PALEOMAGNETISM

Paleomagnetic and rock magnetic measurements at Site 1253 were aimed at estimating magnetostratigraphy of sediments, age, and internal structure of the igneous units. After measuring the NRM, all sections of the archive half of the core were demagnetized using alternating-field (AF) magnetization up to 40 mT in increments of 5 mT at 5-cm intervals. Additionally, the maximum demagnetization level was changed to 50 mT in two intervals: 400.50-480.71 (Sections 205-1253A-5R-1 through 17R-4) and 542.79-543.49 mbsf (Section 205-1253A-31R-4) in order to remove magnetic overprints.

One to three oriented discrete samples were routinely collected from each section of the working half of the core primarily for analysis for shore-based anisotropy of magnetic susceptibility and rock magnetic study. All discrete samples were used to constrain interpretations of the magnetic record of the archive long core by providing additional measurements of polarity and basic magnetic characterization. The discrete samples were demagnetized up to 80 mT in 5-mT increments to permit principal component analysis. The rock magnetic investigations on the ship consist of progressive acquisitions of isothermal remanent magnetization (IRM) and stepwise thermal demagnetization on discrete samples. Magnetic stability analyses following the Lowrie-Fuller test (Lowrie and Fuller, 1971) were also conducted to investigate the magnetic domain state of the discrete samples.

Sediments

The majority of NRM inclinations for archive sections show steep angles of ~30°-60°N. Continuously high NRM intensities are also observed, whereas rapid changes in magnetic susceptibility are measured by MST core logger (see "Physical Properties"). In most of the sections, these high magnetic intensities were easily demagnetized in the earliest AF demagnetization steps at ~10 mT. This suggests that the steep NRM inclinations and constantly high magnetic intensities have been caused by a strong drilling-induced magnetic overprint. The drilling-induced magnetization generally shows steep angles of magnetic inclination and relatively high magnetic intensity that are thought to be a weak IRM (Musgrave et al., 1993). Changes in magnetic inclination and intensity observed after the AF demagnetization suggest the magnetic overprints were successfully removed in weak AF demagnetization levels. Changes in magnetic inclinations observed from pelagic sediments provide information about magnetic polarity changes and are used for the interpretation of sedimentary age after correlation with the standard magnetic polarity timescale of Berggren et al. (1995a, 1995b) and with previous paleomagnetic results at Site 1039 (Kimura, Silver, Blum, et al., 1997).

Inclination changes of archive halves in the sedimentary sections after AF demagnetization of 40 mT are relatively unclear, although most inclinations are indicating normal polarity (Fig. F64). Comparison between the observed inclinations from the archive sections and discrete samples indicate that postdepositional remanent magnetization (pDRM) of mostly hemipelagic sediments above the gabbro represent normal polarity. However, a few intervals and discrete samples also indicate at least two reversed polarities are possibly identified at ~375-376 (Core 205-1253A-2R) and ~388-390 mbsf (Sections 205-1253A-3R-2 through 3R-3). The paleomagnetic study at Hole 1039B interpreted two positive inclinations below 370 mbsf as magnetic Subchron C5Bn.2n (15.034-15.155 Ma) and Subchron C5Cn.1n (16.014-16.239 Ma), respectively. Based on these previous identifications, a positive polarity at ~370-374 mbsf (Sections 205-1253A-1R-1 through 2R-1) may be interpreted as Subchron C5Bn.2n. Then, a polarity boundary at ~388 mbsf (Section 205-1253A-3R-3) might be identified as the appearance of Subchron C5Cn (16.726 Ma).

Magnetic inclination of sediments from 430 to ~450 mbsf show negative polarity. Only two discrete samples indicate positive polarity (Fig. F64); however, identification of magnetic chrons (or subchrons) was unsuccessful because of poor resolution of the paleomagnetic data.

