Data from the cosmogenic nuclide dating program following AAS 4318 ‘CADAGIA’ field surveys

The CADAGIA program was conducted across four field seasons between 2015-2019 to provide regional-scale record of East Antarctic Ice Sheet change, focussing on the postglacial period (i.e. post ~20 ka BP). Surveys were conducted on nunataks and other ice-free areas between Vestfold Hills and Windmill Islands. Survey and sampling sites were chosen to explore regional ice sheet behaviour. We avoided areas where existing cosmogenic dating programs had been completed (e.g. Bunger Hills, and central-western Vestfold Hills). Detailed mapping and sampling transects were conducted at Windmill Islands, north and north-eastern Vestfold Hills. These were supplemented reconnaissance sampling at remote outcrops across the scatter of ice free areas across the region. Remote locations were accessed via helicopter or fixed wing supported day trips by small field parties. These primarily include sites co-located with deployment of seismic and GPS recording stations (e.g. Snyder Rocks, Mt Strathcona, Mt Brown, Gillies Islands, Carey Nunatak) but were supplemented where feasible to expand geographic coverage (Ravich nunatak, Hippo and Alligator Islands, Mt Sandow) or target key ice-history locations (Cape Jones and Watson Bluff). Cosmogenic sampling was conducted in conjunction with and guided by in-field observations of bedrock and clast weathering (White, 2021) and glacial landforms. Vertical transects were collected from steep nunataks (Cape Jones, Watson Bluff and Mts Brown and Strathcona), and horizontal transects from the large coastal oases (Windmill Islands and Vestfold Hills). The ‘ideal’ sample site was a local summit or an exposed ridgeline, where bedrock displayed evidence of past glacial erosion and was mantled in a thin scatter of erratics. At each site sampling typically consisted of a single piece of surficial bedrock (with striae or glacial polish where [rarely] available) and three cobble-sized, subrounded, subglacial erratics perched on bedrock. Cobbles were selected to minimise potential for inheritance, post-depositional movement, weathering and exhumation, and to maximise quartz content.Sample selection and collection was conducted by several different field scientists (Duanne White, Steven Phipps, Marcello Blaxell and David Small). Duanne White provided guidance and training on sampling methodologies through field examples in Australia (Marcello Blaxell and Steven Phipps) or at Windmill Islands (David Small). Sampling was also conducted following the same set of written guidelines.Sample sites were located with a Garmin e-trex GPS. Elevations were cross-checked with the Reference Elevation Map of Antarctica, accurate to 10 m. Topographic shielding correction was measured at breaks in slope from the perspective of the sample using a sighting compass and clinometer accurate to less than 1 degree. Shielding was often negligible (less than 5 degree). Samples were photographed in the field from several different angles. Photos are available on request.Sample characteristics (Size, lithology, rounding, and surface weathering) were described in the laboratory on return to Australia. Rock disaggregation and quartz purification was then performed at the University of Canberra and Wollongong (Child et al., 2000; Fink and Smith, 2007; Kohl and Nishiizumi, 1992) Beryllium isotopes were extracted from quartz at University of Wollongong and ANSTO in nine separate batches of 8 to 16 samples by Matt Jeromson and Marcello Blaxell, alongside a full chemistry procedural blank. 10/9Be ratios were measured at the Sirius AMS Facility at ANSTO, Australia.Measured 10Be/Be ratios were corrected by the batch-specific procedural blank (ratios of 0.8 to 3.2 x 10-15) and referenced to AMS standards KN-5-2 (nominal ratio 8.55 x 1012) and KN-5-3 (6.32 x 1012). Independent repeat sample measurements were combined as weighted means with the larger of the total statistical error or mean standard error. Final analytical error in concentrations (atoms/gram quartz) are derived from a quadrature sum of the standard mean error in an AMS ratio, 2% for AMS standard reproducibility, 1% in Be spike assay and 1% for quartz mass. Exposure ages are calculated using CRONUS-Earth online calculator, applying the LSDn scaling scheme (Balco et al., 2008). In the absence of a specific Antarctic production rate model, we apply a recent global calibration dataset (Borchers et al., 2016).This version of the dataset (V1.0) archives samples measured by September 2021.4318 cosmo field dataThis worksheet describes the sample characteristics as measured in the field. This is generally recorded at a site rather than sample level.Lat    Long    Locality    All lat/long are recorded relative to the WGS 84 datum. Site elevation (metres)Handheld GPS elevation, relative to sea level. 