Modeling oxygen isotopes in ice sheets linked to Quaternary ice-volume variations

Climate change through the Quaternary was dominated by repeated build-up and retreat of large ice sheets, especially in the Northern Hemisphere. The best indicator for past variations of global ice volume is the oxygen-isotopic composition (d18O) of seawater (dw) recorded in foraminiferal calcite in marine sediments. But the interpretation calcite-d18O (dc) records is not straightforward, mainly because dc depends on both dw and seawater temperature. Furthermore, in common paleoceanographic practice the relationship between dw and ice volume is assumed linear, but two main factors may induce considerable nonlinearity: the mean isotopic composition of ice, which varies during a glacial cycle, and the ocean circulation. In the present study, these two factors were investigated at glacial-interglacial timescale using a 2.5-dimensional thermomechanical ice-sheet model including oxygen-isotope transport, combined with different ocean models. First, using a well-mixed ocean, it was found that the effect of mean-ice d18O variations can be neglected in reconstructing ice volume from marine d18O records. Second, the effect of ocean circulation was investigated qualitatively by replacing the well-mixed ocean by a spatially-resolved ocean model. It was concluded that the ocean circulation is an important element to be taken into account in the climatic and stratigraphic interpretation of dc records and that leads/lags inferred on the basis of proxies depending on more than one physical variable, such as dc, can be misleading in interpreting causal relationships. Finally, the interaction between climate variability at orbital and millennial timescales was investigated in order to gain further understanding of the origin of d18O variations in marine sediments and ice cores. It was found that YD-type events may occur during any deglaciation of the past 800 kyr, so that YD is probably not a one-time event, but an intrinsic feature of the climate change at millennial timescales.