Gabbro

Demagnetization of the gabbro intrusions, conducted only within intact portions of core, revealed that these intrusions have high variations of magnetic inclinations with high magnetic intensity (Figs. F64, F65). After AF demagnetization to 40 mT, some changes in magnetic inclinations were observed; however, the high NRM intensity in most of the gabbro is easily demagnetized at weak AF demagnetization levels (Fig. F65). The ratio of magnetic intensity after the AF demagnetization (20 mT) to NRM intensity also suggests unstable magnetization of the gabbro. The gabbros' magnetic coercivity is very weak, whereas they show high NRM intensity. Three peaks in the intensity ratio, indicating relatively high coercivity magnetization, corresponds with high magnetic intensities after AF demagnetization. Probably, this unstable remanence of magnetization strongly depends on the grain size of magnetic minerals after AF demagnetization. Rock magnetic tests carried on discrete samples revealed most samples contain multidomain (MD) magnetite in which the grain sizes are large enough (>100 µm) to cause unstable magnetic behavior during the demagnetization (see below). Fortunately, some intervals of the intrusions have relatively stable magnetization behavior during demagnetization that coincides with four zones of high magnetic intensity (Fig. F65). The top of the first high intensity interval is observed at ~400 mbsf (Section 205-1253A-4R-CC) in the uppermost gabbro intrusion; the intensity decreases with depth in the intrusion. The lower igneous unit has three zones of high magnetic intensity at ~462 mbsf (Core 205-1253A-14R) to 478.4 mbsf (Section 17R-2, 130 cm), 513.20 mbsf (Section 25R-1, 20 cm) to ~523 mbsf (Section 27R-3), and ~572 mbsf (Section 38R-1) to ~593 mbsf (Section 4R-3).

Although most of the igneous section exhibits unstable magnetizations, magnetic inclinations between the archive sections and discrete samples show similar trends that indicate the gabbros may retain initial magnetizations that represent different age polarities (Fig. F64). Identified normal chrons (or subchrons) are shown in Table T7. Because discrete samples show negative inclination, positive inclinations of the archive-half core section at 491.38-492.80 mbsf (Section 205-1253A-20R-2) are not identified as normal polarity. The top of the high-intensity zone at 513.65 mbsf (Section 205-1253A-25R-1) nearly coincides with the start of a long normal polarity at 513.20 mbsf (Section 25R-1, 20 cm). Both polarity and magnetic intensity changes may indicate age boundaries within the igneous units. The polarity boundaries and the changes in the magnetic intensity ratio were compared with lithostratigraphic observations and identification of mineralogy in these units (see "Petrology").

Rock magnetic tests were conducted to investigate stability of magnetization and to identify magnetic minerals in the sediments and magmatic rocks. All samples for the tests were collected from minicore and cube samples taken from working-half core sections.

Saturation Isothermal Remanent Magnetization and Lowrie's Tests of Hemipelagic Sediments

For investigation of saturation isothermal remanent magnetization (SIRM), seven small samples (1 cm³) were collected from cube samples and their IRM was measured after acquisition of an isothermal magnetization applied from 20 to 1200 mT. After 24 steps of IRM measurement, thermal demagnetization of multicomponent IRM (Lowrie, 1990) was used as the primary means of identifying magnetic minerals. For these experiments, orthogonal isothermal magnetizations of 100, 300, and 1200 mT were applied on +x-, +y-, and +z-axes of the sample, respectively. The samples were then demagnetized using 16 thermal steps from 50° to 650°C (Fig. F66). The IRM intensity curves for most of the samples smoothly increase up to 300 mT then saturate at ~600 mT. Thermal demagnetization curves indicate there are at least three unblocking temperatures (T_ub). The magnetic intensities suddenly decrease at 170°C and decay rapidly up to ~320°C. These two T_ub indicate goethite (T_ub 150°C) and pyrrhotite or greigite (T_ub 320°C) and are present in the hemipelagic sediments. The rapid decrease of magnetic intensity on the +x-axis up to 320°C suggests that pyrrhotite (or greigite) dominates rather than goethite. Therefore, gentle intensity curves from 300 to 600 mT on the SIRM are believed to reflect the presence of pyrrhotite. A third T_ub is identified at ~550°C, and all intensity components smoothly decay from 320°C to 550°C and indicate that the pDRM carrier might be magnetite (T_ub = 575°C), and its magnetization is very stable.

