Book excerpt: As the climate has warmed over the past decades, the amount of melt on the Greenland Ice Sheet has increased, and areas higher on the ice sheet have begun to melt regularly. This increase in melt has been hypothesized to enhance ice flow in myriad ways, including through basal lubrication and englacial refreezing. By developing and interpreting thermal ice-sheet models and analyzing remote sensing data, I evaluate the effect of these processes on ice flow and sea-level rise from the Greenland Ice Sheet. I first develop a thermal ice sheet model that is applicable to western Greenland. Key components of this model are its treatment of multiple phases (solid ice and liquid water) and its viscosity-dependent velocity field. I apply the model to Jakobshavn Isbræ, a fast-flowing outlet glacier. This is an important benchmark for my model, which I next apply to the topics outlined above. I use the thermal model to calculate the effect of englacial latent-heat transfer (meltwater refreezing within englacial features such as firn and crevasses) on ice dynamics in western Greenland. I find that in slow-moving areas, this can significantly warm the ice, but that englacial latent heat transfer has only a minimal effect on ice motion (10%). By contrast, in fast-flowing regions, which contribute most (60%) of the ice flux into the ocean, evidence of deep englacial warming is virtually absent. Thus, the effects of englacial latent heat transfer on ice motion are likely limited to slow-moving regions, which limits its importance to ice-sheet mass balance. Next, I couple a model for ice fracture to a modified version of my thermal model to calculate the depth and shape evolution of water-filled crevasses that form in crevasse fields. At most elevations and for typical water input volumes, crevasses penetrate to the top ~200–300 meters depth, warm the ice there by ~10°C, and may persist englacially, in a liquid state, for multiple decades. The surface hydrological network limits the amount of water that can reach most crevasses. We find that the depth and longevity of such crevasses is relatively robust to realistic increases in melt volumes over the coming century, so that we should not expect large changes in the englacial hydrological system under near-future climate regimes. These inferences put important constraints on the timescales of the Greenland supraglacial-to-subglacial water cycle. Finally, I assess the likelihood that higher-elevation surface melt could deliver water to regions where the bed is currently frozen. This hypothetical process is important because it could potentially greatly accelerate the seaward motion of the ice sheet. By analyzing surface strain rates and comparing them to my modeled basal temperature field, I find that this scenario is unlikely to occur: the conditions necessary to form surface-to-bed conduits are rarely found at higher elevations (~1600 meters) that may overlie frozen beds.