Floating ice shelves that fringe the Antarctic Ice Sheet are subject to basal melting, and this is a significant source of uncertainty in future sea level rise projections. Part of this uncertainty is driven by the unpredictable internal variability in ocean and atmospheric processes that influence temperature around these ice shelves. Large climate model ensembles can potentially be used to force numerical ice sheet models and enable uncertainty quantification of the resulting projections. However, such ensembles are computationally expensive to run, necessitating alternative methods. Here, we demonstrate a technique to generate independent realizations of internal climate variability from a single climate model run to force an ensemble of ice sheet model simulations. We build on prior model emulation methods to efficiently sample plausible realizations of internal variability in Antarctic ocean forcing, and force the MPAS Albany Land Ice ice sheet model. By using empirical orthogonal function decomposition and Fourier-phase randomization, we generate variability fields that are statistically consistent with the climate model output used. The ensemble spread of the evolving ice sheet state can then be further decomposed into constituent signals to identify the relevant scales of spatiotemporal variability important for future ice sheet projection uncertainty.