Lowrie-Fuller Tests for Hemipelagic Sediments and
Igneous Rocks

The Lowrie-Fuller test (Lowrie and Fuller, 1971) was performed on seven hemipelagic sediments and seven gabbro samples. To interpret these data in terms of magnetic domain state, the relative shapes of the anhysteretic remanent magnetization (ARM) and the SIRM demagnetization curves were combined on a plot (Fig. F67A). The samples were progressively demagnetized up to 80 mT in increments of 5 mT after acquisition of the ARM, represented as a weak-field thermal remanent magnetization (TRM), and produced by demagnetizing the samples in a 150-mT AF in the presence of a 0.05-mT direct-current field. The samples then acquired a laboratory SIRM in a 1-T magnetic field that represents TRM. After the acquisition, the samples were progressively demagnetized up to 80 mT in 5-mT increments. It should be emphasized that any interpretation relies on the assumption that the samples contain a simple magnetite or titanomagnetite mineralogy. Given this assumption, the ARM demagnetization curve decreases rapidly relative to SIRM for a few samples of the gabbro samples, suggesting the presence of MD grains. For most of the samples, the two curves are similar and clear constraints on magnetic domain state are not possible at this time.

In the hemipelagic sediments, the resistance of the ARM to early steps of AF demagnetization relative to SIRM suggests the presence of single- to pseudosingle-domain grains (Fig. F67B).

For detailed analyses of the paleomagnetic and rock magnetic investigations, all discrete samples were demagnetized up to 80 mT in increments of 5 mT. In the hemipelagic sediments, demagnetization and magnetic component curves for both the inclination and the declination for most samples smoothly decrease toward the origin of the Zijderveld diagram (e.g., Sample 205-1253A-2R-2, 12-14 cm) (Fig. F68). However, a few samples show odd behavior during AF demagnetization (e.g., Sample 205-1253A-2R-6, 91-93 cm). The demagnetization curves are similar to the curve of magnetic iron sulfides, such as greigite, in which magnetization trends generally toward the opposite direction to the AF and causes an increase in the magnetic intensity during high AF demagnetization. However, decay of the magnetic intensity up to 80 mT emphasizes there are not serious effects of magnetic noise from the iron sulfides on the demagnetization of the samples. These curves may also reflect artificial disturbances of sediments during drilling so that these results should be excluded from the paleomagnetic study. Pervasive and severe drilling-induced disturbance on the sediments was observed as scattered plots of demagnetization steps on the Zijderveld diagram (e.g., Sample 205-1253A-2R-1, 15-17 cm) (Fig. F68).

In igneous samples, demagnetization curves of samples taken from working-half sections show relatively high magnetic intensity. These intensities resist weak AF demagnetization levels (Fig. F65) but smoothly decrease with increasing demagnetization level (e.g., Samples 205-1253A-5R-3, 133-144 cm, and 12R-1, 32-33 cm) (Fig. F68). Magnetic overprints in these samples are successfully removed at 20 mT. Additionally, the rock magnetic tests revealed their magnetic domain state to be MD (Fig. F67). Other discrete samples show a rapid decrease in magnetization in early AF demagnetization steps (e.g., Sample 205-1253A-22R-4, 63-64 cm) (Fig. F68). Although the magnetic behavior should be tested using other rock magnetic methods, this rapid decrease in magnetization of the igneous units (Fig. F65) indicates that much coarser magnetic grains are dominant in rocks than intrusions showing stable and high-coercivity magnetization.

The magnetic inclinations observed from the discrete samples are also used for estimations of paleolatitudes for the hemipelagic sediments and the igneous units. For this analysis, only inclinations of samples that had maximum angular deviation values <4° were used. Additionally, the calculations used only positive inclinations interpreted as normal magnetized samples in the Northern Hemisphere. The paleolatitude for each unit was calculated from the geocentric dipole relationship:

tan (I) = 2 tan (latitude),

where I is the average of the inclination groups. An angle deviation of the paleolatitude was calculated by Fisher's deviation method (Fisher, 1953). The averaged inclinations of each lithologic unit are shown in Figure F69. Although these results show high standard deviations, the estimated paleolatitude of hemipelagic sediment above the gabbro sill is 8.6°N (upper confidence limit = +11.6° and lower confidence limit = -5.6°) and that of the upper gabbro and lower igneous unit are estimated to be 13.4°N (+24.6° and -1.0°) and 11.0°N (+14.3° and -7.8°), respectively. These values are close to the current latitude at this site, ~9°N.