Varve data of sediment cores SO90-39KG/56KA as stacked record

Data were (1) cleaned (elimination of event deposits, (2) despiked (removing values beyond 2.5 std dev), (3) log-transformed, (4) averaged with 5-yr moving window.数据集依次经过以下预处理步骤：(1) 数据清洗：剔除事件沉积样本；(2) 去尖峰处理：移除超出2.5倍标准差（standard deviation）的数值；(3) 对数变换；(4) 采用5年滑动窗口进行均值计算。