Previous   |    Next

doi:10.2204/iodp.proc.318.102.2011

Paleomagnetism

Samples, instruments, and measurements

Paleomagnetic investigations during Expedition 318 relied on three main types of measurements: (1) those of the natural remanent magnetization (NRM) of archive-half sections before and after alternating-field (AF) demagnetization, (2) those of discrete samples with at least partial AF demagnetization curves and thermal demagnetization on selected specimens, and (3) those of the anisotropy of magnetic susceptibility (AMS) (including bulk magnetic susceptibility, χ). Anhysteretic remanent magnetization (ARM) and isothermal remanent magnetization (IRM) were also studied on selected samples.

Archive halves of cores were measured on the 2G Enterprises cryogenic magnetometer with an inline AF demagnetizer. Discrete samples were measured either on the cryogenic magnetometer or on a spinner magnetometer. AMS and bulk susceptibility were measured with the Kappabridge KLY4-S spinning magnetic susceptometer. ARMs were imparted with the DTech 2000 AF specimen demagnetizer and IRMs were imparted with an ASC impulse magnetizer. Please note that although it is the custom in the paleomagnetic literature to refer to specimens as the objects which are actually measured, in the case of IODP material, there is no difference between the “sample” in the sense of the IODP convention and the specimen in the paleomagnetic sense. We will use the term sample throughout the paleomagnetic reports, by which we mean both sample and specimen.

Cores were not oriented with respect to geographic North during Expedition 318. The sites are quite near the magnetic South Pole. The direction of the geomagnetic field is ~89° up or steeper, and any orientation relying on magnetic methods would be highly unreliable.

The sample and software coordinate systems for the 2G Enterprises Model-760R superconducting rock magnetometer (SRM) and the archive and working halves are shown in Figure F9. Both are “right handed” such that if the +x-axis is along the direction of the thumb, the +y-axis is along the index finger and the +z-axis is along the middle finger. The up arrow is the –z-axis by paleomagnetic convention. In order to maintain a consistent reference frame between the two halves of the core, the +x-axis is always toward the double lines predrawn onto the core. The core is split so that the double lines are on the working half, so the +x-axis is into the core surface for the working half (including discrete samples taken from the working half) and out of the core in the archive half. In the coordinate system defined in the magnetometer software, the +z-axis is into the machine and the +x-axis is up. The same coordinate system is used in the Molspin Minispin magnetometer.

Discrete samples were taken either in plastic sampling cubes or by sawing the samples into cubes with parallel and cut-off saws. We used the Japanese Natsuhara sampling cubes (7 cm³), pushing them into the working half with the “up” arrow pointing in the core up direction (Fig. F10A). The sample +x-axis is toward the double lines on the working half and the up arrow is the –z-axis. Sawed samples were scribed with an up arrow and wrapped in parafilm to prevent them from drying out.

The 2G magnetometer software is quite flexible, allowing measurement in any of the 24 possible positions shown in Figure F10B. Data are output to files rotated in sample coordinates. The Minispin allows measurement in four or six positions using either a long or short spin time. We used the four position/short spin protocol and simply measured samples multiple times if the circular standard deviation exceeded 5°. Multiple measurements were averaged.

At the high southerly latitudes of our site locations, the magnetic field is presently pointing steeply upward. The reverse direction is steeply downward and unfortunately parallel to the infamous “drill string remanence.” Generally speaking, after demagnetization a first-order interpretation would be that “up” directions (–z-axis or negative inclinations) are normal and “down” directions (+z-axis or positive inclinations) are reverse.

All remanence measurements on the archive halves were made using a 2G Enterprises Model-760R SRM equipped with direct-current superconducting quantum interference devices and an inline, automated AF demagnetizer capable of reaching a peak field of 80 mT. Measurements on the archive halves were taken at 5 cm intervals. We measured the NRM followed by measurements after treatment in alternating fields of 5, 10, 15, and 20 mT on selected cores. After examination of representative demagnetization sequences, a level of demagnetization was chosen that removed the drill string remanence yet preserved enough remanence to be measured on the shipboard magnetometer. Archive halves were demagnetized to either 15 or 20 mT based on these data. In a very few cases, archive halves were demagnetized to 25 mT to ensure that the drill string remanence had been removed to the extent possible.