2015/16 used a barometric corrected GPS (Etrex 300), while other seasons used a standard GPS.SiteIDName used to identify site in respective field book. Most are a simple 3 digit code representing the site. Some represent individual samples (usually with a letter suffix). The final seasons used a XXXX/XXXX sample code to generate redundancy in the sample name and more easily detect potential laboratory transfer errors.Erratics/BedrockNumber of each sample type collected at each sitePosition/LandformCharacteristics of the sample location. Most locations reflect exposure from the ice sheet, but some (809-815, 1173/1174-1183/1184) were sampled to test efficacy of wave washed bedrock as a paleo-sea level record.ShieldingAz, ele, az ele etc. = azimuth and elevation of topographic shielding for the sample or site. Shielding is the fraction of cosmic ray flux, calculated using the Greg Balco’s V2 (April 2018) online shielding tool (http://stoneage.ice-d.org/math/skyline/skyline_in.html) 4318 cosmo field dataThis worksheet describes the sample characteristics as measured in the laboratory. Note that some samples were accidentally crushed prior to description, while others may have been described but not yet processed for Be-10 measurement.Sample IDSite ID, plus a letter-suffix to indicate the individual sample name. Letters a-g typically indicate erratics, while a br suffix indicates bedrock. A, B, C axis.Size of the sample in cm, measured using a ruler to the nearest half-cm.LithologySample rock type. Minerals visually identified. Q = quartz, F = feldspar, G = garnet, B = biotite, M = muscovite, A = amphiboleRoundingPowers (1953) roundness scale, A = angular, SA = subangular, SR = subrounded, R = rounded, WR = well rounded.Clast surface characteristicsA selection of surface characteristics used to measure degree of weathering and relative exposure age. Classified as absent (a) present (p) or common (c). ‘cm lost’ is a subjective estimate of how much material may have been lost from the clast by subaerial crumbling or weathering, based on the surface features and comparison with a typical subglacial clast morphology.4318 AMS and chem dataThis worksheet describes the measurements conducted during Be-10 extraction and Accelerator Mass Spectrometry measurement. Samples are described in measurement batches. Sample IDAs per cosmo lab sample description sheet. Site ID, plus a letter-suffix to indicate the individual sample name. Letters “a” to “g” typically indicate erratics, while a “br” suffix indicates bedrock. UOW Lab ID - University of Wollongong lab numberANSTO Be Cathode ID – ANSTO lab number specific the Be-10 measurementANSTO OZ code – ANSTO laboratory number.Total Qtz (g) – quartz mass dissolved for Be-10 measurement    10Be/9Be    d[10Be/9Be] – ‘raw’ measured Be-10/9 ratio measured via AMS (and uncertainty)Blank ID(s)    Blank 10Be/9Be    Blank d[10Be/9Be] – Be-10/9 ratios for the procedural blank measured in each batch of samples            Corr 10Be/9Be    Corr d[10Be/9Be] – corrected Be-10/9 ratio, subtracted for specific procedural blank for that batch.        Carrier     Be Carrier (µg/g)    d[Be Carrier]    Carrier (g)    9Be Spike (µg)    9Be (atoms)                        Details of the 9Be carrier added prior to quartz dissolution. Carrier name, 9Be concentration (and uncertainty), carrier solution mass, and finally the calculated number of 9Be atoms added to the sample.10Be SRM-4325 (atoms.g-1)    d[10Be]    d[10Be]%Calculated Be-10 concentrations and uncertainties in each sample, referenced against NIST SRM-4325 with an assumed isotope ratio of 2.79 x 10-11.            AMS standard    AMS std nominal value     Actual AMS standards used during each sample run, and the assumed value. AMS run date = date Be10/9 oxide was measured.4318 Cronus entry data and agesData used to calculated exposure ages via the Cronus Earth online calculator. See documentation in Balco et al, 2008 and https://hess.ess.washington.edu/ for description of what each column represents. Thickness is a mix of different measurements a best possible for each sample (uncertainties black = measured in lab (0.5 cm), blue = estimate from photos (1 cm), red = estimate from similar samples (2 cm)). Density assumed to be 2.7 in all samples given quartz rich metamorphic lithologies. Be-10 at/g    Be-10 at/g uncertaintySample ages (at last!) and uncertainties. LSDn scaling scheme.

