IODP Proceedings    Volume contents     Search



The main objective of the shore-based paleomagnetic study was to produce a polarity magnetostratigraphy in as much detail as possible. At selected depth intervals in all three holes, attempts were also made to characterize the remanence carrier and to make preliminary estimates of relative paleointensity. These objectives were achieved through measurements of natural remanent magnetization (NRM) and alternating-field (AF) demagnetization of discrete samples (one from each section) taken from the working half. In addition, specific intervals selected on the basis of their high clay content and relatively high susceptibility were sampled at denser intervals ranging from one sample every 50 cm to one sample every 10 cm. No U-channels were measured at the OSP, as explained in "Samples."

Taking discrete samples in this case meant lower overall sample resolution than can be achieved with access to a long-core (pass-through) magnetometer. However, the discrete sample approach has the advantage of better measurement precision and allows more extensive demagnetizations of each sample compared to the 15 mT standard shipboard half-core measurements. Using this method, we were able to separate different components of the magnetic signal and, when possible, isolate the characteristic remanent magnetization (ChRM) using principal component analysis (PCA) (Kirschvink, 1980), yielding short successions of high resolution.

The conventional right-handed ODP coordinate system was used throughout (+z = downcore; +x = perpendicular downward from the split surface of the working half). The cores were recovered without the use of a Tensor tool and therefore were not oriented with respect to horizontal. However, because of the relatively high latitude (~40°N) of the sites, the lack of horizontal orientation did not present a problem for the purpose of constructing a polarity timescale.


Lithologic conditions that frequently alternated between soupy and hard cemented layers prevented the systematic collection of U-channel samples, although a cryogenic super-conducting quantum interference device (SQUID) rock magnetometer (755–4000, horizontal pass-through type) is available at the University of Bremen. Three U-channels were taken in Hole M0027A for postexpedition research (Sections 313-M0027A-222R-1, 68–147 cm; 223R-1, 0–150 cm; and 223R-2, 92–150 cm). These U-channels will be used to investigate in detail the paleomagnetic behavior/transition across the Eocene–Oligocene glacial maximum. At the OSP, oriented cube samples (6.2 cm3) were carefully collected from working-half cores. For clear characteristic magnetization, NRM was measured after detailed stepwise demagnetization on each axis of the sample up to 100 mT soon after a discrete cube sample was collected. All Expedition 313 paleomagnetic results are provided in SI units.


All paleomagnetic measurements were carried out in the paleomagnetic laboratory at the University of Bremen. Discrete sample NRM measurements were made using a 2G Enterprises pass-through direct-current (DC) SQUID cryogenic magnetometer (model 755–4000; horizontal orientation) equipped with an automated AF demagnetizer and an automated sample handler capable of processing 96 discrete samples (8 each × 12 sets) in one batch (Fig. F14). All measurements were made in a ferromagnetically shielded space to prevent samples from being exposed to the ambient laboratory magnetic field.

Measurements and procedures

Measurements of remanent magnetization were carried out on all discrete samples. In addition, all samples were also AF demagnetized at 5 mT increments up to 60 mT to isolate the ChRM. A few samples were then further AF demagnetized at 10 mT up to a maximum field of 100 mT.

During the first batch (96 samples) of AF demagnetization and measurements of core from Hole M0027A, we discovered that many samples were acquiring a magnetization, typically at AF fields higher than 40 mT. We interpreted this to be a gyroremanent magnetization (GRM), which is AF demagnetization–induced magnetic remanence acquired perpendicular to the applied alternating field in strongly anisotropic magnetic material (Stephenson, 1980). In light of this finding, we changed the demagnetization procedure for the other samples following the technique proposed by Dankers and Zijderveld (1981), where the remanence in each axis is measured directly after AF demagnetization of the same axis. The theory is that any remanence acquired in one axis during AF demagnetizations of another axis will be removed when the axis is AF demagnetized by the same field strength.

ChRM was determined using PCA on a sequence of demagnetization steps determined individually for each lithology. For the first batch of samples, which were demagnetized without the technique proposed by Dankers and Zijderveld (1981), the interference of the GRM forced us to use a different approach. Because the z-axis was always demagnetized last, the intensity information in this axis was not affected by the acquisition of a GRM. In these cases, we settled on establishing the direction of the samples using demagnetization steps at fields below 40 mT.

Data quality

The quality of the paleomagnetic data varied markedly for different lithologies. The best results were obtained in clay-rich intervals, whereas in most sandy parts of the cores, data were too noisy to determine a prevailing remanence direction. A few samples were not fully demagnetized even with alternating field strengths of 100 mT, suggesting the presence of a high-coercivity component.


Polarity histories were reconstructed for a few of the clay-rich stratigraphic intervals based on the ChRM inclination data. Polarity zones were tentatively correlated to the GPTS of Cande and Kent (1995) using age control provided primarily by Sr isotope stratigraphy and limited biostratigraphic constraints. The resulting age assignments for the reversals must be considered uncertain because of fairly large errors in age control (~1 m.y.) and fairly rapid polarity changes in the GPTS for the stratigraphic intervals with polarity history.

Data from the sandy sequences often proved to be too noisy for magnetostratigraphic interpretations. The poor data quality is attributed to unstable or weak NRM in the sediments.