Present and future variations in Antarctic firn air content

A firn densification model ( FDM) is used to assess spatial and temporal ( 1979-2200) variations in the depth, density and temperature of the firn layer covering the Antarctic ice sheet ( AIS). A time-dependent version of the FDM is compared to more commonly used steady-state FDM results. Although the average AIS firn air content ( FAC) of both models is similar ( 22.5 m), large spatial differences are found: in the ice-sheet interior, the steady-state model underestimates the FAC by up to 2 m, while the FAC is overestimated by 5-15 m along the ice-sheet margins, due to significant surface melt. Applying the steady-state FAC values to convert surface elevation to ice thickness ( i.e., assuming flotation at the grounding line) potentially results in an underestimation of ice discharge at the grounding line, and hence an underestimation of current AIS mass loss by 23.5% ( or 16.7 Gt yr(-1)) with regard to the reconciled estimate over the period 1992-2011. The timing of the measurement is also important, as temporal FAC variations of 1-2 m are simulated within the 33 yr period ( 1979-2012). Until 2200, the Antarctic FAC is projected to change due to a combination of increasing accumulation, temperature, and surface melt. The latter two result in a decrease of FAC, due to ( i) more refrozen meltwater, ( ii) a higher densification rate, and ( iii) a faster firn-to-ice transition at the bottom of the firn layer. These effects are, however, more than compensated for by increasing snowfall, leading to a 4-14% increase in FAC. Only in melt-affected regions, future FAC is simulated to decrease, with the largest changes (-50 to -80 %) on the ice shelves in the Antarctic Peninsula and Dronning Maud Land. Integrated over the AIS, the increase in precipitation results in a similar volume increase due to ice and air ( both similar to 150 km(3) yr(-1) until 2100). Combined, this volume increase is equivalent to a surface elevation change of +2.1 cm yr(-1), which shows that variations in firn depth remain important to consider in future mass balance studies using satellite altimetry.