Ice sheets once were viewed as passive elements in the climate system enslaved to orbitally generated variations in solar radiation. Today, modeling results and new geologic records suggest that ice sheets actively participated in late-Pleistocene climate change, amplifying or driving significant variability at millennial as well as orbital timescales. Although large changes in global ice volume were ultimately caused by orbital variations (the Milankovitch hypothesis), once in existence, the former ice sheets behaved dynamically and strongly influenced regional and perhaps even global climate by altering atmospheric and oceanic circulation and temperature.
Experiments with General Circulation Models (GCMs) yielded the first inklings of ice sheet's climatic significance. Manabe and Broccoli [1985], for example, found that the topographic and albedo effects of ice sheets alone explain much of the Northern Hemisphere cooling identified in paleoclimatic records of the last glacial maximum (~21 ka).
Several features of ice sheet/atmosphere interactions appear to be robust, as they are common results of several independent GCM investigations as well as of model investigations which use reduced or simplified treatments of atmospheric physics. These include displacement of the jet stream by high ice sheets, significant cooling over and downwind of the ice sheets, and reorganization and strengthening of storm tracks along the large North American (Laurentide) ice sheet and extending across the North Atlantic. A global array of climate records reveals these climatic effects of changes in ice-sheet dimensions at orbital timescales (104-105 yr) (Figures 1a and 1b).
Fig. 1.
(a) The SPECMAP stacked
O record for the last 140,000
years
[from Imbrie et al., 1992]. (b) North Atlantic
13 C record of
variations in the percentage of North Atlantic Deep Water over the last 140,000
years, interpreted to be driven by ice sheets through their
influence on the wind field [Imbrie et al., 1992]. (c) Record of
O variations over the last
45,000
years measured in ice core from Summit, central Greenland by
Greenland Ice-Core Project (GRIP) [Dansgaard et al., 1993]. (d)
Record of variations in ice-rafted debris (IRD) in the North
Atlantic (core Vema 23-081) over the last 45,000 radiocarbon
years [Bond and Lotti, 1995]. Heinrich events 1-4 are shown
as H1-H4.
Associated climate change lags insolation changes in response to the slow time constant of large ice sheets [Imbrie et al., 1992]. Important issues yet to be explored by GCMs are the influence of smaller ice sheets on climate, the detailed distribution of surface temperature, precipitation, and run-off on the surfaces of ice sheets, and interhemispheric communication. What influenced larger ice sheets' atmospheric circulation was their changing three-dimensional geometry through the last glacial cycle [Kutzbach and Guetter, 1986]. Before their role in the climate system can be qualified, ice-sheet geometry must be reconstructed. This geometry is constrained by a variety of geologic records from the glaciated continents and the ocean. The record of eustatic sea level, in particular, is valuable for determining changes in ice volume. Inverse modeling of ice-sheet history is a powerful tool for identifying changes in ice-sheet size from the last glacial maximum to the present. The ICE-4G model of the global ice sheets, which is constrained to match the Barbados sea level record, suggests that ice-sheet elevations during the last glacial maximum were far lower than those reconstructed by CLIMAP Project Members in 1981 [Peltier, 1994]. The influence of large areas of water-saturated sediments beneath the ice sheets that provided little resistance to ice flow may account for these lower elevations. These new ice-sheet boundary conditions are being used in the Paleoclimate Model Intercomparison Project (PMIP) and are likely to generate different climatic effects than the higher CLIMAP ice sheets used in previous GCM experiments.
Coupling ice-sheet models to atmospheric GCMs will help resolve problems such as mass balance parameterizations and identify time-dependent responses to changes in such boundary conditions as ice-sheet dimensions, insolation, and atmospheric composition. Ice-sheet timescales, however, can be far longer than those of atmospheric processes; thus, coupling of ice, ocean, and atmospheric models remains technically difficult, and possible, perhaps, only through asynchronous stepping at low resolution. Atmospheric conditions at the surface of the ice sheets derived from future GCM experiments directed toward atmosphere/ice-sheet coupling will undoubtedly be expected to maintain the ice sheets in steady state at or near the configurations suggested by the geologic record.
Finally, GCMs show that - with no change in atmospheric composition - most climatic response to ice sheets is confined to the Northern Hemisphere. Interhemispheric coupling is a crucial process that requires further investigation; it may reveal information about synchronous climatic variations across the equator. Such investigation may require coupled ocean-atmosphere GCMs that treat heat transport in the ocean, such as heat transport linked to the deep thermohaline circulation that originates in the North Atlantic.
High-resolution paleoclimate records from ice cores and marine and terrestrial sediments suggest climate variability at higher frequencies (10³ yr) than those predicted by orbital forcing (Figure 1c). Terrestrial and marine records also reveal ice-sheet variability that corresponds with high-frequency climate change (Figure 1d). Ice sheet variations and abrupt climate change at these millennial timescales are linked by data from the marine record of ice-rafted debris in the North Atlantic. Heinrich events, which are episodes of massive discharge of icebergs to the North Atlantic [Broecker, 1994], and their newly discovered higher-frequency cousins [Bond and Lotti, 1995] indicate partial collapse of the Laurentide Ice Sheet and possibly other ice sheets in Iceland and Europe as well. Data from the Labrador Sea reveal a trail of ice-rafted debris and suggest that Heinrich events originated from an ice stream draining the Laurentide Ice Sheet through Hudson Strait [Andrews et al., 1994]. Bond and Lotti [1995], however, find that more than one ice sheet - and more than one sector of the Laurentide Ice Sheet - contributed ice-rafted debris to the North Atlantic during each Heinrich event as well as during many of the higher-frequency ice-rafted debris events. These higher-frequency events seem to march 3 or 4 steps for every Heinrich event. The Greenland ice core record shows that ice-rafted debris events occurred at the end of cooling episodes in the North Atlantic region and are followed by rapid warming approaching interglacial conditions (Dansgaard-Oeschger events).