Because of possible misidentifications of polarity resulting from drill string remanence, we obtained samples from each core, usually one per section. Discrete samples were taken in intervals with the least disturbed bedding (from coring or soft-sediment deformation). These were measured (as time allowed) either in the SRM using a specially designed discrete sample tray (Fig. F11) or in the Minispin magnetometer. Because measurement of discrete samples on the 2G magnetometer is not yet done routinely aboard the JOIDES Resolution, we attempted to establish a protocol that maximized efficiency while maintaining data reliability.

The discrete sample tray has spaces for 16 samples at 10 cm intervals. This is less than the width of the response function, so we had some concern about interference of sample moments. To investigate this possibility, we made measurements at 0.5 cm intervals on 16 standard samples placed in the top-toward orientation in Figure F12. Sample positions are shown as triangles on the right panel and the data acquired by the SRM software are plotted as declination, inclination, and intensity versus track position (centimeters below reference line). From these data, we determined the exact positions of the sample spaces in terms of machine offsets and also determined that for the weak samples encountered during Expedition 318, interference of magnetic moments would not be a problem. We measure 14 samples in a single run, leaving the outermost sample positions empty.

In order to optimize the measurement procedures, we measured seven standard samples using the discrete sample tray (Fig. F11) in the 24 possible positions in the SRM (Fig. F10). Measurements after subtraction of the background and sample tray measurements are plotted in Figure F13A. In general, the circular standard deviations of the replicate measurements were of the order of 5°. It is impractical to measure samples in 24 positions, so we selected three positions that measured each axis along both positive and negative directions: top-toward, top-to-right, and away-up. Means and circular standard deviations for these three selected measurements are shown in Figure F13B. The mean of the three measurements was within a few degrees of the means of the original 24 positions; we consider this sufficiently precise for the purposes of this study. Therefore, all cryogenic magnetometer measurements on discrete samples were carried out in these three positions and averaged.

Samples can be demagnetized using the inline AF demagnetizer on the SRM, a Dtech D2000 AF demagnetizer, or the Schonstedt thermal demagnetizer. To compare the Dtech and inline AF demagnetizers, test samples were given an anhysteretic remanence (see below) and demagnetized using the inline AF demagnetizer on the SRM. After remagnetization with a second ARM, they were demagnetized again using the Dtech D2000 AF demagnetizer. For both tests, samples were measured in three positions: top-toward, top-to-right, and away-up. In the inline AF demagnetizer, each specimen was also demagnetized along three axes (x, y, and z) prior to measurement. In the Dtech AF demagnetizer, the specimens were demagnetized along three axes (x, y, and z) and measured in the first position then demagnetized along the y-axis and measured in the second position. Finally, the specimens were demagnetized along the z-axis and measured in the third position. We show the data for the two experiments in Figure F14. Considerable scatter in the measurements can be attributed to both instrument noise (Fig. F13) and acquisition of spurious magnetizations in the AF demagnetizers. The fields in the instruments are axial and on the order of tens of nanoTesla, depending on the orientation of the ship, and cannot be demagnetized further. However, no consistent pattern is apparent in the directions acquired, therefore it is not a simple ARM or gyroremanent magnetization (see Tauxe et al., 2010, for details) that would be acquired parallel and perpendicular to the axial field, respectively. In any case, after averaging the three measurements at each step and calculating a best-fit line through the measurements, the directions were within 5° of the known direction of the ARM (along the x-axis). Given the ability to measure 14–16 samples in a single demagnetization/measurement sequence, we chose to use the inline SRM AF demagnetizer for analysis of most of the shipboard samples. Each measurement and demagnetization was carried out along the three directions (top-toward, top-to-right, and away-up).

