IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.322.104.2010

Paleomagnetism

We completed natural remanent magnetization (NRM) measurements and alternating-field (AF) demagnetization on a total of 269 discrete samples from Hole C0012A. We analyzed the results in Zijderveld diagrams (Zijderveld, 1967) and calculated the characteristic remanent magnetization (ChRM) direction using principal component analysis (Kirschvink, 1980). Volume magnetic susceptibility of these discrete samples was measured after AF demagnetization with the Kappabridge KLY 35, except for 96 discrete samples from Cores 322-C0012A-4R through 20R, which were measured for anisotropy of magnetic susceptibility (AMS) with the Kappabridge KLY 35 before the AF demagnetization experiments. The Königsberger (Q) ratio was also determined for all measured samples. Because of a malfunction of the spinner magnetometer toward the end of the expedition, only two samples from basaltic basement yielded measurements of remanent magnetization on board the ship. Additional postcruise measurements with a cryogenic magnetometer at Geological Survey of Japan, National Institute of Advanced Industrial Science and Technology, made it possible to characterize the remanent magnetization of basaltic samples collected for shipboard study.

NRM, magnetic susceptibility, and Königsberger ratio

Downhole variations of paleomagnetic data obtained from Hole C0012A are shown in Figure F35. Variations of magnetic properties among various lithologies are similar to those observed at Site C0011. Overall, silty claystones in the lower part of Unit III and siltstone in Unit IV and the upper part of Unit V have relatively low NRM intensity (averaging ~2 mA/m) compared to those of sandstones in Unit II (average ~9 mA/m) and volcaniclastic sandstones in the lower part of Unit V (~80 mA/m). In lithologic Unit II, NRM intensity also shows two extremely high peaks of the order of 1 A/m at 161 and 172 m CSF, corresponding to the presence of volcaniclastic sandstones. In lithologic Unit III, both NRM intensity and magnetic susceptibility show a stepwise decrease downhole. The mean NRM intensity value decreases from ~30 mA/m above 270 m CSF to ~7 mA/m below 270 m CSF, indicating a possible change in the depositional environment and perhaps associated with the observed sulfate reduction (see "Biogeochemical processes" in "Inorganic geochemistry") and production of methane (see "Organic geochemistry") below 270 m CSF. In contrast, NRM shows a stepwise increase in Unit V from 10 mA/m of the upper sandstone to 80 mA/m of the lower part of the sediments at 480 m CSF, which is the strongest NRM in Hole C0012A.

Magnetic susceptibility and NRM intensity variations through sedimentary units are closely correlated. Magnetic susceptibility values are generally ~2 × 10–4 to ~4 × 10–4 SI in lithologic Units II–IV but are significantly higher (>7 × 10–4 SI) for volcaniclastic sandstones in the lower part of lithologic Unit V. Like the NRM record, two sharp increases in magnetic susceptibility are present at 161 and 172 m CSF in Unit II, corresponding to tuffaceous sandstone and volcaniclastic sandstone at these depths, respectively. Magnetic susceptibility is ~2 × 10–4 SI throughout lithologic Unit I and gradually increases to ~5 × 10–4 SI at the bottom of lithologic Unit II. In lithologic Unit III, magnetic susceptibility shows a stepwise decrease from ~5 × 10–4 to ~2 × 10–4 SI at around 270 m CSF, which is synchronous with the change in intensity of magnetization. Magnetic susceptibility stays around 2 × 10–4 SI through lithologic Unit IV down to the middle of lithologic Unit V, where magnetic susceptibility abruptly increases associated with tuffaceous sandstone and volcaniclastic sandstone. The Q ratio for NRM after AF demagnetization at 10 mT (NRM10 mT) for the majority of samples (except for the sediments in Unit I) is <1, suggesting that the total magnetization of the sediments contain dominantly induced magnetization. As shown in Figure F35 (last column), the disappearance of several high peaks and large decrease in Q ratios between NRM before demagnetization (NRM0 mT) minus NRM10 mT (blue symbols) and NRM10 mT (red symbols) reveal the existence of the pervasive drilling-induced remanent magnetization (DIRM). The Q ratio peaks of various sandstones (triangle and square data points) in Figure F35 are the results of DIRM because Q ratio values after AF demagnetization at 10 mT (red symbols) fall in the same range as those for muddy sediments (Fig. F35). This observation suggests that the low-coercivity magnetic minerals in various sandstones carry an unstable remanence that is more susceptible to an external magnetic field.

