Antarctic Sea Ice Response

Antarctic Sea Ice Response in CMIP5 Pre-industrial, Historical and Ozone Perturbation Simulations

Abstract

Multi-decadal Antarctic sea-ice trends from three CMIP5 modelling systems (ACCESS1.0, ACCESS1.3 and CSIRO Mk3.6) are compared against recent observations.  Several hypotheses are advanced to explain the recent small positive trend in overall extent. Area weighted trends in some ensemble members of the CSIRO Mk3.6 system match the observations reasonably well. Whilst the CSIRO results with ozone forcing are more realistic than runs without the recent ozone perturbation, with regional patterns being enhanced in the Bellingshausen and Weddell seas and East Antarctica, from these results we do not exclude the hypothesis that the current positive trend in observed net sea ice extent is due to natural variability. The regional patterns of ice advance and mid-winter ice maximum are well represented in the ACCESS1.0 and 1.3 modelling systems and are driven by atmospheric pressure, air and ocean temperature trends; inclusion of ozone forcing did not have a discernible effect.

The ice-ocean fluxes and other forcings of the ocean multi-decadal variability have been investigated in the 500 year long pre-industrial control for both the CSIRO and ACCESS systems, to determine how the natural variability of the system is aligning with the signal in the historical runs.