Reconstructions of Antarctic sea-ice at the peak of the last interglacial (marine isotope stage 5e)

海冰在调节全球海洋生态系统、海洋环流和人为气候变化方面发挥着基础性作用。据预测，在未来几十年里，南半球和北半球的海冰都将萎缩。这种不确定性的产生是因为最近南半球没有海冰退缩事件来约束模型。为了解决这个问题，我们最近发表了一份基于冰芯数据的最新间冰期海冰范围评估报告。这一重建表明，南极冬季海冰覆盖面积可能减少65% (+/-)7% (Holloway et al. 2016)（图1）。有趣的是，这表明最后一次间冰期可能是预测到21世纪末南半球海冰萎缩（高达58%）的一个极好和独特的模拟（Stocker et al., 2013）。然而，Holloway等人（2016）对海冰减少65%的计算仅基于冰芯数据。因此，这种提议的退缩迫切需要与海洋核心数据进行核对，因为这些数据可以更直接地测量海冰边缘过去的位置。一些现有的基于微化石的南大洋重建包含了在末次间冰期高峰期间海面温度升高和海冰范围大大减少的证据（Allen & Peck, submitted and refs）。其中)。但是，这些先前的研究都是在太靠北的海洋沉积物岩心上进行的，无法准确地约束海冰边缘的位置，而且它们没有足够的时间控制或样本分辨率，无法准确地解决最后一次间冰期海冰最小值的时间和大小。

Sea-ice plays a fundamental role in regulating global marine ecosystems, ocean circulation and anthropogenic climate change. Sea-ice is predicted to retreat in both the Southern and Northern Hemispheres over the coming decades. This uncertainty arises because there has not been a recent sea-ice retreat event in the Southern Hemisphere to constrain models. To help tackle this problem, we have recently published a new assessment of last interglacial sea-ice extent, based on ice core data. This reconstruction indicates a likely 65% (+/-)7% reduction in Antarctic winter sea-ice cover (Holloway et al. 2016) (Figure 1). Intriguingly, this suggests that the last interglacial is potentially an excellent and unique analogue for the predicted shrinking of Southern Hemisphere sea-ice (up to 58%) by the end of the 21st century (Stocker et al., 2013).However, the Holloway et al. (2016) calculation of a 65% reduction in sea ice is based solely on ice core data. This proposed retreat therefore urgently requires checking against marine core data, since these allow a more direct measurement of the past position of the sea-ice edge. Some existing microfossil based Southern Ocean reconstructions contain evidence of warmer sea surface temperatures and greatly reduced sea-ice extent during the peak of the last interglacial (Allen & Peck, submitted and refs. therein). But these previous studies are on marine sediment cores located too far north to accurately constrain the position of the sea-ice edge, and they have insufficient chronological control, or sample resolution, to accurately resolve the timing and magnitude of the last interglacial sea-ice minimum.