Reconstructions of paleo winter sea ice concentrations and revised age models for marine sediment cores SO136-111 and E27-23.

The Holocene sea-ice project brings together for the first time, records from the Antarctic continent and deep sea sediments that will allow us to calibrate three sea-ice extent surrogates, validate their use in contrast to satellite observations and explore climatic influence on the physio-ecological environment over the last 10,000 years.Spreadsheet 1 (appendix A): Complete list of Accelerator Mass Spectrometry (AMS) dating completed on E27-23 from various identified sources with original 14CAge and reported error. Three dates identified as Burckle pers comm. here were provided by Dr Lloyd Burckle (LDEO) to Dr L. Armand for this work. Outlier attributions are identified; the term Averaged identifies the two samples where final calibrated dates were averaged in this work. All remaining AMS dates were converted to calendar ages using the linear-based CALIB07 (Stuiver and Reimer, 1993) with calibration to the Marine13 dataset (Reimer et al., 2013) at 95% confidence (sigma 2) and included a correction for the surface water reservoir age of ~752 years at the site of core E27-23 resolved from the marine radiocarbon reservoir correction database and software available from http://radiocarbon.LDEO.columbia.edu/ (Butzin et al., 2005). The percent Marine Carbon relative attribution is provided. The Median age (Cal Yr BP) used as the final age at each respective (mid) depth is provided.In Appendix A the dates are all ages in years, however some are uncalibrated ages and others are Cal yr BP (= calendar years before present).So in terms of headings in Table A:Raw 14C age yr BP - is the raw age provided by radiocarbon dating without any corrections applied. It is in years before present.Corrected raw age (RA=752) - is the raw age with a local RA (Reservoir Age) correction applied and is still in years before present.The remaining ages are calendar years before present having been calibrated.All formats follow recommendations for reporting raw 14C dates and their calibration ages.Spreadsheet 2 (appendix B): Comparison of calibration output from the input of accepted 14C dates using OXCAL 4.2 (Bronk Ramsey 2009; Blaauw 2010), and CALIB07 (Stuiver and Reimer, 1993), both using the Marine13 calibration curve (Reimer et al., 2013) at 95.4% confidence (sigma 2) and including a correction for the surface water reservoir age of ~752 years at the site of core E27-23. The calibration output difference between the median Cal Yr BP, regardless of calibration method employed, was greater than or equal to 40 Cal Yr BP. Calibration data from the output of CALIB07 has been used in this paper to determine chronostratigraphy.Spreadsheet 3 (appendix C): The foraminiferal stable isotope data from E27-23. Ratios of oxygen (delta 18O) measured from the planktonic foraminifer Neogloboquadrina pachyderma sinistral (greater than 150 microns). Isotope values are reported as per mil (%) deviations relative to the Vienna Peedee Belemnite (VPDB).Spreadsheet 4 (appendix D): The paleo winter sea-ice concentration (wSIC) estimates for marine sediment core SO136-111. The calendar ages, in thousands of years before present (kyr BP), are provided for each sample from core SO136-111. For each of the samples in core SO136-111, we have provided the estimates winter sea-ice concentration (%), along with the associated lower and upper bounds for the 95% confidence interval around the estimated winter sea-ice concentration (%), for both GAM/WSI/13 and GAM/WSI/ETS. The final two columns provide the estimated average annual monthly sea-ice cover for each sample within core SO136-111, originally estimated using the Modern Analogue Technique, by Crosta et al. (2004). Finally, we provide the estimated summer sea surface temperature, again using the Modern Analogue Technique, from Crosta et al. 2004. Spreadsheet 5 (appendix E): The paleo wSIC estimates for marine sediment core E27-23. The calendar ages, in thousands of years before present are provided for each sample from core E27-23. For each of the samples in core E27-23, we have provided the estimated winter sea-ice concentration (%), along with the associated lower and upper bounds for the 95% confidence interval around the estimates for winter sea-ice concentration (%).

