Information on geomagnetic field intensity in the past is essential for understanding the behavior and mechanism of the geodynamo. A fundamental unresolved problem of relative paleointensity (RPI) estimations from marine sediments is that changes in the constituents of magnetic mineral assemblages may influence RPI estimations, called lithological contamination. A negative correlation between RPI and the ratio of anhysteretic remanent magnetization (ARM) susceptibility to saturation isothermal remanent magnetization (kARM/SIRM), which is a proxy for the proportion of magnetofossils to detrital magnetic minerals, was previously reported from deep-sea sediments. This could be caused by lower RPI recording efficiency of the magnetofossil component than the detrital component. To elaborate further this issue, we have conducted a paleo- and rock magnetic study of a sediment core taken from the central north Pacific. RPI estimated from a slope in a diagram plotting a pair of natural remanent magnetization (NRM) and ARM at each alternating-field demagnetization step (NRM-ARM demagnetization diagram) has a negative correlation with kARM/SIRM. Principal component analysis of first-order reversal curve diagrams indicates a downcore increase of the magnetofossil proportion with increased kARM/SIRM. These results reinforce the lower RPI recording efficiency of magnetofossils. In this core, the magnetic coercivity ranges of the magnetofossil and detrital components overlap, which produces a linear NRM-ARM demagnetization diagram. This hinders a possibility of obtaining uncontaminated RPI from a coercivity window representing the magnetofossils or detrital magnetic components, which was tried by some previous studies. A linear NRM-ARM demagnetization diagram, which was sometimes used as a criterion of reliable RPI estimations, does not necessarily mean the absence of lithological contamination to the RPI. In sediments with changing proportion of magnetofossils, normalization with IRM may work better than ARM.Graphical
Read full abstract