The Last Glacial Maximum (LGM) form of the Laurentide ice-sheet has been actively debated for a century or more. A continuing focus of debate has been upon the issue of the number of “domes” that characterized the complex during its time of maximum extent. Although ice mechanics based attempts to characterize this structure invariably deliver Hudson Bay centered monodomal models, such analyses cannot be considered definitive.

An alternative approach to the reconstruction of LGM ice-sheet topography is that based upon the geophysical inversion of relative sea level (RSL) observations. When employed in conjunction with accurate information concerning the space-time distribution of land ice, this method may deliver accurate estimates of ice thickness. The Laurentide component of the global ICE-4G model obtained in this way is monodomal but unconstrained in the continental interior.

Very recently evidence has been forthcoming that argues strongly that the ICE-4G reconstruction contains a significant error over the Keewatin region of the Northwest Territories and to the southwest of Hudson Bay, an error that may be corrected only by the incorporation of a significant dome of ice over Keewatin and increased concentration to the south.

The new data that require this modification to ICE-4G include improved constraints upon the LGM lowstand of eustatic sea level derived from observations of shelf inundation at sites remote from the ice-sheets which strongly suggest that the volume of continental ice in ICE-4G at LGM was too low by approximately 10%. That this missing ice must have been located over the Laurentide platform is strongly suggested by a combination of space geodetic and absolute gravity measurements. A new multi-domed model of the Laurentide ice-sheet that enables the global model of the last deglaciation event to satisfy all of these constraints will be discussed.