Measurements of NRM after 20-mT demagnetization agree very well with shipboard measurements for the top 3 m (Fig. F1). Below 3 mcd, the measurements begin to diverge and below 12 mcd, the intensity of the U-channels is an order of magnitude weaker than that of shipboard measurements, indicating severe postcoring alteration. This decrease in intensity is attributed to the dissolution of fine magnetic grains during storage (Yamazaki et al., 2000).

ARM is sensitive to fine magnetic grains (0.05 to ~1 Ám) of low-coercivity minerals such as magnetite and maghemite. At 3 mcd, a severe two-orders-of-magnitude drop in ARM is observed (Fig. F2). SIRM at 1 T is a measure of the total content of remanence-carrying grains (magnetite, maghemite, and hematite, for example). Again, the sharp drop at 3 mcd is evident (Fig. F3). By applying an isothermal remanent magnetization (IRM) of -0.3 T, the low-coercivity grains such as magnetite and maghemite are magnetized in the opposite direction of the high-coercivity grains (hematite and goethite) (Fig. F4). The S-ratio (SIRM [+1 T]/IRM [-0.3 T]) is a measure of the downcore variations in high- and low-coercivity minerals. High S-ratio values indicate more low-coercivity minerals such as magnetite, whereas lower values indicate more high-coercivity minerals. The S-ratio is above 0.9 from 0 to 2.5 mcd, spikes sharply to 0.5 at 2.5 mcd, and ranges from 0.8 to 0.85 below 2.5 mcd. (Fig. F5). This is consistent with the loss of fine-grained magnetite as noted by Yamazaki et al. (2000). Further evidence of this loss can be seen by normalizing ARM (sensitive to grain-size and concentration) to SIRM (sensitive to concentration) to get a parameter sensitive to grain-size. Figure F6 shows a 50% drop in the ARM/SIRM parameter, indicating a loss of fine-grained material. This intense dissolution of fine-grained magnetic material inhibits paleoclimate interpretations from these measurements.