This is a hotly (or perhaps that should be coldly) disputed question. Over the past 50 million years, the Earth's climate has cooled. Large ice sheets formed on Antarctica 35 million years ago, but did not form in the Northern Hemisphere until about 2.7 million years ago. Earth scientists mostly agree that global climatic cooling is associated with decreasing levels of carbon dioxide in the atmosphere. They also generally agree that ice sheets grow only if sufficient moisture is available to produce winter snow, which in is then able to survive the summer heat. So what triggered the onset of the Quaternary ice ages 2.7 million years ago? A trigger mechanism for the onset of glacial conditions seems essential because orbital forcing has always occurred but it has not always induced a strong reaction in the climate system. Why, for example, did glacial stages occur in the Pleistocene but not in the Palaeogene or Neogene? As Walter Wundt (1944) observed, it would seem that only under certain geographical and geological circumstances do orbital variations permit ice ages to make an appearance.
A precondition of glaciation appears to be a cold climatic background. Now, many factors could cause the atmosphere to become cooler (see Raymo 1994). Key factors seem to be the carbon dioxide (and possibly methane) content of the atmosphere, the arrangement of continents and oceans, and the presence of large mountain ranges, the last two of which affect continental temperatures and potential moisture sources. It is possible that the redistribution of continents and oceans and the growth of mountain ranges through the Cenozoic era led gradually to a cooling of the atmosphere that, by Pleistocene times, was cold enough to permit ice sheets to form during insolation minima. However, early experiments with a general circulation model, run with orbital configurations corresponding to times of rapid ice sheet growth, raised doubts about the ability of Croll-Milankovitch forcings to trigger the growth of ice sheets (Rind et al. 1989). Under none of the orbital configurations was the model able to maintain snow cover through the summer at locations suspected of being sites where major ice sheets began forming, despite reduced insolation during the summer and autumn. The model also failed to preserve a layer of ice, 10 m thick, placed in localities where ice existed during the Last Glacial Maximum. Only by adjusting ocean surface temperatures to their values at the height of the ice age could the model manage to preserve a smallish patch of ice in northern Baffin Island. To David Rind and his colleagues, the experiments brought out a wide discrepancy between the response of a general circulation model to orbital forcing and geophysical evidence of ice sheet initiation, and indicated that the growth of ice occurred in an extremely ablative environment. To explain how ice sheets might grow in an ablative environment requires a more complicated model or else a climatic forcing other than orbital perturbation - a reduction in carbon dioxide content is a possibility.
Studies of marine plankton (diatoms, coccolithophores, and foraminifera) from the floor of the subarctic Pacific Ocean help clarity the role of moisture supply in initiating Northern Hemisphere ice sheets (Haug et al. 2005). Alkekone unsaturation ratios and diatom oxygen-isotope ratios suggest that, 2.7 million years ago, summers warmed and winters cooled, so inflating the seasonal temperature contrast of the subarctic Pacific Ocean. The warmer summer seas heated the atmosphere, enabling it to hold more moisture. The upshot was that, like 'a snow gun blasting away at ski slopes', prevailing westerly winds blew the moisture onto the cold North American continent where it fell as snow and accumulated as ice (Billups 2005). The cause of the sudden increase in late summer temperatures seems to relate to a change in the mixing between surface and deep ocean waters. Surface waters will not warm if they mix efficiently with cooler and deeper water, a situation that seems to have prevailed up to 2.7 million years ago, as indicated by diatom abundance. After that time, diatom abundance nose-dived, probably because the nutrient supply fed by mixing with deeper waters stopped, at least on a seasonal basis, as a halocline developed. The reduction in vertical mixing of the ocean waters allowed the surface sea to warm during late summer and early autumn, which led to the loading of the 'snow gun' that triggered the onset of the Northern Hemisphere glaciation. Had not the ocean mixing stopped, the obliquity minimum that occurred 2.7 million years ago would have reduced water vapour transport and starved ice-sheets of a snow supply (Haug et al. 2005).
