Session: Glacial Climates (LGM, Last deglaciation, Ice sheet uncertainties, Glacial-interglacial cycles)
Author: Xu Zhang / email@example.com / Alfred Wegener Institute Helmholtz centre for Polar and Marine Research
Co-author: Evan Gowan, Alfred Wegener Institute Helmholtz centre for Polar and Marine Research;
Martin Werner, Alfred Wegener Institute Helmholtz centre for Polar and Marine Research;
Gerrit Lohmann, Alfred Wegener Institute Helmholtz centre for Polar and Marine Research;
An uncertainty in PMIP4 glacial simulations is the ice sheet reconstruction (height and geometry) that appears to play an important role on glacial climate (e.g the strength of Atlantic meridional overturning circulation). Here we use an isotope-enabled fully coupled AOGCM to assess the transient and equilibrium features of glacial climate in response to five available ice sheet reconstructions of the Last Glacial Maximum (Licciardi et al 1998; ANU; GLAC-1; ICE-6g; ICESHEET). Here we mainly focus on different LIS reconstructions since the Laurentide Ice Sheets (LIS) play a more important role on Atlantic meridional overturning circulation than the Fennoscandia Ice Sheets (FIS). According to the ice volume, LIS reconstructions from Licciardi and ICESHEET (Gowan et al 2016) can be classified into a low class, while the rest belongs to a high class. The initial ocean state of the experiments are identical to each other and derived from a previous PMIP3 LGM simulation. During the integrations, we find that the equilibrium timescale in ocean circulation is much shorter in the high class than in the low class. In particular, the ocean circulation experiences a pronounced weakening in its transient phase under the low ice sheets, followed by a rapid recovery to a strong ocean state. As the system gets equilibrium, the simulated climate states regarding sea surface temperature, sea ice cover and precipitation patterns are also distinguished, especially between low and high ice sheet classes. Our results suggest that various ice-sheet reconstructions will be a considerable source of uncertainty for inter-comparisons of PMIP4 LGM simulations.