Paleomagnetic stability tests

Remanent magnetization of discrete samples was investigated using stepwise AF demagnetization techniques in order to extract the primary component of magnetizations acquired at the time of deposition. Because of time constraints and because AF demagnetization appears to be more effective in removing the DIRM as seen in the shipboard paleomagnetic investigation for Site C0011, AF demagnetization was preferred over the thermal technique for sediments and rocks in Hole C0012A. A histogram of inclinations isolated from the 269 discrete samples is shown in Figure F36. Inclinations from these discrete samples are mostly concentrated at ±60° (Fig. F36; positive and negative peaks). These values are slightly higher but more or less close to the theoretically predicted value for the latitude of this site (±52.1°). Furthermore, the two polarity inclinations are not statistically different from each other, indicating that ChRM is a record of the paleomagnetic field close to the time of formation of the recovered sediments.

Figure F37 illustrates the magnetic stability behavior of several representative samples from various lithologic units. Figure F37A, F37B, F37C, and F37D shows examples of normal and reversed polarity intervals giving reliable results for polarity determination. The behavior demonstrates the removal of nearly vertical downward DIRM remanence after AF demagnetization at 10 mT and the isolation of a stable magnetization component that univectorially decays toward the origin of the vector plots (Zijderveld, 1967). As at Site C0011, we also noted that several samples from Hole C0012A display higher remanent magnetization stability (Fig. F37E, F37F), indicating the presence of a higher coercivity component, which cannot be completely demagnetized by AF demagnetization at 60 mT. Overall, the quality of paleomagnetic data throughout Hole C0012A is much higher than that at Site C0011, enabling the correlation between magnetostratigraphic data with the standard geomagnetic polarity timescale (GPTS) more confidently (see "Integrated age model and sediment accumulation rates").

Polarity sequence and magnetostratigraphy

We used ChRM inclinations from discrete measurements to define magnetic polarity sequences for Site C0012. Similar to the magnetic records at Site C0011, several relatively well defined polarity intervals have been identified in downhole magnetostratigraphic records. As with most applications of magnetostratigraphy, one of the greatest problems is correctly matching the observed sequence of magnetic polarity zones with the appropriate part of the GPTS. The polarity interval between 7.5 and 8.5 Ma could be well correlated with the standard GPTS unambiguously. For instance, shipboard biostratigraphic data suggest that sediments within 107.17–200.63 m CSF are older than 7.07 Ma but younger than 8.78 Ma (see "Biostratigraphy"). This information suggests that the observed normal polarity interval between 142.11 and 169.39 m CSF (Sections 322-C0012A-11R-2, 86 cm, through 14R-1, 88 cm) should correspond to the normal polarity Chron C4n.2n (7.695–8.108 Ma).

For the other part of Hole C0012A, the interpretation of the paleomagnetic polarity record is less straightforward. However, available biostratigraphic data from this site assisted in deciding which observed magnetic polarity zone or set of zones can be correlated with which magnetic chron on the GPTS (Tables T11, T6).

Integrated age model and sediment accumulation rates

Magnetostratigraphic and biostratigraphic (calcareous nannofossil) datum events are summarized in Tables T11 (Model A) and T6 (Model B), and the main features of the magnetostratigraphic interpretation along with the inferred biostratigraphic zones at Site C0012 are presented in Figure F38. The age determinations based on magnetostratigraphy are modeled in two ways. Although both of the models rely on biostratigraphy (nannofossil datum; blue curve), Model A (green curve) stresses on internal consistency of magnetozone sequences, whereas Model B (purple curve) follows the nannofossil datum as faithfully as possible.

