On the reconstruction of pleistocene ice sheets: A review

Pleistocene ice sheets can be reconstructed through three separate approaches: (1) Evidence based on glacial geological studies, such as erratic trains, till composition, crossing striations and exposures of multiple tills/nonglacial sediments. (2) Reconstructions based on glaciological theory and observations. These can be either two- or three-dimensional models; they can be constrained by ‘known’ ice margins at specific times; or they can be ‘open-ended’ with the history of growth and retreat controlled by parameters resting entirely within the model. (3) Glacial isostatic rebound after deglaciation provides a measure of the distribution of mass (ice) across a region. A ‘best fit’ ice sheet model can be developed that closely approximates a series of relative sea level curves within an area of a former ice sheet; in addition, the model should also provide a reasonable sea level fit to relative sea level curves at sites well removed from glaciation.This paper reviews some of the results of a variety of ice sheet reconstructions and concentrates on the various attempts to reconstruct the ice sheets of the last (Wisconsin, Weischelian, Würm, Devensian) glaciation. Evidence from glacial geology suggests flow patterns at variance with simple, single-domed ice sheets over North America and Europe. In addition, reconstruction of ice sheets from glacial isostatic sea level data suggests that the ice sheets were significantly thinner than estimates based on 18 ka equilibrium ice sheets (cf. Denton and Hughes, 1981). The review indicates it is important to differentiate between ice divides, which control the directions of glacial flow, and areas of maximum ice thickness, which control the glacial isostatic rebound of the crust upon deglaciation. Recent studies from the Laurentide Ice Sheet region indicate that the center of mass was not over Hudson Bay; that a major ice divide lay east of Hudson Bay so that flow across the Hudson Bay and James Bay lowlands was from the northeast; that Hudson Bay was probably open to marine invasions two or three times during the Wisconsin Glaciation; and that the Laurentide Ice Sheet was thinner than an equilibrium reconstruction would suggest.