CADAGIA项目于2015-2019年间开展了4次野外作业，旨在获取东南极冰盖变化的区域尺度记录，研究重点为冰消期（即约2万年BP以来的时段）。野外调查在韦斯特福尔德丘陵（Vestfold Hills）与风车群岛（Windmill Islands）之间的冰原岛峰（nunatak）及其他无冰区开展。选点旨在探究区域冰盖行为特征，避开了已开展宇宙成因测年（cosmogenic dating）项目的区域（如邦格丘陵、韦斯特福尔德丘陵中西部）。

研究团队在风车群岛、韦斯特福尔德丘陵北部及东北部开展了详细测绘与采样断面工作，并补充对区域内零散无冰区的偏远露头进行了踏勘采样。偏远点位由小型野外分队通过直升机或固定翼飞机支持的日间行程抵达，主要包括与地震及全球定位系统（GPS, Global Positioning System）台站部署点位重合的区域（如斯奈德岩、斯特拉思科纳山、布朗山、吉利斯群岛、凯里冰原岛峰），并在可行范围内补充点位以扩大地理覆盖范围（如拉维奇冰原岛峰、河马与短吻鳄群岛、桑多山），或针对关键冰历史点位（如琼斯角与沃森断崖）进行采样。

宇宙成因采样结合并参考了野外观测的基岩与岩屑风化情况（White, 2021）及冰川地貌特征。从陡峭冰原岛峰（琼斯角、沃森断崖、布朗山与斯特拉思科纳山）采集垂直断面，从大型沿海绿洲（风车群岛与韦斯特福尔德丘陵）采集水平断面。“理想”采样点位为局部峰顶或裸露山脊，此处基岩显示出古冰川侵蚀痕迹，并被少量漂砾覆盖。每个点位的采样通常包括一块表层基岩（若存在擦痕或冰川磨光面则优先选取，此类情况较为罕见），以及三块嵌于基岩之上、砾石大小、次棱角状的冰下漂砾。选取砾石时尽可能降低继承效应、沉积后移动、风化及抬升暴露的潜在影响，并最大化石英含量。

采样选取与采集工作由多名野外科学家完成（杜安·怀特、史蒂文·菲普斯、马塞洛·布拉克斯尔与戴维·斯莫尔）。杜安·怀特通过澳大利亚境内的野外实例（马塞洛·布拉克斯尔与史蒂文·菲普斯）或风车群岛的野外实操（戴维·斯莫尔）为团队提供采样方法的指导与培训，所有采样工作均遵循统一的书面指南开展。

采样点位通过佳明e-trex GPS（Garmin e-trex GPS）定位，高程通过《南极洲参考高程图》交叉验证，精度为10米。2015/16季采用经气压校正的GPS（Etrex 300），其余季采用标准GPS设备。点位编号（SiteID）为用于在对应野外记录本中标识点位的编号，多数为简单的3位代码；部分编号代表单个样品（通常带有字母后缀）。最后一季采用XXXX/XXXX格式的样品编号，以提升采样名称的冗余性，更易检测实验室传输过程中可能出现的错误。

