Back to the map pages

The onset of Northern Hemisphere Glaciation during the Tertiary and Quaternary.

Mark Maslin & Jonathan Adams

The onset of Northern Hemisphere glaciation - when large ice sheets first spread over the northern continents - culminated in the intense glacial-interglacial cycles that define the Quaternary, and began a unique period in Earth history in which both poles (rather than just one of them) have remained ice locked.

The earliest recorded glaciation in the Northern Hemisphere is between 10 and 6 Ma during the late Miocene (e.g., ODP Leg 151, 1994; Jansen et al., 1990; Wolf and Thiede; 1991;Jansen and Sjøholm, 1991; Wolf-Welling et al., 1995; Haug et al., 1995a; Haug, 1995). This involved a significant build up of ice on Southern Greenland. However, the process did not gain much momentum until 3.5-3 Ma, when the Greenland ice sheet expanded to include Northern Greenland. Recent evidence reviewed in Maslin et al. (1998), suggests that the initiation of large-scale Northern Hemisphere glaciation was the relatively sudden culmination of a longer term, high latitude cooling (Wolf and Thiede, 1991; ODP Leg 151 and 152, 1994; Wolf-Welling et al., 1995, 1997, Maslin et al. 1996, 1998).

After the glaciation of Greenland, the progressive spread of ice-sheets through the northernmost latitudes may have have occurred during episodes which came on relatively suddenly, at least if we look back on them against a timescale of several million years. There are various stepwise episodes of widespread increase in iceberg activity (indicated by grit dropped to the sea floor by the melting bergs), and corresponding falls in sea level indicating ice buildup on the land surfaces. These major transitions were at 2.74 Ma (corresponding to rapid glaciation of the Eurasian Arctic and Northeast Asia), 2.70 Ma (glaciation of Alaska) and major glaciation of the North American continent (2.54 Ma) (Tiedemann et al., 1994, Shackleton et al., 1995). However, the the step-like nature of the ice rafting records may conceal a fundamentally more gradual process of ice build-up inland, indicated by the progressive 18O enrichment of deep sea (benthic) isotope records (Tiedemann et al., 1994; Shackelton et al., 1995). This is because the ice-rafting records indicate only when the continental ice sheets were mature enough to spill over the edge of the landmass onto the adjacent oceans. Thus, it is not certain that these ice build-up events were really as 'sudden' as they might appear.

Several very gradual processes (taking millions of years) seem likely to have been important in setting the scene for Northern Hemisphere glaciation (Maslin et al. 1996, 1998). Tectonic changes, such as the uplift of the Himalayan-Tibetan Plateau (Ruddiman and Raymo, 1988, Ruddiman et al., 1989, and Ruddiman and Kutzbach, 1991; Raymo, 1991, 1994a, and Raymo and Ruddiman, 1992), the deepening of the Bering Strait (Einarsson et al., 1967) and/or the Greenland-Scotland ridge (Wright and Miller, 1996) and the emergence of the Panama Isthmus (Keigwin, 1978, 1982; Keller et al., 1989; Mann and Corrigan, 1990; Haug and Tiedemann, in press) have been suggested (Hay, 1992; Raymo, 1994b) as important factors. These processes seem too gradual to account entirely for the speed of Northern Hemisphere glaciation, and in particular the rapid growth phases that may be indicated by ice-rafting events.

Trying to explain the sudden-ness with which evidence for ice sheets appears in the record in each region, Lourens and Hilgen (1994), Maslin et al. (1996, 1998), and Li et al. (1998) suggest that tectonic changes may slowly have brought global climate to a critical threshold, at which point the relatively rapid variations in the Earth's orbital parameters (and thus insolation) triggered more extensive Northern Hemisphere glaciation. This theory is supported by some computer simulations of the ice-and-climate system; these suggest that it is possible to build up Northern Hemisphere ice sheets over a relatively brief 200,000 years, at around 2.75 to 2.55 Ma, by varying only the seasonal solar radiation pattern controlled by the orbital parameters (Maslin et al. 1998, Li et al. 1998). This is a relatively brief timespan considering the slowness of changes in background conditions, and the dramatic nature of the shift in the Earth's surface conditions that was involved. One should however be wary of the relative simplicity of the model, and the approximation of some factors (e.g. neglecting the differences in continental arrangement between the Pliocene and the present).