From these age models we infer that a significant increase in sediment accumulation rate occurred in Unit I at ~130 m CSF from ~2 to ~6 cm/k.y. for Model A (green curve) or at ~110 m CSF from ~1 to ~6 cm/k.y. for Model B (purple curve). For both of these models, the lithologic Unit I/II boundary can be assigned an age of 7.8 Ma. These sedimentation rate values are not corrected for compaction.

For Unit II, the two age models (A and B) are the same to 175 m CSF (~8.3 Ma). Model A suggests successive changes in sediment accumulation rates from ~6 to ~1 cm/k.y. at ~180 m CSF, from ~1 to ~7 cm/k.y. at ~190 m CSF, and from ~7 to ~3 cm/k.y. at ~210 m CSF. Model A gives a Unit II/III boundary age of 10.2 Ma (Fig. F38A). The interpretation of Model A strongly relies on the recognition of an apparently long normal interval between 216.84 and 238.04 m CSF as long normal Chron C5n.2n (9.987–11.040 Ma), although there are considerable breaks between cores. This model correlates the successive sequence of relatively short polarity intervals found in the magnetostratigraphic record between 180 and 195 m CSF to Chrons C4Ar.2r through C4r.1r, which forms a straight line. A drawback of Model A might be the correlation of the dominantly reversed polarity interval between 197 and 207 m CSF with successive chrons between Chrons C4Ar.3r and C5n.1r, where polarity is dominantly normal in the middle and only the margins are reversed. This can be compromised by attributing the normal chron (C5n.1n) to the breaks of the core.

On the contrary, the discrepancy between Model A (green) and the nannofossil age model (blue) can be compromised by Model B (purple), which mostly follows nannofossil datum events with less emphasis on the consistency of the magnetic polarity zonations in the middle part of Unit II. Model B imposes the normal sequence between 213 and 240 m CSF to between polarity Chrons C4An and C5n.2n. The two relatively long reversed Chrons C4Ar.1r and C4Ar.2r could be imposed on the possible hiatus close to the base of Unit II (~220 m CSF) suggested by nannofossil (see "Biostratigraphy") and sedimentological features (see "Lithology"). Only the hiatus can explain the absence of reversed polarity intervals and relatively slow sedimentation rate of ~1.3 cm/k.y. for this period. One advantage of this model is the fairly uniform sedimentation rate of ~6 cm/k.y. between 111 and 213 m CSF. Model B gives a Unit II/III boundary age of 9.4 Ma.

For Model A, in the middle part of Unit III, the sedimentation rate increases slightly from ~3 to 4 cm/k.y. at ~250 m CSF and increases further to ~6 cm/k.y. at ~280 m CSF (Fig. F38B). The Unit III/IV boundary age can be estimated as 12.8 Ma according to the magnetostratigraphy, which is also consistent with the age model based on nannofossil datum. Model B merges to Model A at ~240 m CSF and both models are mostly consistent to the bottom of Unit III. The Unit III/IV boundary age for Model B is 12.7 Ma, which is slightly younger than Model A.

The age models for Units IV and V are less certain, especially for the lower part of the hole where core recovery was rather poor. Although there is a considerably reliable nannofossil datum at 417 m CSF (Zone NN6/NN5 boundary; Table T11), we seek for the age model (Model A) away from the control point trying to follow a straight line from Unit III down to the upper part of Unit V (430 m CSF) with a sedimentation rate of ~6 cm/k.y. Model A requires a significant increase in sedimentation rate to ~46 cm/k.y. below ~430 m CSF down to ~470 m CSF, which can be justified by imposing on the possible rapid deposition expected for the thick sandstones recovered for the interval (see "Lithology"). The age model infers a Unit IV/V boundary age of 14.4 Ma. The sedimentation rate decreases to ~8 cm/k.y. below 470 m CSF to 500 m CSF. Below 500 m CSF, the age model is less definite because of low recovery and a suspected hiatus inferred from nannofossil data.

