The Matuyama-Brunhes polarity reversal in four Chinese loess records: high-fidelity recording of geomagnetic field behavior or a less than reliable chronostratigraphic marker?

2014 
Abstract The paleomagnetic recording fidelity of Chinese loess has been a matter of long-term dispute. To help shed light on the acquisition/retention mechanisms of natural remanent magnetization in loess, we report on high-sampling-resolution paleomagnetic constraints on the position of the Matuyama–Brunhes Boundary (MBB) in four typical loess sections in different marginal areas of the Chinese Loess Plateau (CLP), each of which has experienced different East Asian winter/summer monsoon strengths and pedogenic intensities. Detailed thermal demagnetization experiments for 1850 specimens reveal that none of the high-accumulation-rate loess with the weakest pedogenesis at Caotan (western CLP), the massive Pedogenesis loess characterized by extensive development of thick calcareous nodule layers at Jixian (SE CLP), and the moderately pedogenized loess at Fanshan (NE CLP) have preserved high-fidelity geomagnetic records across the MBBs. Jiuzhoutai (western CLP), one of the thickest loess profiles in the world, contains the thickest, and hitherto the most complex MBB that is too extensive to be regarded as rapid fluctuations of geomagnetic field during the M/B reversal. At Jixian, a predominance of pedogenesis-induced chemical remanent magnetization overprinting has heavily obscured the original remanence. The significant downward displacement of the MBB at Fanshan and even its total absence at Caotan, together with extensive remagnetization of the coarse-grained ‘upper sandy loess layer’ (L9) in both sections, further suggest that the processes of remanence acquisition/retention of Chinese loess in the marginal areas of the CLP are rather complicated. Thus, interpretations of the fine structure of geomagnetic transitional field behavior and ‘excursions’, especially in the Matuyama Chron, should be regarded with caution.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    72
    References
    20
    Citations
    NaN
    KQI
    []