George Kukla and Joyce Gavin (2004) make a radical proposal to explain the onset of the last glaciation. They contend that the main impact of past orbital changes on climate was in changing the strength of the solar beam in early spring and autumn, and was not, as is the customary view, in varying summer insolation at high latitudes. At the last interglacial-glacial transition, this shift led to a warming of low latitude oceans and cooling of the northern lands. The increased equator-to-pole and ocean-to-land temperature gradients facilitated the poleward transfer of water onto land-based ice. The earlier ocean warming, combined with decreased water vapour greenhouse forcing over land in spring and earlier establishment of snowfields in autumn, led to the growth of ice sheets and to intermittent episodes of accelerated calving. This model only addresses the first 20,000 years of a glacial cycle. It does not explain the full development of a glaciation and the processes precipitating the collapse of the ice sheets some 100,000 years later, although the relative impact of orbital variations, as opposed to the dynamics of ice and ocean currents, might well decrease during the course of an interglacial-glacial succession. In detail, the Kukla-Gavin (2004, 44) hypothesis runs as follows:
1 During the last interglacial, the climate was quasi-stable and roughly similar to the climate of the current interglacial, with the dominant strength of the solar beam in boreal autumn magnified by high obliquity.
2 Some 116,000 years ago, the equinoctial seesaw (ESS) - the difference between the strength of the solar beam at the top of the atmosphere at spring and autumn equinoxes - shifted from an autumn mode to a spring mode (Figure 4.3). This led to an increase of the El Niño frequency and intensity at the expense of the La Niña frequency and intensity, and a warming of tropical oceans. At the same time, the northern lands cooled owing to a decreased greenhouse forcing and the earlier growth of seasonal snowfields. Additional moisture for the build-up of high latitude ice came from water bodies that warmed earlier in spring and remained warm later into the autumn. The weaker autumnal insolation intensified the temperature difference between relatively warm waters and cooler lands.
3 The meridional circulation strengthened in boreal autumn and winter in response to steeper insolation and temperature gradients between colder high latitudes and warmer tropics. Moreover, the large temperature contrast between the oceans and the land caused intensification in the transfer of ocean water onto the land-based ice.
4 Accumulation of snow in nivation zones increased, favouring the growth of glaciers in high latitudes. The changed distribution of the ice mass accelerated the outflow of ice into the forelands and open ocean. The sea-level dropped.
5 Episodic surges from glacier margins into the ocean lowered sea-surface temperatures and salinity, which in turn extended the duration of seasonal pack ice and enabled more intense and farther reaching outbreaks of Arctic air (Leroux 1993). These changes affected the thermohaline circulation (Broecker 1991), and might have led to a worldwide alteration of oceanic circulation and encouraged the flip-flop behaviour of local climates (Bond and Lotti 1995).
6 Eventually, a major outbreak of polar ice into the oceans cooled the subtropical and tropical oceans to such a degree that the meridional temperature and precipitation gradients decreased (Bush and Philander 1998; Broecker 1991), starving the glaciers.
In short, this new mechanism for starting up the global refrigeration system has the oceans in low latitudes as the key recipient of the insolation signal. Conventional mechanisms, in contrast, regard the increased albedo and the concomitant drop of land temperature as the only trigger of glacial conditions. Under the new mechanism, the warming of low latitude oceans and the increased temperature gradient between the warmer oceans and cooler land combine with decreased water-vapour greenhouse forcing in autumn and with the earlier establishment of snowfields, leading to the accumulation of polar ice and drop of sea-level. The hypothesis rests upon the correlation of radiometrically dated palaeoclimatic evidence with computed past orbital variations. No other climate model, with one exception (see Clement et al. 1999), yet supports it. Even so, the existing data show that the warming of tropical oceans, the probable increase of global mean temperature, and the growth of polar ice accompanied the past orbital shift, which is qualitatively similar to, but stronger than, current orbital shift. Thus, the current global warming may be a product of both humanmade and natural causes.
Was this article helpful?