Dansgaard-Oeschger（D-O）事件是指发生于末次冰期晚期的一系列具有百年至千年尺度的气候突变事件，在全球范围均有记录，产生原因仍不清楚。有研究指出，全新世气候的不稳定性在千年尺度上类似于末次冰期。以往分析D-O事件的时间序列分析方法多以线性理论为基础并且包含了对数据的平稳性假设，而来自格陵兰岛NGRIP冰芯的氧同位素记录却表现出显著的非线性特征，这可能是导致多种分析结果的主要原因。申请者基于格陵兰岛与南极洲的冰芯资料发现D-O事件包含显著的非线性关系与分形特征。本基金希望研究不同纬度带的古气候资料在D-O事件期间的多重分形结构，包括数据的长记忆性、标度突变现象以及多重分形谱等，分析不同纬度带的气候变化在D-O事件期间的多标度特性，定量评估D-O事件的全球特征，为发展古气候模式提供一定的理论依据。

Dansgaard-Oeschger (D-O) events, a series of abrupt climate events on centennial-to-millennial timescales, are found globally and still controversial. It was suggested that millennial-scale climatic instability is not restricted to the last glacial period but is a pervasive feature of Holocene climatic instability. Predominate time series analysis methods applied to D-O events are on the basis of linear theory and assumption that the data is stationary, whereas, oxygen isotope record from the NGRIP ice core from Greenland shows significant nonlinearity, which may be the main cause of diversity of results. Our recent research found significant nonlinearities and fractal characters in the ice core records from both Greenland and Antarctic within D-O events. This project aims to explore the multi-fractal structures of paleoclimate records from different latitudes during D-O events, including the long-term correlation, the scaling mutation and the multi-fractal spectrum. We will also quantitatively estimate the global character of D-O events through the multi-scale law of climate variability for a more plausible plaeoclimate model.