We made pass-through magnetometer measurements and magnetic susceptibility measurements on all split-core archive sections at 2-cm intervals. In order to isolate the characteristic remanent magnetization (ChRM), we subjected the cores to alternating-field (AF) demagnetization. The number of AF demagnetization steps and the peak-field intensity varied depending on lithology, the natural remanent magnetization (NRM) intensity, and the amount of time available for analysis. On average, sediment half cores were demagnetized using three AF steps in addition to the measurement of NRM. Some half cores were demagnetized using as many as six AF steps. The maximum applied field ranged from 20 to 60 mT. In order to assess the half-core data and to identify magnetic carriers, we demagnetized several discrete samples using both AF and thermal techniques. We analyzed the results in Zijderveld and stereoplot diagrams, and, where possible, we calculated the ChRM direction using principal component analysis (Kirschvink, 1980). In addition, we determined magnetic susceptibility on all whole cores from Site 1276 at 2.5-cm intervals as part of the multisensor track (MST) analysis and we measured split-core sections at 2-cm intervals with the point-susceptibility meter on the archive multisensor track (AMST). The two magnetic susceptibility data sets compare well with one another.
Paleomagnetic data obtained at Site 1276 exhibit considerable variations in demagnetization behavior among various lithologies, with the most common features summarized as follows. Variations in magnetic susceptibility broadly correlate with the variations in NRM intensity (Fig. F143). A pervasive remagnetization imparted by the coring process is commonly encountered, as was noted during previous legs (e.g., Gee et al., 1989; Zhao et al., 1994). This remagnetization is characterized by NRM inclinations that are strongly biased toward steep downward inclination, and it can be removed at initial stages of demagnetization. An example is shown in Figure F144, where the NRM inclinations observed in Section 210-1276A-35R-5 are biased toward steep positive values (Fig. F144A). With 30-mT demagnetization, we observed a shift toward shallower inclinations (Fig. F144A) and a significant decrease in intensity (Fig. F144B). Although the maximum level of AF demagnetization on the ship's cryogenic magnetometer was not always able to remove these overprints, ChRM directions can generally be determined from the pass-through measurements. Examples of good-quality AF demagnetization results are shown in Figure F145.
In general, we find that the magnetic properties of sediments recovered from Hole 1276A correlate with the lithology (see Fig. F143). Below we describe the magnetic characteristics of the main lithologic units.
In the middle to upper Eocene Cores 210-1276A-2R through 8R, variably burrowed, varicolored mudstones and claystones have low NRM intensity and low magnetic susceptibility. NRM intensities generally range between 3.7 x 10-5 and 3 x 10-1 A/m (mean = 3.5 x 10-3 A/m). Magnetic susceptibility averages 4.3 x 10-5 SI units. A few discrete peaks of higher NRM and susceptibility values could in some cases be tied directly to the visible presence of pyrite. By analogy to previous studies of lower Oligocene to middle Eocene sediments on the Iberia margin, we speculate that primary magnetite in these claystones and mudstones has suffered diagenetic reduction to pyrite, thereby destroying the primary remanent magnetization. For these cores, reliably defining the primary remanent magnetization was impossible.
Sediment recovery in the upper Paleocene to middle Eocene section was somewhat lower than in Unit 1. Although the NRM intensity of the unit is still low (ranging from 1.4 x 10-4 A/m to 2 x 10-1 A/m; mean = 4.5 x 10-3 A/m), we were able to define the ChRM direction from a few intact cores (see Fig. F145). Magnetic susceptibility in this interval ranges from 6.9 x 10-4 to 6 x 10-5 SI units (mean = 3.1 × 10-5 SI units).
The lower Campanian to lower Paleocene Unit 3 consists of mainly claystone and mudstone with the exception of Cores 210-1276A-24R and 25R, where marlstone and grainstone are present (see "Lithostratigraphy"). The claystones and mudstones have relatively high NRM intensity and magnetic susceptibility, caused by the presence of numerous dark burrowed beds that have relatively high concentrations of magnetic minerals. The mean values for NRM intensity and magnetic susceptibility are 6 x 10-2 A/m and 2.4 x 10-4 SI units, respectively. One interesting observation is that the AMST and MST magnetic susceptibility data and pass-through cryogenic magnetometer NRM intensity records show an anomalous peak for sandstones in interval 210-1276A-15R-4, 132-142 cm (930.82-930.92 mbsf) (see Fig. F143). This sharp increase in magnetic susceptibility is present at approximately the lower/upper Paleocene boundary. The cause of this susceptibility high is not clear and awaits determination by shore-based studies. In addition to this susceptibility high, the interval between 1013.40 and 1018.72 mbsf (Sections 210-1276A-24R-2, 30 cm, and 24R-5, 112 cm) in Unit 3 displays high NRM intensity and magnetic susceptibility values (Fig. F143), corresponding to marlstones and grainstones.
