To the Editor — Satellite observations show that ice losses from Antarctica have accelerated over the past 25 years1. Since 1992, the continent has contributed 7.6 mm to global sea levels, with 40% of this occurring in the past 5 years. Glaciers draining West Antarctica have retreated, thinned and accelerated due to ocean-driven melting at their termini, and the collapse of ice shelves at the Antarctic Peninsula has led to reduced buttressing and increased ice discharge2. Of the 3.2 mm yr−1 sea-level rise (SLR) measured during the satellite era3, Antarctica has contributed 0.27 mm yr−1. The magnitude of SLR from Antarctica is the largest source of uncertainty in global sea-level projections, which are key to appropriate climate change policy.
Projections of the global sea-level budget in the Fifth Assessment Report (AR5) of the IPCC3 are driven by emission scenarios that account for permutations of the physical, socioeconomic and legislative factors that will shape the century-scale increase in global temperature. These representative concentration pathways (RCPs) allow for unabated (RCP8.5), stabilizing (RCP6.0, RCP4.5) and decreasing (RCP2.6) emissions, in addition to the Special Report on Emissions Scenarios (SRES) used in the AR44. The scenarios predict between 280 and 980 mm of global mean SLR by 2100, around a central estimate of 570 mm. The contribution from Antarctica is uncertain, due to challenges in simulating the regional meteorology and the ice sheet’s dynamical response, and falls within −75 and +160 mm.
The accuracy of sea-level predictions is important because there are consequences associated with under- or overestimating the societal response required. Recent advances in the capability of ice-sheet models have improved the skill of simulations when compared to historical trends5. In AR5, the Antarctic regional meteorology was determined from an ensemble of global coupled atmosphere–ocean models6, and the ice-sheet models incorporated full numerical descriptions of ice flow and grounding-line migration3. The expected range of dynamical ice loss was assessed through depth-averaged ice flow simulations7. When combined, these contributions produce the lower, central and upper estimates of sea-level change due to Antarctica reported in AR5 (Fig. 1).
Because the satellite record of Antarctic ice-sheet mass balance1 now overlaps with a decade of the AR5 projections3, we can perform a meaningful comparison between the measured and predicted change (Fig. 1). Between 2007 and 2017, satellite observations show that Antarctica lost 1,883 Gt of ice, equivalent to a contribution of 0.55 mm yr−1 to global SLR. This value is around 30 times greater than the IPCC’s lower estimates, which predicted an average contribution of just 0.02 mm yr−1, and is now at odds with the satellite record. The rate of ice loss is also 80% higher than the AR5 central projections (0.36 mm yr−1) as a consequence of the observed acceleration, and is in fact closest to the upper range (0.68 mm yr−1).
If Antarctic ice losses continue to track the upper range of the AR5 projections, the continent will contribute 151 mm, on average, to global sea levels by 2100. When compared with the central estimate (50 mm), this amounts to an extra 101 mm of SLR. An even greater contribution is possible, because the AR5 projections did not account for the effects of increasing emission concentrations on ice-sheet dynamics, or for the possible impacts of processes such as ice cliff instabilities. Additional ice losses from Antarctica are of particular concern for cities in the Northern Hemisphere, where (owing to gravitational redistribution of ocean mass) SLR will be around 30% higher than the eustatic mean8.
Evaluating the predictions of ice-sheet losses gives a clearer picture of how reliable climate models are, raising confidence in the forecasts on which policy can be based. A higher-than-expected Antarctic contribution to future SLR has significant implications for coastal communities; it has been estimated, for example, that 10 cm of additional SLR will more than double the frequency of storm flooding in the tropics9. And without adaptation, economic losses in coastal cities caused by such flooding could reach US$1 trillion by 205010. As the upper range of Antarctic sea-level projections predict an additional contribution of this scale, these risks should be considered when developing strategies to prepare for future climate change.
This work was supported by the UK Natural Environmental Research Council (grant number cpom300001) and the European Space Agency. T.S. is funded through the NERC Ice Sheet Stability (iSTAR) programme and NERC grant number NE/J005681/1.