Increased future ice discharge from Antarctica owing to higher snowfall

Climate models predict that precipitation will increase in Antarctica, leading to potential ice mass gain and an offset to sea level rise, but here it is shown that enhanced snowfall on Antarctica is likely to increase ice discharge and thereby negate 30% to 65% of the snowfall-induced ice gain. Climate models predict that Antarctica will gain a considerable amount of ice through additional snowfall under warming, which enhances the moisture-carrying capacity of the atmosphere. Less clear is the matter of how much ice might be lost at the same time, and what the overall effect on sea level might be. Here Ricarda Winkelmann et al. show that increases in ice discharge caused by changes in ice dynamics are likely to negate between 30% and 65% of the snowfall-induced ice gain. Their models predict a dynamic ice loss of up to 1.25 metres in the year 2500 in the strongest warming scenario, a strong sea-level-increasing counter to the potential negative contribution to global sea level by the Antarctic Ice Sheet. Anthropogenic climate change is likely to cause continuing global sea level rise1, but some processes within the Earth system may mitigate the magnitude of the projected effect. Regional and global climate models simulate enhanced snowfall over Antarctica, which would provide a direct offset of the future contribution to global sea level rise from cryospheric mass loss2,3 and ocean expansion4. Uncertainties exist in modelled snowfall5, but even larger uncertainties exist in the potential changes of dynamic ice discharge from Antarctica1,6 and thus in the ultimate fate of the precipitation-deposited ice mass. Here we show that snowfall and discharge are not independent, but that future ice discharge will increase by up to three times as a result of additional snowfall under global warming. Our results, based on an ice-sheet model7 forced by climate simulations through to the end of 2500 (ref. 8), show that the enhanced discharge effect exceeds the effect of surface warming as well as that of basal ice-shelf melting, and is due to the difference in surface elevation change caused by snowfall on grounded versus floating ice. Although different underlying forcings drive ice loss from basal melting versus increased snowfall, similar ice dynamical processes are nonetheless at work in both; therefore results are relatively independent of the specific representation of the transition zone. In an ensemble of simulations designed to capture ice-physics uncertainty, the additional dynamic ice loss along the coastline compensates between 30 and 65 per cent of the ice gain due to enhanced snowfall over the entire continent. This results in a dynamic ice loss of up to 1.25 metres in the year 2500 for the strongest warming scenario. The reported effect thus strongly counters a potential negative contribution to global sea level by the Antarctic Ice Sheet.