These sediments are mainly sandy mudstones and muddy sandstones that slowly accumulated during Turonian to latest Santonian time (see "Biostratigraphy"). The characteristic color of Unit 4 is moderate brown, which is perhaps related to the presence of fine-grained iron oxides (see "Lithostratigraphy"). Accordingly, the mean NRM intensity (3 x 10-2 A/m) and magnetic susceptibility (1 x 10-4 SI units) are relatively high. A strong drilling-induced overprint is present throughout Unit 4, which severely limits paleomagnetic work.
Sediments in lithologic Unit 5 are uppermost Aptian(?) to lower Turonian claystones and mudstones that have low NRM intensities (mean = ~2.5 x 10-3 A/m) and magnetic susceptibilities (mean = ~8 x 10-5 SI units). There are more significant variations in susceptibility in Unit 5 than in Unit 4. Characteristic susceptibility peaks reflect carbonate and sandstone layers, and troughs correspond to green and gray claystone and mudstone. The susceptibility peaks of carbonates may reflect an iron component in these rocks, most likely in siderite. Finely laminated claystones (black shales) are well developed and numerous throughout Unit 5. The magnetization of these dark claystones is typically an order of magnitude (in some cases even two orders) weaker than that of light-colored sediments. Although the ChRM directions can still be defined from these black shales using stepwise demagnetization (Fig. F145), we suspect that the magnetic remanence is secondary. Interestingly, some muddy sandstones below 1125 mbsf in Sections 210-1276A-35R-7 through 36R-6 (~1135 mbsf) show high peaks in both NRM intensity and magnetic susceptibility (Fig. F146). The high-intensity peaks were maintained even after 30-mT AF demagnetization, suggesting that it is unlikely that they are caused by drilling-induced remagnetization.
Remanent magnetization of discrete samples from these five lithologic units was investigated using stepwise AF or thermal demagnetization. In most cases, the steep downward component of magnetization imparted by the coring process is easily removed by AF demagnetization. Thermal demagnetization also successfully removed this drilling-induced component. Most samples show unblocking temperatures between 350° and 550°C, indicating that titanomagnetites are likely the main magnetic carriers in the samples.
Drilling in Subunit 5C yielded diabase in one sill and one deeper sill complex. We performed detailed AF demagnetization on all coherent diabase pieces that could be oriented unambiguously with respect to the top of the core (Figs. F147, F148). All pieces contained a vertical secondary component, most likely induced during drilling. In almost all cases, we were able to remove this secondary magnetization by 10- to 20-mT AF demagnetization and by isolating the ChRM direction at higher fields (e.g., Fig. F149). The inclinations of the ChRM direction for the two diabase sills are all positive. The simplest explanation of the positive inclinations is that they represent normal-polarity magnetization, probably acquired within the Cretaceous Normal Superchron (CNS). This interpretation is compatible with time of emplacement inferred from other shipboard studies (see "Igneous and Metamorphic Petrology" and "Physical Properties"). In order to test inclinations determined from the half-core data, several discrete samples from the two sills were thermally treated and measured. The results (squares in Figs. F147, F148) are in good agreement with the half-core measurements. However, more complete shore-based demagnetization and statistical analysis of the data are needed.
A problem was encountered during half-core measurements on diabase pieces because the diabase cores are broken into segments. Because the short pieces have no azimuthal orientation relative to one another and the measurement region of the pass-through magnetometer (~15 cm) is comparable to piece size, we observed significant magnetic interference between the pieces; this leads to artificially low NRM intensities and varying inclinations from piece to piece. For this reason, we omitted data points for short pieces in Figures F147 and F148. We used the following criteria to select only reliable inclination values. The pieces had to be at least 7 cm long, the maximum angular deviation of the principal component analysis had to be <3°, and the pieces had to be homogeneously magnetized. Analyzed pieces that fulfill these criteria are summarized in Table T14. Significant changes in inclination and intensity values in the upper sill (Table T14) suggest that this sill may contain two subunits.
Although the NRM intensities and inclinations are quite variable, we identified distinct differences between the upper and lower diabase sills. The lower sill (Sections 210-1276A-99R-1 through 99R-6; ~1719-1725 mbsf) is generally more weakly magnetized: the average intensity and magnetic susceptibility are 0.8 A/m and 3.2 x 10-3 SI units, respectively (Fig. F148). The upper sill (Sections 210-1276A-87R-6 through 88R-7; ~1613-1623 mbsf) is more magnetic and has average intensity (>3.8 A/m) and magnetic susceptibility (5.6 x 10-2 SI units) values that are consistently higher than the lower sill. The result is that the average Königsberger ratio, Q (defined as the ratio of the NRM intensity to the induced magnetization in the local Earth's field) (see "Paleomagnetism" in the "Explanatory Notes" chapter), for the upper sill is ~1.9, whereas Q values in the lower sill have a mean of 9.5 (Table T14).
