IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.324.106.2010

Paleomagnetism

Recovered core materials in Hole U1349A show diverse magnetic properties. We analyzed 32 discrete samples from volcanic rocks. The topmost recovered cores as well as the deepest cores (324-U1349A-5R and 16R) are mostly composed of volcanic breccias (see "Igneous petrology"). We are not certain if the primary magnetization directions of the material in these cores survived during autobrecciation and lava deposition. This material likely carries a depositional remanent magnetization (DRM). The basalts recovered from Cores 324-U1349-7R through 15R show evidence of a possible chemical remanent magnetization (CRM). Consequently, any directional result must be interpreted with care.

We used the 2G Enterprises cryomagnetometer for some of the volcaniclastic sample measurements because the natural remanent magnetization (NRM) of these samples was too weak to be measured on the Molspin Minispin magnetometer (a few tens of milliamps per meter). Alternating-field (AF) demagnetization and thermal demagnetization were carried out using the DTech degausser and Schonstedt oven, respectively. The NRM bulk magnetic susceptibilities of most of our samples are on the order of 10–3 SI, which is an order of magnitude smaller than the values of basalts recovered from Holes U1346A and U1347A.

Working-half discrete sample measurements

Alternating-field demagnetizations

Three types of behaviors that roughly correlate to the igneous lithology were observed in AF demagnetization measurements in Hole U1349A. The first type, observed in the three samples from Core 324-U1349A-5R, is characterized by a very low NRM. Two out of the three samples show several components that are difficult to separate on the orthogonal vector diagram (Fig. F64A). Only one sample gives a demagnetization result with a maximum angular deviation <8°.

The second type of behavior, present in Cores 324-U1349-7R through 14R, is very different from that observed in recovered fine-grained igneous rocks in Holes U1346A and U1347A (Fig. F64B). The magnetization is very stable, with median destructive fields (MDFs) on the order of 20–25 mT and an initial plateau on the demagnetization spectra that is characteristic of single-domain grains. A large part of the remanent magnetization (up to 25%) is not demagnetized at 150 mT, the maximum AF field for the DTech degausser. Both the drilling-induced and viscous remanent magnetization overprints are very small in samples from this section. All AF demagnetizations are univectorial between ~10 and 150 mT, and the inclinations given by principal component analysis (PCA) (Kirschvink, 1980) are all negative and shallow with an average of –5.0° ± 5.0° (1σ).

The third type of behavior is observed in samples from Core 324-U1349-16R. Two samples are volcaniclastics and one is a from a lava pod (see "Igneous petrology"). These samples have a very low magnetization, lower MDFs than the samples from the section above, and several components on the Zijderveld diagram. However, it is possible to define a linear component pointing toward the origin after demagnetization to ~20 mT.

Interpretations of resulting inclinations from these three types are difficult for various reasons. For the samples of the first type (volcaniclastics), the magnetization is not a thermoremanent magnetization (TRM) but a DRM. It is difficult to estimate when and in which conditions it was acquired. The second-type samples (amygdaloidal basalt) are ideal field recorders from a rock magnetism point of view because their remanence is extremely stable and they give very consistent inclination values (Table T10). Thin section observations indicate that both hematite and titanomagnetite that have suffered oxy-exsolution are present in these cores (see "Igneous petrology"). However, it is unclear which of these two minerals carries the magnetization component that defines the characteristic remanent magnetization. If it is the titanomagnetite, then the magnetization is a true TRM. Otherwise, it is a CRM acquired when the hematite was formed (see "Alteration and metamorphic petrology" and "Thermal demagnetizations"). In the third group, the only sample that carries a true TRM is a fine-grained basalt sample with an inclination of –13.8°.

Thermal demagnetizations

No samples taken from volcaniclastics were used for thermal demagnetizations because this rock type is too fragile for multiple heating steps. Three characteristic behaviors are apparent in thermal demagnetization results from basalts that roughly correlate to the igneous lithology. The first type of thermal demagnetization behavior (samples from Sections 324-U1349A-7R-3 through 13R-5) is characterized by high, small-range unblocking temperatures around 500°C (Fig. F64C). This temperature indicates that the magnetization is probably carried by almost pure magnetite. Demagnetization results typically show a single component once the drilling overprint is removed after heating to ~400°C. Some samples show one point that deviates from the main direction around 475°–525°C, which might be a systematic error in that particular temperature step or might represent the statistical variability in the directions. Sample 324-U1349A-12R-4, 40–42 cm, has a different behavior; its magnetization decreases almost linearly from 100°C and then drops at 500°C, probably indicating the presence of titanomagnetite (-maghemite) with a large range of Ti content (Hunt et al., 1995). This sample did not allow us to define a stable component by PCA with a maximum angular deviation <8°. All other samples gave consistent negative shallow inclinations (Table T10). Possessing the second type of thermal demagnetization characteristics, samples from Sections 324-U1349A-14R-2 through 15R-1 (Fig. F64D) exhibit magnetization that decreases almost linearly from 100°C to ~450°–500°C and then drops at ~540°C. Similar to Sample 324-U1349A-12R-4, 40–42 cm, this seems to indicate a range of Ti content as well as the presence of an almost pure magnetite phase. After removal of the overprint by 150°C, the magnetization follows a univectorial decay toward the origin. PCA gives negative inclinations quite different from each other. One sample had severely erratic behavior and had to be rejected (Sample 324-U1349A-15R-3, 30–32 cm). Possessing the third type of thermal demagnetization characteristics, samples from Sections 324-U1349A-15R-3 through 15R-5 had inconsistent and erratic behavior and were not taken into account in the rest of the analysis.

During thermal demagnetization, no particular pattern was found in the bulk susceptibility variation after each heating step. In some cases, the susceptibility stayed constant until 640°C, in other cases it doubled when heated over 500°C, but in most cases it decreased more or less linearly between 300° and 600°C to half its room temperature value. This variability of behaviors indicates complex magnetic mineralogy with a mixture of different phases. Onshore rock magnetism experiments will be crucial to understand this complex behavior for the interpretation of these results.

Implications for lava deposition history

The average inclination in Hole U1349A was calculated from both AF and thermal demagnetization results with maximum angular deviation values <8° from Cores 324-U1349A-7R through 14R (massive basalt flows). The average inclination is –4.3° ± 5.9° (1σ) (Fig. F65), suggesting the lavas in Hole U1349A were formed near the magnetic equator, assuming that the magnetization carried by these rocks is a true TRM and not a CRM.

The inclination inconsistency in the volcaniclastic section is probably a result of directions of primary magnetization that were rotated during deposition of these volcanic sediments.