The rock-magnetic and paleomagnetic investigations at Site 1200 were conducted to document the magnetic mineralogy, evaluate the potential contribution of serpentinized ultramafic rocks to the magnetic anomalies in forearc regions and the magnetic signatures of seamounts, investigate the magnetic properties of the serpentine sediments in comparison to the properties of the ultramafic parent rock, and potentially reorient the recovered rock fragments into a common reference frame. The process of serpentinization produces secondary magnetite by the alteration of primary olivine, orthopyroxene, clinopyroxene, and spinel. Because the composition and concentration of the iron oxide phase should be related to the degree of serpentinization, measurements of the magnetic properties of rocks and serpentine muds from Site 1200 may contribute to our understanding of the serpentinization process.
Magnetic properties and the potential contribution of serpentinized oceanic mantle rocks to magnetic anomalies have been examined by several investigators, including Stokking et al. (1990), Kelso et al. (1996), and Kikawa et al. (1996), and the magnetic properties of serpentinites from ophiolite complexes have been studied by Beske-Diehl and Banerjee (1980), Luyendyk and Day (1982), and Swift and Johnson (1984). However, with the exception of the study of Stokking et al. (1990), little is known about the magnetic properties of serpentinites in forearc settings.
Magnetic measurements (alternating-field [AF] demagnetization and magnetic susceptibility) were performed on 26 oriented ultramafic minicore and minicube samples from Hole 1200A, 4 unoriented samples originally taken for physical properties measurements, and 11 oriented cubes of serpentine sediment samples from Holes 1200D and 1200E. AF demagnetizations of discrete samples were routinely performed at steps of 0, 5, 10, 15, 20, 25, 30, 35, 40, 50, 60, 70, and 80 mT. The results of these measurements are displayed in Table T6. In addition, small rock chips were used to study thermal demagnetization of isothermal remanent magnetization (IRM), for IRM acquisition experiments, and measurements of the backfield coercivity of saturation remanence. To evaluate the significance and consistency of the characteristic remanent magnetization (ChRM) obtained from the ultramafic clasts, a 72-cm-long hard rock piece was stepwise demagnetized to 80 mT in 2-cm intervals.
The natural remanent magnetization (NRM) intensities of the ultramafic rocks range from 0.03 to 2.66 A/m (arithmetic mean = 0.49 A/m); values are displayed in Table T6 and Figure F43. These values are slightly higher but in a similar range as measurements reported in similar material from Conical and Torishima Forearc Seamounts by Stokking et al. (1990), which range from 0.01 to 0.59 A/m (arithmetic mean = 0.20 A/m). The volume magnetic susceptibilities of the serpentinites in Hole 1200A (Table T6; Fig. F43) range from 1.65 x 10-3 to 1.63 x 10-2 (arithmetic mean = 5.58 x 10-3). These results are also very similar to the data from Conical Seamount (0.05 x 10-3 to 9.78 x 10-3; arithmetic mean = 3.83 x 10-3). Both the magnetic intensity and magnetic susceptibility are very high, and the positive linear correlation between both parameters (r = 0.73) suggests that both signals are dominated by the concentration variation of the magnetic carrier mineral.
The magnetic properties of the serpentine mud are similar to the properties of the parent rocks. The NRM intensity ranges from 0.15 to 0.95 A/m (arithmetic mean = 0.44 A/m), and volume magnetic susceptibilities vary between 2.32 x 10-3 and 1.18 x 10-2 (arithmetic mean = 6.81 x 10-3).
The mineralogy of the phases that contribute to the magnetization and magnetic susceptibility was examined through magnetic studies, including AF demagnetization of the NRM and IRM, Curie temperature determinations, the acquisition of an IRM, and the measurement of the backfield coercivity of saturation remanence (Bcr).
Thermal demagnetization of a few selected rock chips shows a near-linear decrease of magnetic intensity between room temperature and ~500°C, above which the demagnetization rate increases (Fig. F44). All measured samples have Curie temperatures (Tc) of 585°C, which suggests that the dominant magnetic mineral is pure magnetite (Tc = 585°C). Minor substitution of other elements (e.g., Ti, Al, or Mg) decreases the Curie temperature of magnetite. No other sharp decrease in magnetization, which would suggest that another mineral is contributing to the NRM, occurs. Therefore, it is likely that very pure magnetite is the dominant carrier of remanence and susceptibility. The magnetite is predominantly a secondary mineral formed during serpentinization.