Data from the archive halves and discrete samples were saved in .srm and .dsc file formats and uploaded to LIMS. The data files were also converted to the Magnetics Information Consortium (MagIC) standard format with a set of programs that have been added to the PmagPy software package, available for download at magician.ucsd.edu/​Software/​PmagPy. In particular, we used ODP_srm_magic.py and ODP_dsc_magic.py for the split halves and discrete-sample measurements, respectively. Data downloaded from LIMS in .csv format can be converted to MagIC format using the ODP_csv_magic.py program. The MagIC data format allows analysis of demagnetization diagrams, equal area projections, and various plots versus depth using the standard programs in the PmagPy software package.

We also used the Minispin magnetometer along with the Dtech D2000 AF demagnetizer for certain samples because these instruments can be run in parallel with the cryogenic system. Cross-calibration of intensities is a work in progress, but the directions are completely compatible. Data from the Minispin magnetometer are not currently uploaded into LIMS but can be converted to the MagIC standard format using ODP_spn_magic.py in the PmagPy package. These data are archived by IODP and are available upon request from the IODP Data Librarian.

Our investigations on discrete test samples indicate that demagnetization above ~60 mT with both the inline and Dtech D2000 AF demagnetizers became highly scattered and no longer tracked to the origin, despite averaging (e.g., Fig. F14), so discrete samples were not demagnetized above this level.

AMS measurements were made on the Kappabridge KLY4S instrument using the SUFAR program supplied by AGICO, and bulk susceptibility was measured with the SUFAM program. Later in the expedition the AMSSpin program (available for download from magician.ucsd.edu/​Software) was adapted to the shipboard Kappabridge and used for analysis. This is a much more user-friendly application, and although the data analysis is slightly different than in the AGICO software, the data output are statistically indistinguishable. The Kappabridge measures anisotropy of magnetic susceptibility by rotating the sample along three axes, stacking the data, and calculating the best-fit second-order tensor. It also measures bulk susceptibility (χ). Tensor elements were converted to eigenparameters (eigenvectors V₁, V₂, and V₃ with associated eigenvalues τ₁, τ₂, and τ₃, in which τ₁ is the maximum and τ₃ is the minimum; terminology of Tauxe et al., 2010). These can be interpreted in terms of particle alignment within the sample (Fig. F15). Normal sedimentary fabrics are oblate with vertical axes of minimum susceptibility (e.g., Fig. F15B). Disturbance by slumping or other deformation generally yields triaxial fabrics (e.g., Fig. F15D). We used the bootstrap method of Tauxe et al. (2010) to estimate confidence bounds on AMS eigenparameters.

ARMs can be imparted in the Dtech D2000 AF demagnetizer by applying an alternating field in the presence of a direct-current bias field (Fig. F16). In general, ARMs were imparted in an alternating field of 100 mT with a direct-current bias field of 50 μT. Plots of ARM versus susceptibility can be interpreted in terms of changes in grain size and concentration of the magnetic carriers (Fig. F17; see Tauxe et al., 2010, for details).

Isothermal remanences were imparted using the ASC impulse magnetizer. These data can be used to characterize magnetic mineralogy and coercivity distributions. Also, when plotted against ARMs or susceptibilities, they can provide information on changes in magnetic grain size and concentration.

Interpretation of magnetostratigraphy

As noted previously, the “drilling remanence” is generally an overprint in a downward directed vertical field. Ordinarily, it is readily distinguished from the original remanent magnetization, but at high latitude it is subparallel or antiparallel to the primary magnetization, making interpretation of polarity zones hazardous. Therefore, we paid careful attention to possible correlations of “polarity zones” with changes in whole-core susceptibility, character of demagnetization diagrams, and possibly also ARM versus susceptibility plots prior to making an interpretation of polarity.

Disturbed sediments, either caused by coring or geological processes (slumping, faulting, etc.) also disturbs the magnetic directions and fabric. High-resolution core photographs were examined for each core section and disturbed intervals were edited out of our magnetic records. Also, the shipboard sedimentologists were frequently consulted during the data editing process.

Once a polarity stratigraphy was established for a given hole, we correlated the pattern to the GPTS (Fig. F18). This was done with close collaboration with the shipboard biostratigraphy team. We use the chron terminology of Cande and Kent (1995) and the numerical ages of Gradstein et al. (2004) listed in Table T5.

Top of page   |    Previous   |    Next