It has been over a decade since publication of the CLIMAP [Denton and Hughes, 1981] global reconstruction of ice sheets at the last glacial maximum for the CLIMAP Project. To an extent never envisaged at the time, this reconstruction has been used by groups worldwide as boundary-condition input for general circulation models (GCMs) and other numerical experiments to reconstruct glacial-maximum atmospheric and oceanic circulation. Other GCM experiments based on rough approximations of ice-sheet size through the deglacial interval also have been run [Kutzbach and Guetter, 1986], and the climatic output has in turn been compared against ground-truth data such as vegetation and lake levels [COHMAP, 1988].

Recently, several results have provided fresh impetus for updating the CLIMAP reconstruction: coral reef data [Fairbanks, 1989] placing sea level near -120 m at the last glacial maximum at Barbados, which also provide a deglacial record of sea level; U-series dates on these corals [Bard et al., 1990] showing that 14C chronologies are anomalously young through the deglaciation, which place the last glacial maximum at 21,000–20,000 years ago; major improvements [Tushingham and Peltier, 1991] in iteratively matching hypothesized ice-sheet thickness histories based on an independent model of the Earth's visco-elasticity to radiometrically dated sea-level rebound histories; and glacial and marine geological fieldwork by many scientists. These findings provide significant new constraints on the past size of the ice sheets at the last glacial maximum and throughout deglaciation. As an added impetus, the NATO Paleoclimate Modeling Intercomparison Project (PMIP) has called for an updated reconstruction of glacial-maximum ice sheets. Their objective is to compare the performance of GCMs when tested on a realistic (glacial) world having ice-sheet boundary conditions very different from today's.