Model B departs from Model A (green curve) from 270 m CSF downward. Model B favors an age of 13.5 Ma for the Unit IV/V boundary (Fig. F38B; purple text). Model B also infers a smoother sedimentation rate for the base of Unit V but a significantly increased sedimentation rate from Unit III (6.8 cm/k.y.) to Unit IV (10 cm/k.y.). According to these age models, the age of the thick sandstone units in the upper part of Unit V (430~470 m CSF) is estimated at 14.8 Ma (green curve, Model A) and 13.9 Ma (purple curve, Model B).

We emphasize that although there are options for the selection of age models, the ChRM of rocks from Hole C0012A is stable and of high quality as mentioned above. Thus, the possibility of misinterpretation of the polarity of the magnetozone is small compared with Hole C0011B, where interpretation of the magnetic polarity pattern itself was very difficult for the lower part of the hole. The missing intervals due to poor core recovery and a possible hiatus leave room for the possibility of several interpretations such as Models A and B for Site C0012. With the stability of ChRM, it is clear that possible future continuous piston coring at the site should unify the age model.

Paleomagnetic reorientation of the cores

Paleomagnetic declinations were used extensively for reorienting cored material. Table T12 lists the paleomagnetic directions used for reorienting coherent blocks cored by RCB for structural parameters relative to the geographic coordinate system (see "Structural geology"). Although clockwise rotation of ~30° is expected for the Shikoku Basin between 5 and 15 Ma according to Sdrolias et al. (2004), we assumed that the cored material did not experience significant rotation relative to the geographic coordinate system after the deposition. Further analysis to detect reliable magnetization components by distinguishing both viscous remanent magnetization and primary remanent magnetization could be conducted by using a superconducting quantum interference device magnetometer in combination with AF and thermal demagnetizations in shore-based studies.

Paleomagnetic characterization of basement rocks

Toward the end of Expedition 322, we collected four basaltic samples from lithologic Unit VII and measured them with detailed AF demagnetization experiments. Reversed polarity magnetization was observed on two of these basaltic samples (Samples 322-C0012A-53R-2, 35–37 cm, and 54R-1, 77–79 cm) after removal of the DIRM at 5 mT or less. Figure F39 shows examples of the demagnetization behavior of these samples. At least three components of magnetization are revealed by the AF demagnetization (Fig. F39A). DIRM is removed after AF demagnetization at 5 mT and shows a stable magnetization component (declination = 41.4°, inclination = –30°, and maximum angular deviation = 0.7°) decreasing toward the origin. However, after 15 mT AF demagnetization, the magnetization direction shows complicated behavior and starts to radiate out after 25 mT AF demagnetization. After 35 mT AF demagnetization, the magnetization turns around and moves toward origin with the magnetization direction antiparallel to the direction observed from 7.5–15 mT AF demagnetization (declination = –142.8°, inclination = –34.7°, and maximum angular deviation = 4.7°). The magnetization intensity at 5 mT is 15 A/m after removal of DIRM, whereas after 35 mT it is 52 A/m, which is unusually strong. Similarly, Sample 322-C0012A-54R-1, 77–79 cm, showed complex behavior through AF demagnetization (Fig. F39C, F39D).

