中文摘要：

中国黄土的剩磁获得机理和Lock—in深度问题是目前黄土古地磁研究的焦点和核心问题之一，它直接关系到黄土时间标尺的精确度和海陆古气候对比的可靠性。文章以黄土高原东南缘三门峡曹村剖面获得的高分辨率古地磁极性界线作为年代控制点，通过黄土-古土壤序列1．1Ma以来的磁气候记录与近年来中、高纬度大西洋ODP／DSDP钻孔中获得的具有高分辨率古地磁年代约束的深海氧同位素记录进行了详细对比分析，提出与传统的海陆对比不同的基于古地磁界线的海陆对比方案：松山／布容地磁极性转换界线记录在曹村剖面S8的顶部，可对应深海沉积的MIS19，而不是前人认为的MIS21。黄土高原不同地区记录东亚冬、夏季风的差异性和变化性可能是全球具同时性的松山／布容地磁极性转换在黄土高原不同地区产出层位不一致性的主要原因，而基于Lock-in模型的天文年代标尺则可能高估了黄土／古土壤的年龄。

英文摘要：

The recording inconsistencies of magnetic polarity reversals among many Chinese loess sections and between the loess and the marine records have hampered precise section-to-section and land-ocean age comparisons. To better assess the fidelity of polarity reversals and precisely estimate the extent of lock-in depth of the remanence magnetization in Chinese loess, we present a high-resolution paleomagnetic and detailed mineral magnetic investigation of a loess section at Sanmenxia. The Sanmenxia loess section （34°38′N, 111°09′E） is located at the southeastern margin of the CLP, 15km southwest of Shaanxian County, Henan Province. Oriented samples were collected at a 10cm interval from the top of Lt to the upper part of L13. From L8 to the upper part of S8 and from the bottom of S9 to L13, samples were continuously collected to get the complete Matuyama/Brunhes（MB） reversal and Jaramillo Normal Subchron （JNS）. In laboratory, every independently oriented sample was cut into two equal specimens（named sets A and B）. The paleomagnetic results based on set A speciments indicate that MB occurs at the top of soil S8 , and the upper and lower boundaries of JNS are encountered at the top of soil S10 and the boundary of S12/L13, respectively. For all set B speciments, the following mineral magnetic characteristics were determined： （1） magnetic susceptibility determined on a Bartington MS2B at 0.47kHZ frequency; （2）SIRM imparted in a DC field of 2 Tesla by a Redcliffe pulse magnetizer; （3）reversed-field IRM induced in DC field of 0.1T to get the ratio S-0.1 = IRM -0.1T/SIRM2T. We then reveal a polarity-based correlation of magnetic susceptibility and S-0.1 records from Sanmenxia to several well-documented oxygen-isotope and magnetostratigraphic records from the North Atlantic, e.g., ODP sites 982, 983, and 984, and DSDP site 607 over the last 1.1 Ma. Different from previous correlations, we propose here that S8 in Sanmenxia should correspond to MIS 19, the thin S7 to a distinct second-orde