Search

Measurements of the 17Oexcess of H2O were obtained from ice cores in West and East Antarctica. Combined with previously published results from East Antarctica, the new data provide the most complete spatial and temporal view of Antarctic 17Oexcess to date. There is a steep spatial gradient of 17Oexcess in present-day precipitation across Antarctica, with higher values in marine-influenced regions and lower values in the East Antarctic interior. There is also a spatial pattern to the change in 17Oexcess between the Last Glacial Maximum (LGM) and Holocene periods. At coastal locations, there is no significant change in 17Oexcess. At both the West Antarctic Ice Sheet Divide site and at Vostok, East Antarctica, the LGM to Early Holocene change in 17Oexcess is about 20 per meg. Atmospheric general circulation model (GCM) experiments show that both the observed spatial gradient of 17Oexcess in modern precipitation, and the spatial pattern of LGM to Early Holocene change, can be explained by kinetic isotope effects during snow formation under supersaturated conditions, requiring a high sensitivity of supersaturation to temperature. The results suggest that fractionation during snow formation is the primary control on 17Oexcess in Antarctic precipitation. Variations in moisture source relative humidity play a negligible role in determining the glacial-interglacial 17Oexcess changes observed in Antarctic ice cores. Additional GCM experiments show that sea ice expansion increases the area over which supersaturating conditions occur, amplifying the effect of colder temperatures. Temperature and sea ice changes alone are sufficient to explain the observed 17Oexcess glacial-interglacial changes across Antarctica.

An intermediate complexity isotope model (ICM) is used to investigate the sensitivity of water isotope ratios in precipitation, including 17Oexcess, to climate variations in the Southern Hemisphere. The ICM is forced with boundary conditions from seasonal NCEP/DOE II reanalysis data. Perturbations to the surface temperature and humidity fields are used to investigate the isotopic sensitivity. The response of 17Oexcess to a uniform temperature change is insignificant over the ocean, while there is a large magnitude response over the ice sheet, particularly in East Antarctica. A decrease of ocean surface relative humidity produces increased 17Oexcess and dexcess, with a coherent response over both the ocean and Antarctica. For interior East Antarctica, the model simulates a seasonal cycle in 17Oexcess that is positively correlated with δ18O and of large magnitude (~50 per meg), consistent with the observations from Vostok. The seasonal cycle in 17Oexcess for interior West Antarctica is predicted to be considerably smaller in magnitude (12 per meg), and is negatively correlated with δ18O, consistent with new data from a firn core near the West Antarctic Ice Sheet Divide site. Over the ocean, the ICM predicts much smaller seasonal cycles in 17Oexcess. Oceanic source changes (i.e., humidity) are insufficient to explain the amplitude of the simulated seasonal cycle over the Antarctic continent. Spatial differences in the seasonal response of 17Oexcess to local temperature reflect the balance of equilibrium and kinetic fractionation during snow formation.