Remanent magnetization measurements were conducted after the cruise with a cryogenic magnetometer in combination with an AF demagnetizer on sister specimens of Samples 322-C0012A-53R-2, 35–37 cm (Fig. F39B), and 54R-1, 77–79 cm (Fig. F39E). Neither specimen showed the complex demagnetization behavior with antiparallel magnetization observed during shipboard measurements. Sample 322-C0012A-53R-2, 35–37 cm (Fig. F39B), shows stable magnetization of reversed polarity decreasing toward the origin, which agrees with the magnetization component below 15 mT measured with the shipboard spinner magnetometer (Fig. F39A). The magnetization intensity shown in Figure F39B is much lower than that in Figure F39A. This may be explained by the fact that the sister specimen was taken from the outer rim of the minicore (Sample 322-C0012A-53R-2, 35–37 cm), which shows altered color. Sample 322-C0012A-54R-2, 77–79 cm (Fig. F39E), shows normal polarity, which does not agree with the magnetization component of Sample 54R-1, 35–37 cm (Fig. F39B), measured with the shipboard spinner magnetometer. The relatively low coercivity component of Figure F39E and the position of the sister specimen (outer rim of the minicore) may indicate that the magnetization component is mainly composed of DIRM. Another sister specimen was taken between the above two specimens for Sample 322-C0012A-54R-1, 35–37 cm, and measured with a cryogenic magnetometer with stepwise thermal demagnetization up to 600°C (Fig. F39F). The magnetization shows reversed polarity, which agrees with the magnetization component below the AF demagnetization field of 23 mT (Fig. F39D). Specimens from two other samples taken as shipboard paleomagnetic samples were subjected to both AF and thermal demagnetizations and showed relatively stable maganetization of reversed polarity. In conclusion, the basaltic basement rocks taken below the sediment at this site may retain reversed polarity magnetization as the primary magnetization. Table T13 is the summary of the paleomagnetic directions obtained for basaltic samples measured on board the ship and onshore either with AF or thermal demagnetization.

Although the origin of the antiparallel magnetization observed only for the specimens measured with the shipboard spinner magnetometer cannot be understood completely at present, the significant increase in the magnetization intensity after AF demagnetization antiparallel to the original magnetization combined with unstable behavior in-between may suggest the presence of self-reversed remanent magnetization. This phenomenon may be explained as a special case for some magnetic minerals that were in the limited place of only two samples, which cannot even be observed in the sister samples from the same minicore. Another possibility may be that the phenomenon occurred only for the samples taken just after drilling. The magnetic mineral characteristic of the phenomenon disappeared several weeks after the cruise. This could possibly be attributed to self-reversal of pyrrhtite during AF demagnetization, which was already reported by Bina et al. (1999). A third possibility is that the phenomenon was caused by the malfunction of the spinner magnetometer that occurred on the same day basaltic basement rocks were measured. This may be justified by the noisy behavior during AF demagnetization measured with the shipboard spinner magnetometer for Samples 322-C0012A-55R-1, 23–25 cm, and 57R-1, 44–46 cm.

Anisotropy of magnetic susceptibility

We were able to measure AMS with the Kappabridge KLY 35 for minicore samples from Cores 322-C0012A-4R through 20R, with generous help provided by the shipboard scientific party. For the lower part of Hole C0012A, limited samples were taken from tuffaceous/volcaniclastic sandstone layers and sand/silty sandstone layers, and AMS measurements on these samples were performed after AF demagnetization. As stated in the "Site C0011" chapter, based on the results of the comparison experiment, AMS after AF demagnetization could be used for sandstones, although it was recommended that AMS should be measured before AF demagnetization (e.g., Jordanova et al., 2007).

Figure F40 shows the AMS parameters measured for various lithologies in Hole C0012A. Kmin inclinations (Fig. F40A) are dominantly steeper than 60° except for a few horizons. For example, four samples at 90 m CSF (from Sections 322-C0012A-5R-1 through 5R-4) have Kmin inclinations ~0. Shape parameter (T) also indicates prolate ellipsoid (Fig. F40C). This observation may indicate possible vertical stretching of the core because of RCB coring of soft sediments, which was also visually observed as vertical striations (see "Lithology"). An interesting feature is that a shallower Kmin inclination (~15°) is present at the Unit I/II boundary at ~150 m CSF (Fig. F40A). The anisotropy degree parameter (P′) increases at the base of Unit II from ~1.02 to ~1.10 and also at bottom of the Unit V (Fig. F40B). These increases in anisotropy degree may be related to the change in the sedimentary depositional environment.

The Kmin inclination values for the four basaltic rock samples of lithologic Unit VII are close to zero, and shape parameter is negative (i.e., prolate). The characteristics of these AMS parameters could indicate the presence of linear fabric caused by basaltic flow during the basalt emplacement or by late-stage extensions of the seafloor.