Previous   |    Next

doi:10.2204/iodp.proc.318.104.2011

Paleomagnetism

Methods and instruments

Archive-half cores were measured on the 2G Enterprises cryogenic magnetometer and demagnetized at 20 mT. Generally, one section per core was also demagnetized at 5, 10, and 15 mT prior to the 20 mT step. The anisotropy of magnetic susceptibility (AMS) and bulk susceptibility were measured on all discrete samples, typically one per core section. Selected samples were also analyzed using step-wise alternating-field demagnetization on either the cryogenic or spinner magnetometers.

Results

All paleomagnetic measurements from archive halves are shown in Figure F16. The quality of these data can be assessed by:

1. Examining the sedimentary fabrics using the AMS data,

2. Examining the high-resolution core photographs and sedimentary logs for evidence of drilling or sedimentary disturbance of the core material,

3. Examining the behavior of the sediments during progressive demagnetization, and

4. Comparing the inclinations observed with those expected from paleosecular variation at this latitude.

We plot all susceptibility data from the discrete samples in Figure F17. Sedimentary fabrics are expected to be oblate with indistinguishable intermediate and maximum eigenvalues (triangles and squares, respectively, in Fig. F17A) and vertical axes of minimum susceptibility (inclinations of the V₃ eigenvector shown in Fig. F17B). The AMS data from the top few cores have virtually indistinguishable eigenvalues (they are isotropic) and the V₃ eigenvectors are strongly deviated from the vertical. These samples suffer from drilling disturbance, and no reliable paleomagnetic data can be derived from them. However, the AMS data of Core 318-U1356A-14R (130 mbsf) and below are similar to those expected for a primary sedimentary fabric. Note also that bulk susceptibility measurements (Fig. F17C) are calibrated volume susceptibilities and could be used to calibrate the whole-core magnetic susceptibility data.

We inspected each core section using digitally enhanced photographs (see example in Fig. F18). The gray portion of the paleomagnetic data in this example was deleted because of the evident microfaulting (e.g., Fig. F18D) and contorted bedding (e.g., Fig. F18E). The lower portion of the core suffered only minor core breaks (e.g., Fig. F18F) and therefore was considered reasonably undisturbed. Fortunately, at high latitude rotations round the vertical axis only result in offsets in declination (e.g., at ~50 cm in Fig. F18B) but do not affect our ability to discriminate polarity, which is based solely on the inclination. Other types of core disturbance (e.g., from the drilling process or from slumping) were also edited out. The remaining data, after demagnetization at 20 mT, are shown in Figure F16 (black circles).

Progressive demagnetization data can be used to assess whether the samples are stably magnetized or have unremoved drilling-induced overprints. We carried out progressive demagnetization on all archive halves as well as selected discrete samples (see examples in Fig. F19). The behavior of the discrete samples and archive-half sections was similar (compare Fig. F19A, F19C from discrete samples with Fig. F19B, F19D from archive halves). The normally magnetized sediments (Fig. F19A, F19B) exhibit the behavior characteristic of removing a steeply downward directed overprint, likely the drill string overprint. In the reversely magnetized sediments, the steep downward direction is also removed, but the effect is more subtle. In most cases, the drill string overprint was removed by 15 or 20 mT, and we feel reasonably confident that the directions derived by demagnetizing the archives to 20 mT in most cases represent true polarity.

Finally, we compare the observed inclinations with those expected at the site latitude in the Eocene (~60°S latitude) from the statistical model of paleosecular variation of Tauxe and Kent (2004). We plot the inclinations of the natural remanent magnetizations from archive measurements (after editing out disturbed portions of the core) in Figure F20A. After demagnetization at 20 mT, the two polarity groups of normal and reverse inclinations become much more distinct (Fig. F20B). The expected distribution of inclinations from the statistical model of paleosecular variation is shown in Figure F20C. The data from the archive section half measurements are biased toward shallow inclinations compared to the field model, an effect commonly observed in sediments, particularly ones as compacted as those considered here. The clear separation of inclinations into two modes, however, means that our interpretations of polarity are reasonably robust, except for cases of shallow inclination.

Correlation to the geomagnetic polarity timescale

Three intervals with sufficiently continuous recovery make a correlation to the GPTS possible (Fig. F21; Table T1) with constraints provided by biostratigraphic identifications (see “Biostratigraphy”).

The top portion (Fig. F21A), spanning Cores 318-U1356A-14R through 51R, correlates to polarity Chrons C5AAn to C5Dn. The polarity stratigraphy from Cores 46R through 51R does not have a straight forward fit to the GPTS, so a hiatus could be placed between Cores 46R and 47R. The reversal within Core 47R could correspond with Chron C6Cn.2n(y), according to nannofossil constraints (see “Biostratigraphy”).

Cores 318-U1356A-68R through 92R correlate to polarity Chrons C7An to C12r. Cores 78R and 79R have scattered directions. Examination of the progressive demagnetization curves from this interval suggests that these cores could have incompletely removed drill string overprints on normal polarity intervals. If true, this interval may correspond to Chron C10n. Shore-based demagnetization studies are necessary to confirm this suspicion. Core 92R has some apparently normal directions in it and the correlation of directions is unclear. Many cryptochrons are identified in Chron C12r (e.g., Cande and Kent, 1992) and it is possible that these directions correspond to one of the cryptochrons.

The failure to recover undisturbed sediments from Cores 318-U1356A-94R through 97R preclude interpretation of the magnetostratigraphy. Recovery improved below these cores, and the portion spanning Cores 98R through 106R is shown in Figure F21C. Starting from the bottom, the lowermost normal polarity interval in Core 105R and the top part of 106R corresponds to Chron C24n.3n. Core 104R is dominantly reverse with a short normal interval correlating to Chron C24n.2n. The top of Core 104R and all of Core 103R record Chron C24n.1n. Core 102R has a reverse magnetization and fits comfortably within Chron C23r, and the bottom of Core 101R could record the base of Chron C23n.2n. However, the correlation of Cores 100R through 98R is not straight forward. Cores 100R and 99R have mostly reverse magnetizations, whereas Chron C23n.2n is normal. The transition recorded at the base of Core 101R correlates to Chron 23n.2n, consistent with a constant sedimentation rate model and available biostratigraphic constraints. Extrapolating upward from the base of Chron C24n, there must be a hiatus between the bottom of Core 101R and 100R. The smallest gap, consistent with biostratigraphic constraints (see “Biostratigraphy”), would place the transition recorded in Core 99R at the base of Chron 23.1n. Core 100R would then be Chron C23n.1r in age. The duration of the hiatus under this interpretation would be at least 0.9 m.y. The top of Chron C23n is recorded in Core 98R.

Top of page   |    Previous   |    Next