Dome Fuji Ice Core Preliminary Temperature Reconstruction, 0-340 kyr

Preliminary reconstruction of air temperature at Dome Fuji (Antarctica) for the past 340,000 years from isotopic data (d18O and dD) of the first Dome Fuji ice core. The Dome Fuji temperature (delta-Tsite, deviation from the mean of the last 10 kyr) was reconstructed by correcting raw isotopic records of the ice core for vapor source temperature and seawater isotopic composition variations, using an isotopic inversion (Vimeux et al., 2002; Stenni et al., 2004). Due to high noise level in the deuterium excess data especially between 20 and 80 ka, we use a filtered record (passband&gt;5000yr, stopband&lt;2000yr). Note that this data is not intended to resolve millennial changes because of the filtering of the raw data as well as averaging of delta-Tsite to 500-yr intervals. We used a recent seawater d18O reconstruction by Bintanja et al (2005, Nature). We estimate the sensitivities of isotopic values (Uemura et al., 2004) to site (Dome Fuji) and source temperatures and seawater isotopic composition, by using a Rayleigh-based model. We tune this model with the Dome Fuji present-day surface data. We currently lack isotopic data of surface snow from the coast to Dome Fuji to define the air mass trajectories, thus this data should be considered preliminary. Nevertheless, the tuned sensitivities are close to those for Vostok and Dome C (Vimeux et al., 2002; Stenni et al., 2004). The data are put on DFO-2006 age scale and averaged at 500 yr intervals. The value is the mean in the range from 'Top age' to the next 'Top age'. For plotting the data, use staircase (cityscape) function using 'Top age', or linear function using 'Center age'. Note that age is in b2k (before AD 2000).我们基于首支南极富士穹顶（Dome Fuji）冰芯的同位素数据（δ¹⁸O与δD），完成了过去34万年富士穹顶气温的初步重建。富士穹顶气温序列（ΔT_site，站点温度距平，即相对于过去10千年均值的偏差）通过同位素反演（isotopic inversion）方法重建：针对冰芯原始同位素记录，校正水汽源区温度与海水同位素组成的变化，所用方法参考Vimeux等人（2002）及Stenni等人（2004）的研究。由于氘过量（deuterium excess）数据存在较高噪声，尤其在20~80 ka区间内，我们采用了滤波后序列（通带>5000年，阻带<2000年）。需注意，由于对原始数据进行了滤波处理，且将ΔT_site序列平均至500年间隔，本数据集无法解析千年尺度的气候变化。本研究采用了Bintanja等人2005年发表于《自然》（Nature）期刊的海水δ¹⁸O重建序列。我们基于瑞利分馏模型（Rayleigh-based model），估算了同位素值对站点（富士穹顶）、源区温度及海水同位素组成的敏感性，相关参数参考Uemura等人（2004）的研究，并利用富士穹顶现今地表观测数据对该模型进行校准。目前我们缺乏从海岸至富士穹顶的地表雪层同位素数据，无法确定气团轨迹（air mass trajectories），因此本数据集仍属于初步成果。尽管如此，校准后的敏感性参数与沃斯托克站（Vostok）及C穹顶（Dome C）的相关参数较为接近，参考结果同Vimeux等人（2002）及Stenni等人（2004）的研究。本数据集采用DFO-2006年代标尺，且已平均至500年间隔。数据值为'Top age'至下一个'Top age'区间内的平均值。绘图时，若以'Top age'为基准，可使用阶梯（cityscape）绘图函数；若以'Center age'为基准，则可使用线性绘图函数。请注意，年代单位为b2k（即公元2000年之前）。