中文摘要：

为了鉴别长期风化的沉底油,以1种原油（A）和2种燃油（B和C）为研究对象,采用重复性限法筛选340 d风化的水面漂浮油和水下沉底油的稳定诊断比,并基于这些诊断比对沉底油进行聚类分析。结果表明,诊断比的稳定性不仅和油种、风化时间有关,还和溢油的存在形态密切相关;轻组分含量相对高的A和B的沉底油比它们对应的漂浮油受到的风化影响更大,而重组分含量高的C正好相反,研究最终得到4个适合鉴别长期风化漂浮油和沉底油的稳定诊断比;聚类分析法可将沉底油与其油源很好地聚类,并可反映沉底油的风化程度。因此,基于稳定诊断比的聚类分析法可用于鉴别长期风化的沉底油,值得进一步推广。

英文摘要：

In order to identify long-term weathered sunken oil, research was conducted with one kind of crude oil （A） and two kinds of fuel oil （B and C） as objects. The stable diagnostic ratios of floating oils and underwater sunken oils which underwent 340-day weathering were screened by repeatability limit method. Moreover, sunken oils were identified by clustering analysis based on these diagnostic ratios. The study suggested-that the stability of diagnostic ratios was not only related to oil types and weathering time, but also closely related to existence forms of spill oil. A and B sunken oils which had more light components tended to be more affected by weathering than the relevant floating oils, but C which had more heavy components was just the opposite. Eventually, four stable diagnostic ratios suitable for identifying floating and sunken oils, which underwent long-term weathering, were obtained; and the clustering analysis could better cluster sunken oils and their sources, and reflect their weathering degree. In conclusion, clustering analysis based on stable diagnostic ratios could be used for identification of long-term weathered sunken oils.