每个点位采集的漂砾/基岩采样数量、点位/地貌特征、地形遮蔽等信息均有详细记录。其中，点位/地貌特征多数反映了冰盖暴露后的状态，但部分点位（809-815、1173/1174-1183/1184）的采样旨在验证经海浪冲刷的基岩作为古海平面记录的有效性。地形遮蔽的方位角与仰角通过瞄准罗盘与测斜仪（精度优于1°）从采样点视角测量坡度断点处的遮蔽情况计算得到，遮蔽效应为宇宙射线通量的衰减比例，可通过格雷·巴尔科的V2版（2018年4月）在线遮蔽工具计算（http://stoneage.ice-d.org/math/skyline/skyline_in.html）。多数情况下遮蔽效应可忽略（小于5°）。采样在野外从多个角度拍摄照片，照片可按需获取。

返回澳大利亚后，研究团队在实验室对采样特征（尺寸、岩性、磨圆度与表面风化情况）进行描述。随后在堪培拉大学与伍伦贡大学完成岩石解离与石英纯化工作（Child et al., 2000; Fink and Smith, 2007; Kohl and Nishiizumi, 1992）。由马特·杰罗森与马塞洛·布拉克斯尔在伍伦贡大学及澳大利亚核科学技术组织（ANSTO, Australian Nuclear Science and Technology Organisation）从石英中提取铍同位素，共分为9批，每批包含8至16个样品，同时设置完整的化学流程空白对照。10/9铍比值在澳大利亚ANSTO的Sirius加速器质谱（AMS, Accelerator Mass Spectrometry）设施中完成测量。

实测10Be/Be比值通过批次专属的流程空白（比值范围0.8至3.2×10^-15）进行校正，并参照AMS标准物质KN-5-2（标称比值8.55×10^12）与KN-5-3（6.32×10^12）进行校准。独立重复样品的测量结果以加权平均值合并，取总统计误差与平均标准误差中的较大值作为误差项。浓度（原子/克石英）的最终分析误差由以下项的正交和计算得到：AMS比值的标准平均误差、AMS标准物质重现性的2%误差、铍标样测定的1%误差，以及石英质量的1%误差。暴露年龄通过CRONUS-Earth在线计算器计算，采用LSDn标度方案（Balco et al., 2008）。由于缺乏专属的南极生产速率模型，本研究采用近期发布的全球校准数据集（Borchers et al., 2016）。

本数据集（V1.0）归档了截至2021年9月的所有实测样品。

#### 4318 宇宙成因野外数据表
本工作表描述了野外实测的采样特征，通常按点位而非单一样本级别记录。
- 纬度（Lat）、经度（Long）、点位名称（Locality）：所有经纬度均基于WGS 84基准面记录。
- 点位高程（Site elevation, 单位：米）：手持GPS测得的相对于海平面的高程。2015/16季采用经气压校正的GPS（Etrex 300），其余季采用标准GPS设备。
- 点位编号（SiteID）：用于在对应野外记录本中标识点位的编号，多数为简单的3位代码；部分编号代表单个样品（通常带有字母后缀）。最后一季采用XXXX/XXXX格式的样品编号，以提升采样名称的冗余性，更易检测实验室传输过程中可能出现的错误。
- 漂砾/基岩采样数量（Erratics/Bedrock）：每个点位采集的各类样品数量。
- 点位/地貌特征（Position/Landform）：采样点位的特征描述，多数点位反映了冰盖暴露后的状态，但部分点位（809-815、1173/1174-1183/1184）的采样旨在验证经海浪冲刷的基岩作为古海平面记录的有效性。
- 地形遮蔽（Shielding）：Az, ele, az ele等参数代表采样点或点位的地形遮蔽方位角与仰角。遮蔽效应为宇宙射线通量的衰减比例，通过格雷·巴尔科的V2版（2018年4月）在线遮蔽工具计算（http://stoneage.ice-d.org/math/skyline/skyline_in.html）。