Changes in ice sheet size corresponding to the ice-rafted debris events were smaller than orbitally forced changes. However, they may have played a critical role in abrupt, frequent climate change in the circum-North Atlantic region by releasing icebergs and meltwater into the North Atlantic. These episodic discharges from ice sheets may influence a salt oscillator controlling the North Atlantic thermohaline circulation [Birchfield et al., 1994] that affects regional climate. Its influence is recorded in the Greenland ice core paleotemperature record.
The cause of this episodic discharge, however, is unknown. The answer depends on whether ice sheet discharges are caused by external climate forcing or by an internal ice sheet surge mechanism that operates independently of climate [Broecker, 1994]. Results of low-order dynamical modeling suggest that unstable ice sheet behavior such as Heinrich events may develop from periods of basal melting, resulting in a surge [MacAyeal, 1993; Verbitsky and Saltzman, 1995]. It is equally possible that an external, iceberg-discharge-inducing climate change operative in the atmosphere or ocean is the cause.
Of direct relevance to this issue is new evidence that climatic change in the Southern Hemisphere occurs simultaneously with several ice-rafting events in the North Atlantic [Lowell et al., 1995]. Whether this suggests that global climate change and ice-rafting events result from external forcing or from a wide-ranging influence of ice sheet size and stability on the global climate is unknown. In either case, rapid changes in ice sheet dimensions could be transmitted globally through the atmosphere via topographically forced atmospheric waves and/or through the freshwater influences on ocean circulation, as they may have been on orbital timescales [Imbrie et al., 1992]. Northern Hemisphere ice sheets served many roles in the rich climate variability of the last glaciation. At orbital timescales, they orographically forced the atmosphere and discharged meltwater and icebergs into the North Atlantic as the result of insolation-forced changes in ice sheet dimensions and melting rates. These effects amplified and transmitted insolation changes to other parts of the global climate system. A similar relation may hold for ice sheet variations occurring on millennial timescales. Because much of the climate change at frequencies higher than Milankovitch appears to be associated with ice sheet variations, further understanding the role of ice sheets in the climate system depends on identifying the cause of these variations.
This topic was the basis for a NSF-sponsored workshop held at Oregon State University in December 1993. We thank the participants and Chip Levy for their contributions to this article, which was supported by the Climate Dynamics, Geology and Paleontology, and Polar Glaciology Programs of the National Science Foundation.
Andrews, J. T., H. Erlenkeuser, K. Tedesco, A. E. Aksu, and A. J. T. Jull, Late Quaternary (stage 2 and 3) meltwater and Heinrich events, northwest Labrador Sea, Quat. Res., 41, 26, 1994.
Birchfield, E. G., H. Wang, and J. J. Rich, Century/millennium internal climate oscillations in an ocean-atmosphere-continental ice sheet model, J. Geophys. Res., 99, 12,459, 1994.
Bond, G. C., and R. Lotti, Iceberg discharges into the North Atlantic on millennial time scales during the last glaciation, Science, 267, 1005, 1995.
Broecker, W. S., Massive iceberg discharges as triggers for global climate change, Nature, 372, 421, 1994.
CLIMAP Project Members, Seasonal reconstructions of the earth's surface at the last glacial maximum, Geol. Soc. Am. Map Chart Ser., MC-36, 1981.
Dansgaard, W., et al., Evidence for general instability in past climate from a 250-kyr ice-core record, Nature, 364, 218, 1993.
Imbrie, J., et al., On the structure and origin of major glaciation cycles 1. Linear responses to Milankovitch forcing, Paleoceanography, 7, 701, 1992.
Kutzbach, J. E., and P. J. Guetter, The influence of changing orbital parameters and surface boundary conditions in climate simulations for the past 18,000 years, J. Atmos. Sci., 43, 1726, 1986.
Lowell, T. V., C. J. Heusser, B. G. Andersen, P. I. Moreno, A. Hauser, G. H. Denton, L. E. Heusser, C. Schluchter, and D. R. Marchant, Interhemispheric symmetry of paleoclimatic events during the last glaciation, Science, in press, 1995.
MacAyeal, D. R., Growth/purge oscillations of the Laurentide ice sheet as a cause of the North Atlantic's Heinrich events, Paleoceanography, 8, 775, 1993.
Manabe, S., and A. J. Broccoli, The influence of continental ice sheets on the climate of an ice age, J. Geophys. Res., 90, 2167, 1985.
Peltier, W. R., Ice age paleotopography, Science, 265, 195, 1994.
Verbitsky, M., and B. Saltzman, A diagnostic analysis of Heinrich glacial surge events, Paleoceanography, 10, 59, 1995.
Return to Science and
Society
Return to Starting
Point