As mentioned, shipboard pass-through measurements did not yield reliable primary remanent magnetization from sediments in Unit 1, so the magnetic polarity stratigraphy of the lower to upper Eocene cores cannot be established at present. Based on biostratigraphic data in Unit 2, however, we were able to tentatively correlate certain elements of the observed magnetic polarity interval in upper Paleocene to middle Eocene sediments with the geomagnetic polarity timescale. In particular, the reversed polarity interval in Section 210-1276A-9R-5 may represent Chron C21r of the early middle Eocene. However, because only this section displays an indisputable reversed polarity, the magnetostratigraphic potential for Unit 2 is obviously limited.
Shipboard micropaleontological study suggests that the K/T boundary is in Section 210-1276A-21R-4 of Unit 3 (see "Biostratigraphy"). From the magnetostratigraphic point of view, however, no well-defined reversed zone can be identified at the inferred boundary, even though we expect to find one (i.e., Chron C29r). Failure to detect a reversal could be due to severe drilling-induced overprint. Detailed demagnetization work on discrete samples from this unit is needed.
The biostratigraphic ages for Unit 4 suggest that there should be at least one reversed polarity magnetozone (Chron C33r; 83-79 Ma) in this cored interval (Turonian-uppermost Santonian). However, the recovered sediments are all normally magnetized. It is possible that our sampling interval (2 cm) and the size of the region sensed by the magnetometer are too large to detect fine-scale magnetization changes. Shore-based studies of discrete samples may help to identify small-scale features not discernible in our pass-through measurements. The sedimentation rate is ~2 m/m.y. in this interval; thus, Chron C33r should occur over ~8 m of core. However, Cores 210-1276A-25R through 28R are barren of fossils and it is possible that a hiatus is present and Chron C33r sediments are missing.
The age of the succession in Unit 5 is latest Aptian(?) to early Turonian, which is within the CNS, and, accordingly, cores in Unit 5 are all normally magnetized. However, we cannot exclude the possibility that the normal polarity also could represent late-stage remagnetization in a normal field.
Since the beginning of DSDP in 1968, paleomagnetists have been complaining about the presence of a magnetic remanence that is thought to be acquired during drilling and/or recovery of the core (e.g., see Stokking et al., 1993). Numerous investigations of the magnetic fields of the various tool components used in coring have been conducted previously on the advanced piston corer and extended core barrel (e.g., Fuller and Garrett, 1998; Fuller et al., 1998). Results suggest that there can be very strong local fields in the bits and in the pipe (e.g., Stokking et al., 1993; Herr et al., 1998; Fuller et al., 2000). During Leg 210, we used an RCB that was half nonmagnetic (lower) and half normal (upper) steel for Cores 210-1276A-1W through 57R (this type of core barrel was tried during Leg 209, but there was no documentation of its effect on the magnetization of cores). Starting with Core 210-1276A-59R, we used a full nonmagnetic RCB for the first time. The nonmagnetic barrels were made with 15-15-LC steel, which is a nonmagnetic material commonly used in the petroleum industry. All the odd-numbered cores from 210-1276A-59R through total depth were retrieved with this nonmagnetic core barrel, whereas even-numbered cores were recovered with a regular RCB (magnetic). This allowed us to test the effects of the magnetic vs. nonmagnetic RCBs on remanence intensities and directions. In this experiment, we were fortunate that sedimentary rocks from Cores 210-1276A-58R through 68R in Unit 5 have quite uniform lithology and composition and the sediments all have normal polarity (i.e., there is no complication due to a reversal sequence). As shown in Figure F150A there is a clear indication that cores obtained with a regular core barrel are more strongly magnetized than those with the nonmagnetic core barrel. After 20-mT demagnetization, the difference in remanence intensity clearly remains but it is generally reduced compared with the NRM (Fig. F150B).
Although we focused our comparison on the overall remanence intensity, we also compared inclinations (Fig. F151). Curiously, there appears to be a tendency toward anomalously steep negative inclinations (both before and after 20-mT demagnetization) at the top of every core, regardless of which core barrel was used. A closer examination of the remanence-intensity record reveals that there are intensity spikes exactly at the same positions as the steep inclinations (Fig. F151A). This suggests that some coring component (or perhaps rust from inside the drill pipe) that is strongly magnetized with upward inclination can remagnetize the top portion of the core material. At this time we can only speculate, but it is also possible that the remagnetization occurs upon core retrieval (e.g., it may be affected by the "stabber bar" and/or its swivel connecting to the core cable).
Understanding the magnetic effects of core barrels is very important for the new Integrated Ocean Drilling Program. Our results demonstrate that strong magnetization of conventional core barrels provides a potential "source field" in which the sediments could acquire barrel-induced magnetization. There is no longer any doubt that standard core barrels should be replaced with the nonmagnetic core barrels during future drilling legs, especially those targeted for paleoceanographic investigation at low latitudes.