#### 4318 实验室采样特征数据表
本工作表描述了实验室实测的采样特征。需注意，部分样品在描述前意外被粉碎，另有部分样品仅完成描述但尚未开展铍-10测量处理。
- 样品编号（Sample ID）：点位编号加字母后缀，用于标识单个样品。字母a-g通常代表漂砾，br后缀代表基岩。
- 样品三轴尺寸（A, B, C axis）：以厘米为单位的样品尺寸，使用直尺测量至最近0.5厘米。
- 岩性（Lithology）：通过目视鉴定的岩石矿物组成，对应代码：Q=石英（quartz）、F=长石（feldspar）、G=石榴子石（garnet）、B=黑云母（biotite）、M=白云母（muscovite）、A=角闪石（amphibole）。
- 磨圆度（Rounding）：采用Powers（1953）磨圆度分级标准：A=棱角状（angular）、SA=次棱角状（subangular）、SR=次圆状（subrounded）、R=圆状（rounded）、WR=圆状极好（well rounded）。
- 岩屑表面特征（Clast surface characteristics）：用于表征风化程度与相对暴露年龄的表面特征，分为缺失（a）、存在（p）或普遍（c）三类。“cm lost”为基于表面特征并对比典型冰下岩屑形态得到的、岩屑因地表风化或崩解损失的物质厚度的主观估算值。

#### 4318 加速器质谱与化学分析数据表
本工作表描述了铍-10提取与加速器质谱（AMS, Accelerator Mass Spectrometry）测量的全流程数据，样品按测量批次组织。
- 样品编号（Sample ID）：与实验室采样描述表一致，即点位编号加字母后缀，字母a-g通常代表漂砾，br后缀代表基岩。
- 伍伦贡大学实验室编号（UOW Lab ID）：伍伦贡大学实验室的样品编号。
- ANSTO铍阴极编号（ANSTO Be Cathode ID）：澳大利亚核科学技术组织专属的铍-10测量实验室编号。
- ANSTO OZ代码（ANSTO OZ code）：ANSTO实验室的通用编号。
- 总石英质量（Total Qtz, 单位：克）：用于铍-10分析的溶解石英总质量。
- 原始10Be/9Be比值与不确定度（10Be/9Be、d[10Be/9Be]）：通过AMS直接测得的原始10Be/9比值及其测量不确定度。
- 流程空白参数（Blank ID(s)、Blank 10Be/9Be、Blank d[10Be/9Be]）：对应批次流程空白对照的10Be/9比值及其不确定度。
- 校正后10Be/9Be比值与不确定度（Corr 10Be/9Be、Corr d[10Be/9Be]）：扣除对应批次流程空白影响后的校正10Be/9比值及其不确定度。
- 载体添加参数：包含载体名称、9Be载体浓度（µg/g）、浓度不确定度、载体溶液质量（g）、9Be spike添加量（µg）及添加的9Be原子总数，为石英溶解前添加的9Be载体的完整信息。
- 铍-10浓度与不确定度（10Be SRM-4325 (atoms.g-1)、d[10Be]、d[10Be]%）：基于美国国家标准与技术研究院（NIST）SRM-4325标准物质（假定同位素比值2.79×10^-11）计算得到的样品铍-10浓度及其不确定度。
- AMS标准物质与标称值（AMS standard、AMS std nominal value）：本次测量使用的AMS标准物质及其假定标称值。
- AMS测量日期（AMS run date）：Be10/9氧化物的质谱测量日期。

#### 4318 CRONUS暴露年龄输入数据与年龄数据表
本工作表包含通过CRONUS-Earth在线计算器计算暴露年龄所需的全部输入数据，各列含义详见Balco et al.（2008）及https://hess.ess.washington.edu/公开文档。
- 样品厚度：采用多源数据综合得到的最优值，误差标记规则：黑色=实验室实测（0.5 cm）、蓝色=基于照片估算（1 cm）、红色=基于相似样品类比估算（2 cm）。所有样品密度均假定为2.7，因岩性以富石英变质岩为主。
- 铍-10浓度与不确定度（Be-10 at/g、Be-10 at/g uncertainty）：校正后的样品铍-10浓度及其测量不确定度。
- 最终暴露年龄与不确定度：采用LSDn标度方案计算得到的样品暴露年龄及其总不确定度。