded its offshore bed to such an extent that, during higher-frequency cycles of growth and decay, it could no longer advance into deep offshore water? Why did massive icesheet growth in the Northern Hemisphere occur only in the past 500,000 yr? What controlled the pace of ice-sheet growth and decay? Given the close correlation between insolation changes in the Northern Hemisphere and global ice-sheet growth and decay, what mechanisms transmit the signal worldwide?
A feature of recent decades has been the appreciation of the extreme variability of climate on time-scales finer than that of ice-sheet growth and decay. The ice-core records show abrupt changes in climate, especially Dansgaard-Oeschger events which represent a sharp warming of up to 10°C followed by a slower phase of cooling that recur every few thousand years (Hammer, this volume, Chapter 78). Such variations have drawn attention to possible mechanisms of change and the importance of establishing records of glacier fluctuations over the globe. Some workers argue that glacier fluctuations, including those on a millennial scale, are synchronous over the world (Denton et al., 1999), thus arguing for a rapidly changing driver of change, such as the levels of carbon dioxide or water vapour in the atmosphere. Others stress that glaciers display antiphase behaviour on a millennial scale, thus arguing for oceanic drivers of change, such as the bipolar seesaw (Broecker, 1998; Clark et al., 2002b). This latter mechanism involves the thermohaline circulation whereby warm conditions in the North Atlantic (as occurs now) draw heat from the Southern Hemisphere oceans. Conversely, a cold North Atlantic suppresses the thermohaline circulation and leads to warmer conditions in the Southern Ocean. Both situations can be driven from conditions in the north or south and imply antiphase behaviour of glaciers on a millennial scale. The debate is still alive and its firm solution, at least for the last glaciation, awaits new dating methods that have a resolution sufficient to unpick fluctuations on time-scales of centuries. In this context, there is much to be gained by the use of different approaches, for example, glacio-isostasy (Fleming, this volume, Chapter 45), or study of more recent glacial episodes, such as the Little Ice Age, where other dating methods are available (McKinzey, this volume, Chapter 54).
Was this article helpful?