Key points are not available for this paper at this time.
We use bivariate scatter plots to illustrate variations in selected rock magnetic properties (low‐field susceptibility, anhysteretic and isothermal remanence) of late Neogene and Pleistocene deep‐sea sediments from 16 sites in the Arctic Ocean, North Atlantic, equatorial Atlantic and North Pacific Ocean, and the Arabian Sea. Our intention is to examine the ability of the rock magnetic properties to differentiate the sediments according to factors such as lithology, geographical area, and the dominant mode of terrigenous sedimentation, which at these sites is via ice‐rafting, via bottom currents, or via eolian processes. Overall, correlations between sediment magnetic properties and gross lithology is poor, and factors such as the source and transport path of terrigenous sediment (and detrital magnetic minerals), together with the action of reductive diagenetic processes, are the major controls on the magnetic properties. On the bivariate scatter plots, sites with major ice‐rafted contributions tend to have high sedimentary ferrimagnetic mineral concentrations, relatively coarse ferrimagnetic grain‐sizes, and scattered sample point distributions; in contrast, sites where we infer significant bottom‐current supply of terrigenous material have tightly grouped sample point distributions. Carbonate sediments in which the terrigenous component is supplied by eolian processes tend to have a broad range of magnetic mineral concentration, caused by glacial‐interglacial fluctuations in carbonate accumulation and eolian activity. Sediments containing significant volcanogenic material have high concentrations of relatively coarse‐grained ferrimagnetic material. Reductive diagenesis is a significant determinant of sediment magnetic properties in high‐productivity areas and has the effect of preferentially removing the fine‐grained ferrimagnetic fraction, causing a coarsening of the ferrimagnetic grain‐size distribution and a rise in the antiferromagnetic:ferrimagnetic ratio.
Bloemendal et al. (Fri,) studied this question.