全新世海冰项目（Holocene sea-ice project）首次汇集了南极大陆与深海沉积物的相关记录，借此可对3种海冰范围代用指标进行校准，对比卫星观测数据验证其适用性，并探索过去1万年来气候对生理-生态环境的影响。

附表1（附录A）：针对岩芯E27-23的加速器质谱法（Accelerator Mass Spectrometry, AMS）测年完整列表，数据源自各已确认来源，包含原始碳14年龄与报告误差。其中3个标注为「Burckle pers comm.」的测年数据由劳埃德·伯克尔博士（LDEO）为本研究提供给L·阿尔芒博士。异常值已标注；「平均」一词指代本研究中对最终校准年代取平均值的2个样品。其余所有AMS测年数据均通过基于线性模型的CALIB07软件（Stuiver与Reimer，1993）转换为日历年代，并以95%置信度（2σ）匹配Marine13数据集（Reimer等，2013）进行校准，同时针对E27-23岩芯站位的表层水储库年龄约752年进行了校正——该校正值源自海洋放射性碳储库校正数据库与相关软件，可从http://radiocarbon.LDEO.columbia.edu/ 获取（Butzin等，2005）。表格同时提供了海洋碳占比的相关信息。各对应（中间）深度的最终年龄采用中位日历年龄（Cal Yr BP，即日历年龄距今年数）标注。需注意，附录A中所有年龄均以年为单位，但部分为未校准年龄，其余则为Cal Yr BP（即距今年份的日历年龄）。表格表头说明如下：
- 原始14C年龄（yr BP）：指放射性碳测年得到的原始年龄，未施加任何校正，单位为距今年份。
- 校正后原始年龄（RA=752）：指施加了本地储库年龄（Reservoir Age, RA）校正的原始年龄，单位仍为距今年份。
- 其余年龄均为经过校准的距今年份日历年龄。
所有格式均符合放射性碳原始测年数据与校准年代的报告规范。

附表2（附录B）：对比了采用两种方法对已确认14C测年数据进行校准的结果：一是OXCAL 4.2软件（Bronk Ramsey，2009；Blaauw，2010），二是CALIB07软件（Stuiver与Reimer，1993）；两种方法均以95.4%置信度（2σ）匹配Marine13校准曲线（Reimer等，2013），并针对E27-23岩芯站位的表层水储库年龄约752年进行了校正。无论采用何种校准方法，中位Cal Yr BP的校准结果差值均≥40个日历距今年份。本研究采用CALIB07软件的校准输出结果确定年代地层。

附表3（附录C）：E27-23岩芯的有孔虫稳定同位素数据。数据为从>150微米级别的左旋厚壁新方球虫（Neogloboquadrina pachyderma sinistral）浮游有孔虫中测得的氧同位素比值（δ¹⁸O）。同位素值以相对于维也纳佩迪贝蒙石（Vienna Peedee Belemnite, VPDB）的千分比（‰）偏差形式报告。

附表4（附录D）：海洋沉积物岩芯SO136-111的古冬季海冰浓度（winter sea-ice concentration, wSIC）估算结果。岩芯SO136-111的每个样品均标注了以千年前距今年份（kyr BP）为单位的日历年龄。针对该岩芯的每个样品，本研究提供了冬季海冰浓度估算值（%），以及GAM/WSI/13和GAM/WSI/ETS两种模型下，冬季海冰浓度估算值的95%置信区间上下限。最后两列提供了岩芯SO136-111各样品的估算年平均月度海冰覆盖度——该数据最初由Crosta等（2004）采用现代类比法（Modern Analogue Technique）估算得出。此外，本研究还提供了由Crosta等（2004）采用现代类比法估算得到的夏季海表温度。

附表5（附录E）：海洋沉积物岩芯E27-23的古冬季海冰浓度（wSIC）估算结果。岩芯E27-23的每个样品均标注了以千年前距今年份为单位的日历年龄。针对该岩芯的每个样品，本研究提供了冬季海冰浓度估算值（%），以及冬季海冰浓度估算值的95%置信区间上下限。