A Common Era reconstruction of the Interdecadal Pacific Oscillation from the Law Dome ice core, East Antarctica.

This dataset is an annual reconstruction of the Interdecadal Pacific Oscillation (IPO), a decadal-scale mode of variability in the Pacific Ocean which has climate impacts across the Pacific Basin. This data is a time series spanning CE 1-2011 inclusive (ie, the Common Era). The time series is reconstructed from three primary annually-resolved proxy series from the Law Dome ice core. These three series are the log-transformed seasonal sea salt concentration for the cool season (June to November), the log-transformed seasonal sea salt concentration for the warm season (December to May) and the annual snowfall accumulation rate. The reconstruction uses a Gaussian kernel correlation reconstruction method (Roberts et al., 2019) with 2000 ensemble members, which provides a mean IPO index value for each year, as well as upper and lower quartiles. The reconstruction target time series was the observed Interdecadal Pacific Oscillation spanning 1870-2020, which had been smoothed using a Gaussian window of 13 years. This Gaussian kernel correlation reconstruction is an evolution/replacement of the method and reconstruction presented in Vance et al., (2015) to reconstruct the IPO. This is now our preferred dataset for the Law Dome IPO reconstruction, and supersedes that published by Vance et al., (2015). The time series (dataset) consists of three columns with column headings as follows: Year – where year is the year from the beginning of the Common Era, ie, ‘436.0’ means the year CE 436, and ‘2009.0’ means the year 2009.IPO (mean) – the mean of the IPO reconstruction index value Std Dev) – the standard deviation of the index value for each year.

本数据集为太平洋年代际振荡（Interdecadal Pacific Oscillation, IPO）的年际重建序列。太平洋年代际振荡是太平洋海域的年代际尺度变率模态，其气候影响覆盖整个太平洋海盆。本数据集的时间序列覆盖公元1年至2011年（含首尾两年），即公元纪年（Common Era）范畴内的完整时段。该时间序列基于劳德姆冰穹（Law Dome）冰芯的三个主要年分辨率代用序列重建，这三个代用序列分别为：冷季（6月至11月）经对数变换的季节性海盐浓度、暖季（12月至次年5月）经对数变换的季节性海盐浓度，以及年降雪积累速率。本次重建采用高斯核相关重建方法（Gaussian kernel correlation reconstruction method, Roberts等，2019），共包含2000个集合成员，可输出各年份的平均IPO指数值以及上下四分位数。本次重建的目标时间序列为1870-2020年的实测太平洋年代际振荡序列，该序列已通过13年高斯窗口进行平滑处理。该高斯核相关重建方法是对Vance等（2015）提出的原有IPO重建方法的改进与替代方案，目前为本研究团队推荐的劳德姆冰穹IPO重建数据集，替代了Vance等（2015）发布的原有研究成果。本时间序列（数据集）包含三列数据，列标题如下：1. 年份（Year）：对应公元纪年的年份，例如"436.0"代表公元436年，"2009.0"代表公元2009年；2. IPO平均值（IPO (mean)）：该年份IPO重建指数的均值；3. 标准差（Std Dev)）：对应年份指数值的标准差。