The samples that exhibit very uniform IRM acquisition behavior were saturated by 400 mT and have Bcr values of 12-58 mT (average = 25.3 mT) (Fig. F45), which are virtually identical to the Bcr values reported by Stokking et al. (1990) of 14-50 mT.
The Koenigsberger (QNRM) ratio compares the relative importance of the remanent magnetization with the magnetization induced in the formation by the Earth's magnetic field. Koenigsberger ratios were calculated using the following equation:
QNRM = MNRM/(k x H),where
The value for H used is 31.83 A/m (40,000 nT). The QNRM ratio is used to determine whether the in situ magnetization of the crustal section is dominated by remanent magnetization (QNRM > 1) or an induced component parallel to the current field (QNRM < 1). The Koenigsberger ratios of the hard rock clasts (Table T6; Fig. F43) vary from 0.4 to 8.0 (mean = 2.4), indicating that the in situ magnetization of most of these rocks (26 out of 30) is larger than their induced magnetization. Thus, the in situ magnetization is dominated by remanent magnetization rather than an instantaneous magnetization induced by the Earth's magnetic field.
The samples were easily demagnetized by AF techniques, as illustrated in Figures F46 and F47. The decay of remanent magnetization intensity during progressive demagnetization of most ultramafic clasts is very similar and displays little variability (Fig. F46). The median destructive field (MDF), the field required to reduce the NRM to one-half, is very low throughout the recovered section and ranges from 3.6 to 14.0 mT (average = 5.9 mT) (Table T6; Fig. F43). The MDF of the serpentine mud samples is extremely low and ranges from 2.5 to 3.7 mT (average = 2.8 mT). Subsequent to demagnetization, individual components of magnetization were identified for each sample using a least-squares fitting routine applied to each component of magnetization identified on a vector end point diagram.
Stable magnetic directions were determined for 17 of the 26 oriented hard rock samples and are displayed in Table T6. This component represents a chemical remanent magnetization acquired by magnetite particles as they formed during serpentinization (e.g., Stokking et al., 1990; Kelso et al., 1996). The stable component of magnetization decays to the origin of the vector end point diagrams in most cases. However, for a small number of samples, the stable component of magnetization did not decay to the origin. As expected from their emplacement history, serpentine muds did not display a stable magnetic component.
The AF demagnetization behavior of representative samples is illustrated in vector end point diagrams and shows three different types of behavior (Fig. F47). Type 1 samples have one single component of magnetization (Fig. F47A, F47B) that yields a stable magnetic direction, despite low MDF values. Type 2 samples display a ChRM with a soft overprint that is removed by a 10- to 15-mT field (Fig. F47C, F47D). It was not possible to determine a ChRM from Type 3 samples (Fig. F47F) or from serpentine muds (Fig. F47H). The results of the vector component analysis of the stable magnetizations are presented in Table T6.
Comparison between the demagnetization behavior of NRM and the saturation IRM (SIRM) was conducted to obtain initial information about the domain state of the magnetic carrier mineral (Fuller et al., 1988; Lowrie and Fuller, 1971; Stokking et al., 1990). The SIRM of serpentinized harzburgites is more stable during AF demagnetization than the NRM (Fig. F48), which suggests that multidomain magnetite dominates the magnetic properties. However, a certain number of pseudo-single domain-sized magnetite grains that are capable of carrying a stable magnetic remanence are apparently present, in addition to the multidomain magnetite.
The variability of the ChRM and the potential use of the ChRM to reorient individual rock pieces were investigated by stepwise demagnetization of one 72-cm-long piece of serpentinized harzburgite. The results of this test show a variation in declination between 150° and 330° and a variation in inclination between 38° and 84° at 20 mT over this short interval (Fig. F49). This demonstrates that the magnetization, which was acquired as a chemical remanent magnetization during serpentinization, was either acquired over a relatively long period of time or while the rock was being deformed and rotated. In addition, the degree of serpentinization within individual rock pieces can vary significantly (see "Igneous and Metamorphic Petrology of Ultramafics"), which might contribute to the nonhomogenous magnetization. This result precludes the use of the ChRM as reorientation tool.