Contribution of Atmospheric Rivers to Antarctic Ice Sheet Snowfall

The Antarctic Ice Sheet (AIS) is losing mass at an accelerated pace. On top of that multi-decadal mass loss signal, which is primarily driven by ocean warming, AIS mass balance varies substantially from year to year. These mass balance variations are driven by atmospheric processes, which in turn affect snowfall, the primary input of the AIS mass balance. While annual snowfall rates on AIS are generally low, a substantial portion of the annual snowfall is associated with highly episodic marine air intrusions, associated to synoptic-scale cyclones. Some of these systems are associated with long, narrow plumes of strong horizontal water vapor transport, referred to as atmospheric rivers (ARs). While previous work has identified substantial impact of AR on local snow accumulation, an ice-sheet wide assessment of how much AIS snowfall is brought by ARs is still lacking. Here we provide a conservative estimate of the contribution of ARs to Antarctic Ice Sheet snowfall over the period 1980 to 2019, by combining an AR algorithm specifically optimized for AR detection around Antarctica and precipitation rates, both of which are based on NASA’s MERRA-2 reanalysis. The AR algorithm categorizes ARs based on anomalously strong meridional transport of water vapor in the atmosphere. Integrated over the full AIS (and excluding areas poleward of 80ºS), AR snowfall equals 194 +/- 57 Gt per year, equivalent to a 7.4 +/- 2.1 % to total AIS snowfall, with higher contributions in East Antarctica than in West Antarctica. While the absolute rates of AR snowfall decrease with elevation on the AIS, its relative contribution remains substantial to elevations of 3000 m above sea level, indicating that ARs can penetrate deep into the ice sheet. ARs explains a significant amount of interannual variability in AIS snowfall, as well as annual precipitation trends in some areas of the ice sheet. Our findings highlight the significant impact of episodic precipitation events such as ARs to annual AIS snow accumulation, as well as its temporal variability and recent trends.